articles are: (I) languageand the Habit of - ERIC

DOCUMENT RESUME

ED 022 757TE 000 646

By-Stauffer. Russel G.. Ed.LANGUAGE MO TIC HIGHER THOUGHT PROCESSES.

National Conference on Research LI English. Champaign. IN.

Pub Date 65Note-72p; Reprint of articles originally published in Elementary English; v42 n4-5 Apra and May 1965.

Available from-National Council of Teachers of English. 508 South Sixth Street. Champaign. M. 61820 (Stock

No. 13906. HC $125).EDRS Price PlF -$OSO HC Not Available from EDRS.

ors-COMPOSITION (LITERARY). *CONCEPT FORMATION. ELEMENTARY EDUCATION. *ENGLISH

INSTRUCTION. EXPRESSIVE LANGUAGE. GIFTED. *LANGUAGE, LANGUAGE SKILLS. LINGUISTICS. LITERATURE.

READING INSTRUCTION. SECONDARY EDUCATION. SELF EXPRESSION. SEMANTICS, *THOUGHT PROCESSES .

This collection of seven significant articles on higher thought processes stresses

the important relationship of these processes to the teaching of language skis. The

articles are: (I) language and the Habit of Credulity" by 12ussell G. Stauffer. (2)

'Research on the Processes of Thinking with Some Appkations to Reading by David

H. Russet (3) 'Form Consciousness. an Important Variable in Teaching Language.

Literature. and Composition" by James R. Squire, (4) "The Teaching of Thinking by Hicla

Taba. (5) t Formation in auldren by Harriett Amster. (6) "The Ouality of

Oualification" by Murray S. Miron. and (7) "Expressive Thought by Gifted Children in the

Classroom" by James J. Gallagher. (DL)

I

t041

0Aw

togo

9

ii

1

4

vsi

40

44

.. _....

4a 1

11

.

vlo

vb..

.OP

".

4.

0 141° 1

0 8

a

44

0 if0 i

uta

1?

amili.

vil

4II

,1

a1w

Ii.1It

40

OSII.de

a

.4

rrqtrep.rgp-n., 777,"cm"-nn, r s

The articles in this bulletinwere originally published in

the April and May 1965 issues of

ELEMENTARY ENGLISH(XLII, 4 and 5)Copyright 1965

byNational Council of Teachers of English

WILLIAM A. JENKINS, Edit Of

"PERMISSION 10 EPROM INIS COM MEDANIMAL BY MICROFPE OLIXIMS IEEN NAMEDBy We TS moiL AfaIrffTO ERIC NO MIGNIIZATIONS OPERATING UWEAMIENS WIN INE IL S. OFFICE OF EDUCATION.ROM REPRODUCTION OUTSIDE INE ERIC SYSIDAMIRES PERMISSION OF ME COPYR IONT MEL"

Ii

Contents

Why This Bulletin? v

Language and the Habit of Credulity 1

Russell G. Stauffer

Research on the Processes of Thinkingwith Some Applications to Reading 9

David H. Russell

Form Consciousness, an Important Variable in TeachingLanguage, Literature, and Composition 19

James R. Squire

The Teaching of Thinking 31

Hilda Taba

Concept Formation in Children 40

Harriett Amster

The Quality of Qualification 50

Murray S. Miron

Expressive Thought by Gifted Childrenin the Classroom 56

James J. Gallagher

111

i

A

The National Conference on Research in English

is an organization of one hundred

active members qualified to conduct

scientific research in English.

The purpose of the organization is to

stimulate and encourage research

in the teaching of English

and to publish results of

significant investigations and of

scientific experimentation.

1965 President

JEANNE CHALL

Harvard Graduate

School of Education

Cambridge, Massachusetts

iv

z

1

3

I1:

1i

Why This Bulletin?Ushered into the educational arena of

our society in recent years have been avariety of proposals aimed at both quickand permanent changes. The currents ofsome of these movements have been asstrong as their shoals have been treacher-ous. Some among us have been seriouslyintimidated; some have panicked; somehave stood firm and, as professional peo-people, have sought a high level of com-petence.

Our nationwide concern that all be edu-cated has caused us to realize more clearlythan ever &at, where human learning isconcerned, there is no one best way ofteaching or learning. We also see moreclearly that how children will share ourliterary heritage depends to a considerabledegree on where they are born, how theygrow, how they see the why or worth ofthings, who loves them. Looking at theprivileged and underprivileged among ushas caused us to reexamine the meaningof literacy and to note that intellectualslums exist at all social levels.

Facing up to our intellectual slums isrequiring a new look at how thinking mancatalogs his universe and attempts to orderthings. The respect and forbearance of thethinking mind has taken on new dimen-sions. This has helped us in our reexami-nation of American education and what itdares to do. Frank C. J ennings puts it thisway in his refreshing This Is Reading:

First, it is based upon the clear democraticnotion that all children, whatever their back-grounds, whatever their physical and intellec-tual limitations, be helped to become the verybest kinds of persons. By best is meant thatthey achieve the skills and the knowledge thatwill help them to realize as much as possibleof their grandest dreams. This concept ofAmerican Education is not one of leveling downall people to a kind of social least commondenominator. It is the giving to everyone ofthe will and the desire to reach beyond hisgrasp.'

'Frank G. Jennings, This Is Reading (New York:Bureau of Publications, Teachers College, Colum-bia University, 1965), p. 113.

v

At another point Jennings says: "Whatwe say learning is will determine how weteach. What we conceive learning to bewill depend on what we think mind is.Thus if the mind is merely a collection offaculties, training is required. . . . But ifmind is a function, as most contemporarypsychologists would insist, then learning isa far more complex affair."2 Because wethink with language it is without a doubtthe most momentous product of the humanmind. It reflects man's ability to abstractand to symbolize, to deal with time andplace and conditions, to make comparisonsand to have preferences, and to socialize.Language is the prime tool of the scholar.

Requirements cf a scholar are interest,enthusiasm, an appetite for ideas, and theintellectual courage and integrity necessaryto understanding knowledge and the man-agement of experience. How scholarshipof this caliber is to be attained in a societythat deliberately tries to teach skills orknowledge to others is not a secret. A clearand vigorous statement in this regard wasprepared recently by the California Teach-ers Association's Commission on Educa-ional Policy. After prolonged delibera-

tions the Commission produced a docu-ment that defines the dimensions of excel-lence in education. There are six clusters ofideas in the document which are not to beviewed as constituting an orderly progres-sion or any sort of continuum but ratherto be viewed as being mutually supportive.The six dimensions are:

Capacity for inquiryProblem-solving competenceCommunication and computation skillsFamiliarity with organized disciplinesCultivated enjoymentsDemocratic commitment3

The first three dimensions have the fla-vor of processes, and the other three bear'Ibid., p. 42.'California Teachers Association, Commission onEducational Policy, Dimensions of Excellence inEducation (1705 Murchison Drive, Burlingame,California: CTA Publications Supply, 1965, 25cents).

,

the quality of outcomes. The Commissiondeliberately placed at the head of the listthe dimensions comerned with processesbecause pedagogical instruction should callforth these processes. The subject matterdisciplines should feed these processes inthe process of the student's becoming fa-miliar with knowledge. The curriculumshould engender productive thinking. In-structional situations should be created insuch a way that students can be taught toarrive at decisions by honest and opendeliberation of many choices rather thanby indoctrination of a single choice.

While each dimension plays a signifi-cant role, it is not by accident that theskills of communication and computationthrough language, numbers, and othersymbols have to be regarded as the "fun-damentals." Just as the use of language isessential to communication, so, too, is itessential to inquiry itself. Productive think-ing is dependent upon clear and accuratecommunication.

Reports on a conference concerned withthe basic cognitive processes in childrensuggested that cognitive functioning re-quires the development and use of verbalmediators as well as the selection of di-mensions for stimulus categorization.4 Alsosuggested is the idea that the process beactive, questioning, testing, and inventing;and that it result in information-producingbehavior. The "close hard look" and the"what is it?" response apparently play a

'Basic Cognitive Processes in Children, eds. JohnC. Wright and Jerome Kagan (Monograph of theSociety for Research in Child Development, SerialNo. 86, Vol. 28, No. 2.; Chicago: The Universityof Chicago Press, 1963).

vi

major role in even the most primitive formof cognitive functioning. Succeeding stagesof development apparently involve the de-liberate use of verbal labels and mediatorsand the wise selection of dimensions andstrategies so that purposeful behavior andactive information processing may result.

All this leads to the conclusion that theteacher is even more intimately involvedin the learning process than may some-times be thought. The teacher must arrangethe learning environment so that studentsare provoked to search for and defineproblems, or to ask questions. Further-more, it is the teacher s responsibility tosee to it that the problems defined are sig-nificant; that answers are accepted as pro-visional, and lead to questions of furtherimportance and to generalizations for nextsteps. "Students can be instructed in appre-hending cause and effect, the relationshipsof situational factors, the benefit of organi-zation and system, the dependence uponaccuracy, the need for examining alterna-tives, and the requirement for evaluationof decisions."5

Language emerges from all this as theessential medium in formal education. Ameaningful use of the skills of reading,writing, speaking, and listening is requiredin our schools today more than ever. Asteachers become skilled in bringing out oftheir instruction high competence in theseskills, excellence will result. It is towardthis end that this publication has been pre-pared.

RUSSELL G. STAUFFERDirector, The Reading Study CenterH. Rodney Sharp Professor of Education

'California Teachers Association, Commission onEducational Policy, op. cit., p. 10.

RUSSELL G. STAUFFER

Language and the Habit ofCredulity

It is commonly accepted that languageis the instrument that, better than anyother, enables persons both to develop andto participate in their culture. Granted thatthis is true, it follows that language is asymbol-system of vital importance to eachindividual and to his society.

When an infant discovers that a cry canproduce certain desired results, he hasmade a start on the long road towardacquiring a language and being a part ofa collective communication system. This isespecially so vcrhen the child begins to real-ize that he can participate in the languagecommunity as both a producer and receiverof symbols (4, p. 24 ). The knowledge thatthis presages the difference between manand animal is indeed vivifying. As a pro-ducer and receiver the child differs sharplyfrom the Pavlovian dog who is also re-sponding to a symbol (the bell) with aresponse (salivating ) when the chain ofevents is initiated by man. And this dif-ference is strikingly apparent\ when onenotes that there can be tio reciprocal rela-tion between the dog and the man.

This ability to learn and use languageis different from the power among insectso perform feats of amazing skill. A Sphex

wasp, for instance, has the ability to per-form a delicate and exacting nerve opera-tion on an Ephippiger grasshopper. The

Dr. Stauffer, Director of the Reading Study Center,is the H. Rodney Sharp Professor of Education atthe University of Delaware, Newark.

Elementary English, XLII ( April 1965), 362-369.

1

operation paralyzes the legs, leaving thegrasshopper helpless and a ready supply ofliving food for the wasps to be hatched.

"Once the paralysis is accomplished thewasp drags the victim to its nest by anantenna. When, however, the antenna iscut off, the wasp is baffled, and can't con-ceive of any other way to move the grass-hopper. It has been concluded then thatwhile insects possessed highly specializedinstinctive skills, which they didn't need tolearn by teaching or example, they com-pletely lacked the ability to reason" (16,p. 119).

Unlike the Sphex wasp but like thePavlovian dog, the child makes associationsbetween a symbol and an experience. Butunlike the Pavlovian dog the language sym-bol- are not artificial products of an ex-perimental situation. Most of the first sym-bols a child learns to isolate and use areacquired in a first-hand experience situa-tion. Usually these learnings are related toconcrete things which are perceived direct-ly through the sensesmilk, mother, doll.Whether children first learn to use wordsat age one or at age two and a half thelearning demands are the same. Each childmust associate, select, use, and remember.

Sooner or later children discover thatnot all symbols refer to concepts on theobject level. While it is thought that be-tween the age of eighteen to twenty-fourmonths children live predominantly in thepresent, some ability to project into thefuture is developing. Ames reports that"Words indicating present come in first,

2 LANGUAGE AND THE HIGHER THOUGHT PROCESSES

then words indicating the future, and fi-nally those indicating the past. Thus 'to-day' ( 24 months) precedes 'tomorrow' (30months ) which in turn precedes 'yesterday'(36 months )" (3, p. 122).

It is at about this point in the child'slong, slow progress toward language matu-rity that he attains an awareness, eventhough vaguely, of two basic conditionsFirst, he begins to realize that all the thingsin his world have names and that folksabout him know these names. The periodis sometimes affectionately labeled as the"What's this?" age. Second, little by littlehe learns to deal with words that representrelative and indefinite ideasthe little dog,the little boy, the little car, the little house.Puzzling as it may be, he gradually learnsto use little discriminately and in so doingtakes a big step on the road toward under-standing multiple meanings and figurativelanguage.

As the child learns to deal with wordslike little he learns to deal with incon-gruities of meaning and use. At first heuses the word appropriately in differentcontexts under different communication de-mands almost as if by intuition. Somehowhe learns to recognize a common elementin the different situations in which theconcept of little fits. Not being articulate,though, about the many specifics that enterinto each use of little (little toy-little storm)he operates in part on knowledge and inlarge part on extrasensory intuition. Herehis intuition rests in the degree of constancyassociated with the variables of little.

Even so, when dealing with concepts oflittle, referents usually are a part of thesensory world and can be reexamined.Thus while there is present a certainamount of blindness and a certain demandto accept by intuition or faith, the degreeand amount is not as great as when thechild deals with a concept like tomorrowbecause tomorrow never comes. Yet as

Ames and others ( 12) point out, childrenlearn to deal with the concept tomorrowat an early age. Somehow they must rec-ognize some of the attributes of the con-cept tomorrow. Certainly, though, theremay be judged to be present a greater de-gree of blindness and a greater demandon intuition than when dealing with con-cepts such as little.

It seems then that the intuition demandsrequired when dealing with concepts likeshoes, little, and tomorrow can provide thereadiness for dealing with concepts likedemocracy, peace, and eternity. Could itbe said therefore that the beginning offaith and of credulity originates in the"word" or in a symbol?

Today, tomorrow, then yesterday. Onemight well auger here the hope that springseternal and use this early learning patternas a revelation of man's visionto lookahead from today and with increased ma-turity to have that look ahead be temperedwisely by the past. As Ames (3) goes on tosay, the age of three to five brings with itmuch greater projection into the future,until by age eight even extremes of timespan can be handled adequately.

Gradually, though, the veil of an individ-ual's and his culture's emotional, intellectual,and spiritual world creates gossamers thatmay lead to stereotypes, prejudices, andsentiments as well as to convictions andbeliefs. Now he learns the special idiomsof the teens, the slogans and stock phrasesof the propagandist and the publicityagents, and the stories and puns of theadults. Shifts of meaning are subtle andvaried and their acuteness may often gounrecognized. Fortunately, though, over allis the innateness of laughter and the rangeof things laughed at (11, p. 86). And asStephen Leacock believes "humor in itshighest meaning and furtherest reach .. .

finds its basis in the incongruity of lifeitself.... and becomes the contemplation

LANGUAGE AND THE HABIT OF CREDULITY

and interpretation of our life" (8, p. 15).Then as a person deals with the verbalconventions of his time, he discovers thathumor can be the wedge that amplifies theneed to adjust concep+s to the ever-shiftingrealities of experience.

Children love fun. They seem to be nat-ural pranksters unless, as one author putsit, it's been spanked out of them. The care-free, spontaneous, enthusiastic laughter ofchildren uninhibited by the sober conven-tions of adult society has through the agesbeen a favorite genre of the poets.

From games and jokes and pranks to funwith words seems an easy step and usuallyoccurs early in the word-life of children.Here their imaginations are undiminished.From nursery rhymes and ditties to Winniethe Pooh to Paul Bunyan is easily accom-plished as they disarmingly toy with theprotective masks of phantasy and unreality.They catch shifts of meaning that occur inchildren's puns, in simple name calling like"you're a grandmother," or 'you're a ham-burger with onions." All this requires asense of intuition and of credulity. Humorseems to represent the early developmentof a language thermostat that permits themto keep their language habits on a con-trolled level and thus avoid the emotionalbreakdowns that stem so frequently fromcommunication misunderstandings. A goodsense of humor has for a long time beenreferred to as an emotional stabilizingasset.

So as children matuie and achieve aclearer grasp of the common and uncom-mon concepts in their world of culture, thesophistication will surely remove the in-nocence from their eyes. And it is our dutyto see to it that this change not be marredby the cold brittleness of bitter disillusion-ment. Rather we must see to it that thematuring be activated by the subtle vibra-tions resulting from ^. love for the creative

3

and the substantial resources of an outlookon life that is built on faith.

And so the individual learns that 'thesymbols of language are slippery things.They d a not stay put: they carry multiplemeaniugs, which shift from time to timeand differ from place to place; their os-tensible references overlaid with meaningsof a non-rational character. It ( language)is the result of man's collective quest forexpression, and it therefore reflects thecultural value, the changing attitudes andintentions and preoccupations by whichman lives" (4, p. 17).

And so, too, man learns to accept mean-ings of things not experienced or if expe-rienced not examined, or if examined notgeneralized. He learns to project on lan-guage a credulity that ranges from thenaive to the sophisticated. First, he acceptsnames without question. For a while hetends to apply them indiscriminately on aclass basis calling all children "boy" or allanimals "bow-wow." Even as an adult hetends to accept labels such as salt, grain,and love without a check on their deriva-tion and history or the reasons why theirmany uses developed and how they tendto be related. He learns to associate manymeanings with tomorrow without perhapsever being quite ready to deal with, 'To-morrow and tomorrow and tomorrow,comes ever and anon after today." Hethinks of certain words as representingdefinite concepts and uses them as if theydid until one day perchance he is asked toelaborate on the meaning of "How far is amile?" Then he discovers that saying 5,280feet or 1,700 yards adds only limited clar-ity. It says nothing about the kind of mile,i.e., a mile high, a mile deep, a mile underwater, a mile wide canyon; the time of amile, i.e., a mile a minute, a four minutemile, a two week mile; or reactions to a

°Author's inn&

4 LANGUAGE AND ME HIGHER THOUGHT PROCESSES

mile, how tired is a runner after a mile, amountain climber after ascending a mile,

a golf player after playing six holes, andso on. He now sees this "definite" termmik being overlaid with various credulitiesand multiple ramifications. The many fac-eted world of connotations and the manyfaces of use make a difference.

Now he accepts with a new kind of cre-dulity the fact that our sun is 93,000,000miles away in outer space and that othersuns are even farther away. Light yearsake on an almost incredulous dimensionas the old stereotype for mile is stretchedseemingly beyond the demands made byan ageless generic.

As a result other stereotypes may comeunder surveillance: mother, cowboy, astro-naut, spartan. Biases need to be juggledagain: rich and poor, master and slave, oldand new, near and far.

Throughout this multiplicand of changebounded by finite and infinite limits, humorcan represent the impedance ratio betweenthe polysemantic function of language andrudimentary single equivalence. It isthrough humor that he grows alert to themultiplicity of intents and implicationslatent in a single word or idea or concept,even though at times in their placing andstress they are contradictory. Such extrav-agance of meaning allows for full use ofovertones and undertones of connotationsand denotations, subtleties and acuities,synonyms and analogies.

And so, as man serves his apprenticeshipto communication, his language and histhinking may be packed with fabulouscredulities, beneath which may be, as Hooksays of the peasant, a kind of vulgar em-piricism (6). Needed is much training inthe critical examination of language; infamily circles, in classroom discussions, infriendship groups, and in private medi-tation.

Finally, as the child learns to deal with

ideas through written language, he is in-troduced to more remote persons, places,and times. Now he should be led to dis-cover anew that meaning depends uponthe total incidences of a concept in hisexperiences and upon the context in whichit is being used by the author.

The use of the simple verb to brush can-not be understood even if a dictionary isconsulted unless the reader can examinethe situation in which the word is usedand examine his own experiences. "Tobrush a picture," means one thing and"to brush up in arithmeti4" means quiteanother. Or, if a sentence read, "The set-tlers had a brush with the Indians," itmight be necessary to do as one third-grade boy did: say that he didn't under-stand what brush meant here. With teacherhelp the boy did understand how he couldbrush against someone on the sidewalk andhave a brief contact, perhaps almost askirmish; and so he discovsred that his ex-periences, when restnictured, did providesome meaning for this different use ofbrush. Certainly his meaning was not asprecise and vivid as that of his father whohad a brush with the enemy on Okinawa.Even so, father and son would need a cer-t._ in amount of credulity when readingabout the American settler's and their brushwith the Indians. Furthermore, for eitherthe father or the son the 'experience" ofa "brush with the Indians" could be com-pletely verbal. In other words, it could bethat both father and son might have ade-quate understanding without ever actuallyhaving had a brush with anyone.

Then, too, it may be that the authornever had a brush with the Indians. Thisfact together with the fact that the expe-riences of all three were differentfather,son, authormakes perfect communicationimpossible. The necessity for interpretationalways creates the need for credulity tosome degree. So an intelligent and inte-

1

1

LANGUAGE AND THE HABIT OF CREDULITY 5

gated person with a minimum of credu-lities might be described as "... one whois capable of sorting out his experiences,making pertinent distinctions among them,and seeing them in relation to his systemof values. He is capable of making com-parisons between two sets of data, twoexperiences, two generalized ideas, all with-out confusion between them. He cannot doany of these things at all unless he canattach verbal symbols to them for purposesof labelling, keep those symbols distinct,and manipulate them without confusion"(4, p. 4).

A Second ApprenticeshipThe first language apprenticeship a per-

son serves is to the spoken word. Not onlydo most people first learn to formulateconcepts and communicate ideas as re-ceivers and producers of oral language,but they do so with increasing frequencyand effectiveness as they grow and mature.Many use oral language with a considerabledegree of adequacy without knowing theirlanguages grammar, either its morphologyor syntax. The skill results from actual ex-perience with audience reaction. The dramaof face to face communication spotlightsthe need for studying audience effects, forrestating, and for clarifying abstract ideas.It is this experience of speaker-listener in-terrelationship which emphasizes the dif-ferences between oral and written commu-nication (5).

In written communication, a face to faceinteraction between the author and thereader is an uncommon experience. Thedemands on the writer are greater thanthose made on the speaker. The writermust carefully choose his words. His firstobligation is to write with precision. Cer-tainly he must keep his audience in mindbut he need not avoid using words thatmay tax his reader's vocabulary. Certainlyhe must know the demands of the languageof prose and the language of poetry.

The reader, in turn, must understandthat writing usually involves a more dis-ciplined and compact form of communica-tion. He must, in the final analysis, knowhow to find out the "whole" truth by care-fully examining and weighing the ideasrecorded so as to grasp their full meaningand to identify the hidden implications andthe motives of the author. To do all thishe must learn to read and do productivethinking.

To train a reader so that he can dealwith different interpretations and examinethe assumptions and implications of eachis to help him avert habits of credulity.Such training should be started when thechild first learns to read. Then as he in-creases in ability to recognize printedwords as conveyors of ideas he will be-come more effective in understanding whatis being said and why it is said (13).

Children just learning to read can betaught to sort out their experiences, makepertinent distinctions among them, comparethem with those of the author, generalizeconcerning the ideas presented, and do soin relation to their system of values. Asthey mature emotionally, intellectually, andspiritually they can be taught how man'scollective quest for wisdom is an ongoingprocess that reflects changing attitudes andintentions and preoccupations. They canalso learn how to see things as they are,without illusion or emotional bias, and howto make choices or decisions that are sane,prudent, fair, and reasonable.

Such training is best initiated in a groupsituation in which all the children partici-pating are required to deal with the samematerial. Under these conditions eachmember of the group can act as an auditorfor each other member and require asearching examination of events to deter-mine the facts and to test their quality,validity, truth, and accuracy ( 14).

It becomes apparent immediately that

I

1


such learning to read is a demanding task.It requires a command of thinking skills aswell as reading skills. Obviously, too, read-ing of this high caliber needs to be taught.And the training should be started, as itcould be, at the primary level, by usingmaterial that can be thoroughly intelligibleto the reader or lies within his experiencegrasp. Story materials, if well structured,do lend themselves to such use because, asAdler ( 1 ) indicates, in general the groundplan for reading story material resemblesthe division of rules for reading scientificor expository works.

Gradually, though, as the reading mate-rials provided through the intermediate andsecondary schools cause the reader to gobeyond his immediate experiences, thetraining must prepare the reader to dealwith scientific or expository material as athinking reader just as he learned to dealwith the narrative variety ( 15). Thenknowledge of grammar and logic becomesinaeasingly more valuable. This is espe-cially so when a man reads something thatat first he doesn't understand. And, as Horn(7) points out, this is usually the case inschooL

A child learning to perform as a thinkingreader must learn to do his own thinking.When he turns to a story he must learn touse the title to conjecture about the story.Certainly every reader does this to somedegree but the trained reader does so de-liberately and thoughtfully. Then as hereads the story he checks his speculationsagainst the story to see whether or not hewas right about what he thought wouldoccur. As additional facts are given he mayeither confirm his assumptions and be morecertain about the story outcome or he mayreject his assumptions and declare newones. And so on to the end.

As is evident when the reader is trainedas described here, he is given much op-portunity to deal in discoveries as he un-

ravels the problems set up by the story plot.This approach to reading training might bedescribed as a problem solving approach.By so doing, the opening remarks in thepreface to G. Polya's How to Solve It seemto be appropriate:

A great discovery solves a great problembut there is a grain of discovery in thesolution of any problem. Your problem maybe modest; but if it challenges your curios-ity and brings into play your inventive fac-ulties, and if you solve it by your ownmeans, you may experience the tension andenjoy the triumph of discovery. Such ex-periences at a susceptible age may createa taste for mental work and leave their im-print on mind and character for a lifetime(9).

If there is one quality to which everyable reader must be committed, it is theability to find and test evidence. To ac-complish this high objective, pupils mustbecome skillful in identifying the relevantand the truthful in the light of thoughtfullydeclared objectives. It has been said by thesage that a good question is half an answer.Thus the predicting or declaring of assump-tions provides a fint step in the reading-thinking process. Examining evidence, not-ing relationships and discovering storytrends provides a second step. If basicreading materials designed for use at theprimary level consist of carefully struc-tured story plots, the basic reading trainingcan be directed as described and result inreading-thinking experiences at a suscepti-ble age which may create a taste for mentalwork and leave their imprint on mind andcharacter for a lifetime.

For purposes of clinical study and re-search, thinking may be classified into fivemain types: associative thinking, conver-gent thinking, problem solving, criticalthinking, and creative thinking (11 ). To aconsiderable dergee, though, all of thesethinking types are closely interrelated andare used almost as one or may all be used

LANGUAGE AND THE HABIT OF CREDULITY

by the able reader-thinker in dealing witha reading problem. Constantly as he readsand speculates and reads again the readeris doing associative type thinking; he's put-ting together the two and two of storyevents by making associations, and by re-calling related personal experiences that fitthe scene. As he progresses through thestory the reader does convergent typethinking. From a broad divergent approachbased on limited initial clues he narrowsdown the possibilities as he approaches thestory end. Each purpose declared or eachquestion asked or each purpose changed

creates a demand for problem solving typethinking. And certainly as he proceeds thereader does critical reading. He weighs theevidence found against his experience and

knowledge of acceptable standards and ac-

cepts or rejects, and this requires a criticalattitude. Surely, too, as he proceeds hewill have opportunities to be creative inhis use of the information supplied. Orwhen a story is finished or expository ma-terial has been read he may be ready todeclare two or three other acceptable andcreative endings or next steps.

In all this the role of the teacher is ofvital importance. Again as Polya says:

1. Helping the student. One of the mostimportant tasks of the teacher is to help hisstudents. This task is not quite easy; it de-mands time, practice, devotion and soundprincipl%

The siudent should acquire as much ex-perience of independent work as possible.But if he is left alone with his problemwithout any help or with insufficient help,he may make no progress at all. If theteacher helps too much, nothing is left tothe student. The teacher should help, butnot too much and not too little, so that thestudent shall have a reasonable share of thework (9, p. 1).

The teacher must avoid being the prod-uct of authoritarian indoctrination. Shedoes this by saying frequently: What doyou think will happen next? Why do you

7

think so? Were you right in your thinking?Read the lines that prove you were right.She does net do the thinking for the chil-dren. She does not use the Pavlovian con-ditioned response approach. Rather shequickens the reading performance in anastute way so that essential concepts oftime, space, people, humor, numbers, andmorality are not overlooked.

The teacher can help pupils unobtrusive-ly and naturally when she takes advantageof group thinking and challenging, espe-cially in the problem solving atmosphereof a directed reading-thinking circum-

stance. She allows each pupil opportunityto learn to exercise self-control, to be sys-tematic, to evolve ideas of his own to whichhe is committed by considered judgment,and to tolerate and respect different ideasof others. The amenities of social adjust-ment are required and acquired in a cir-cumstance that is primarily concerned withchildren's mental development. It is in such

an atmosphere that pupils can acquire theattitudes of honest thinking so that later inlife they will always desire to be enlight-ened and informed rather than to be blindand unreasoning. The attitudes of a think-ing reader affect not only what he readsand hcars, what he accepts and rejects, butalso the objectivity of his thinking.

Such training in reading will result ineffective reading and thinking and helpbreak the habits of credulity which un-examined concepts produce. It will helpstudents clarify concepts on the "sense-data" (10) level and in turn those that can-

not be subject to immediate sensation. ksthe pupils develop skill in using thinkingtechniques, they develop an appreciationfor the value of reflective thought. Findingand using facts takes on functional signif-icance as over and over again they look forrelevant facts rather than trying to attainan idle, rote recital of "all" the facts.Gathering evidence, even story evidence,


to support an assumption, requiies anevaluation of the facts, organization of thefacts, and the drawing of sound conclu-sions. In addition, pupils learn from re-peated experience as they read throughstory after story and stop to reflect andconjecture at various points that assump-tions based on insufficient evidence are ten-tative and that judgment must bepended.

ConclusionIt has been stated that reading can be

an obstacle to thinking and help extend thehabit of language credulity initiated throughthe spoken word. Or, it can be a means ofdeveloping habits of clear thinking. Half-truths, superstitions, falsehoods, and prej-udices can be detected and dealt with. Itremains for the reader as well as the lis-tener to be alert, seek out valid inferences,follow a careful chain of reasoning fromfact to fact, and attempt to draw soundconclusions by testing his hypotheses. Whenthis is done the practiced reader will, asAltick says, find ".. . abiding evidence thathe is not so easily deceived as his neighbor"(2, p. 112).

References1. Adler, Mortimer J., What Man Has Made of

Man. New York: Frederick Unger PublishingCo., 1937.

2. Altick, Richard D., Preface to Critical Read-ing. New York: Henry Holt & Co., 1946.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

Ames, L., "The Development of the Sense ofTime in the Young Child," Journal of GeneticPsychology, 68 (1946) 97-125.Commission on Secondary School Curriculum,V. T. Thayer, Chairman, Language in Gen-eral Education. New York: D. Appleton-Century-Co., 1940.Grace, W. J. and J. C. Grace, The Art ofCommunicating Ideas. New York: The Devin-Adair Co., 1952.Hook, Sidney, "The Philosophy of Reading,"The Science and Philosophy of Reading.Proceedings of the Forty-first Annual Educa-tion Conference, University of Delaware,Newark, Delaware, 1959, pp. 20-34.Horn, Ernest, "Language and Meaning," Na-tional Society for the Study of Education.Forty-first Yearbook, Part II. The Psychologyof Learning. Chicago: University of ChicagoPress, 1942, pp. 377-413.Leacock, Stephen, Humor, Its Theory andTechnique. New York: Dodd, Mead, Co.,1935.Polya, G., How to Solve /t. Princeton, N. J.:Princeton University Press, 1948.Russell, Bertrand, The Problems of Philos-ophy. London: G. Allen, 1912.Russell, David H., Children's Thinking. Boston: Ginn and Co., 1956.Schechter, D. E., M. Symonds, and D. Bern-stein, "Development of the Concept of Timein Children," Journal of Nervous and MentalDisorders, 121 (1955) 301-10.Stauffer, Russell G., "Breaking the Basal-Reader Lock Step." Elementary School Jour-nal, 61 (1961) 269-76.Stauffer, Russell G., "The Role of Group In-struction in Reading," Elementary English,41 (1964) 230-4.Stauffer, Russell G., "Reading and the Edu-cated Guess," Changing Concepts of ReadingInstruction. IRA Conference Proceedings,1961. New York: Scholastic Magazines.Storer, John H., The Web of Life. New York:The Devin-Adair Co., 1953.

DAVID H. RUSSELL

Research on the Processes ofThinking with Some Applicationsto Reading

Research on the so-called "higher mentalprocesses" has been a dubious, even pre-carious enterprise in this country for muchof this century. Scholars in most disciplines,and even psychologists themselves, havehad doubts about attempts to study cog-nitive functioning. In the Scientific Ameri-can, Barron (4) reports sending letters towriters asking them to contribute to studiesof creative thinking. He comments as fol-lows on the replies:

In trenchant and not particularly orderlyprose. about a fifth of those who respondedto our original letter pointed out the intrin-sically evil character of psychological re-search. The objections to such research aremainly on these counts: it is vivisection; itis an expression of the effort of organizedsociety to encroach upon the individual androb him of his freedom; it is presumptuousbecause it seeks to describe and to under-stand what is intrinsically a mystery.

The suspicion of studies of thinking hasextended beyond artists, writers, collegeprofessors, and atomic scientists to psychol-ogists themselves. Despite the brilliant ex-cel,' :ons of James, Thorndike, and thetransplanted Lewin, American psycholo-gists in general have been wary of studies

The late Dr. Russell was a ' rofessor of Educationand Psychology at the University of California,Berkeley, and a past president of The NationalCouncil of Teachers of English. This paper isslightly adapted from an address on the same sub-ject given at a conference on "Reading and Think-ing" at the University of Delaware.Elementary English, XLII (April 1965), 370-378,432.

9

of mental life. We have careful laboratoryinvestigations of conditioning eye-blink andelegant procedures for recording the maze-running ability of rats, but we have oftenshied away from the study of the complexintellectual life of children and adults. Thishas probably not been true of Europeanpsychology to nearly the same degree. TheGestaltists, and Burt, Bartlett, and Piagethave been concerned with cognitive proc-esses. Here in America, as Edna Heid-breder put it, we have not always beenasking the important questions about hu-man behavioror at least not until quiterecent times. Within the last ten years,however, there has been a discernible shiftof emphasis in psychological research to-ward some of the many phases of intel-lectual functioning.

It is now about ten years since I at-tempted to put together, in some sort oforganized fashion, the scattered work ofthe last sixty years on higher mental proc-esses. In the book Children's Thinking (45)I agreed with Johnson (30) that, in sur-veying research on children's thinking, itis possible to distinguish between the ma-terials of thinking, which are multitudi-nous, and the processes of thinking, whichare very few. I suggested that it is feasibleto describe, and to some extent to discover,unique characteristics of each of six typesof thinking. These categories I am using inthis article because I believe they all canbe applied directly to the learning of Ian-

1


guage abilities and especially to learningto read, the area from which my exampleswill be drawn. Here then is the hypothesisthat most thinking behavior can be cate-gorized into one or more of the six cate-gories: perceptual thinking, associativethinking, concept formation, problem solv-ing, critical thinking, and creative thinking.It is not the purpose of this article to betaxonomicto define and distinguish thesetypesalthough this can be done. Instead,I should like to indicate a few oustandingresearches or research results in the variouscategories, with an occasional hint of howthese may be applied to the process ofreading. There are, of course, many otherlabels that could be usedthinking I, lsbeen described as relational, fluent, logical,structural, scientific, evaluative, inferential,deductive, and artistic. Spearman (51)wrote of eduction of relations and cor-relates. Guilford (22) uses terms like con-vergent and divergent thinking. Bruner(7) eistinguishes between intuitive andanalytic thinking. The possibilities aremany but this paper uses six labels whichare particularly relevant to the work ofthe teacher.1. Perceptual Thinking. Perceptual think-ing is learned; it goes beyond relativelyunlearned sensation to an awareness of ob-jects and events which are interpreted. Itmay be relatively simple as in pitch dis-crimination or complex as in a recognitionof emotional meanings. It may be objectiveas in naming a primary color or subjectiveas in interpreting pictures or in the "John-son image" during an election. Perceptualresearch flourished early in this century,moving from introspection to nicely con-trolled laboratory responses. Applicationsto letter, word, and phrase recognition areobviously related to reading and thereforeresearches and theories of perception prob-ably need more attention in educationalpsychology.

1

Some of the theories of perception arephysiologically based as in Hebb's (24)cell-assembly theory and some are func-tionally based as in Helson's (26) adapta-tion level or Brunswik's (10) perceptualconstancy. The last ten years has seen em-phasis on the influence of set, attitude, andother personality facrors in perception asin the work of Ames (2), Bruner and Post-man (9), and Blake and.Ramsey (5). Thewell-known Ames' studies illustrated theinfluence of habit in visual perception ofspace relationships. The Bruner-Postmanthree-step cycle of expectancy, input of in-formation, and checking of hypothesiswould seem to offer many leads to read-ing research. For example, they say thestronger the set or hypothesis, or category,the less information needed to confirm it,the more needed to change it. The Blakeand Ramsey book explores some of therelationships between perception and per-sonality.

More recent summaries of research onperception are represented by Wohlwill's(60) review of the development of per-ception abilities in childhood and by theGibson and Olum (18) chapter on experi-mental methods of studying perception inchildren. They find that the research on thequestion of part versus whole discrimina-tion is inconclusive with results dependingupon the materials used in the experiments.Langman (33) listed sixteen visual percep-tion skills and five auditory perceptionskills needed in reading and added seven-teen generalizations used in letter-soundanalysis. Gibson (19) studied the role ofgrapheme-phoneme correspondences in per-ception of words and concluded thatpseudo-words constructed according torules of invariant spelling-to-sound cor-relation are perceived more accurately intachistoscopic presentation than theirmatched words with variable spelling-sound prediction. Gibson also reports sev-

10.4111111.11.zune.....o.

RESEARCH ON THE PROCESSES OF THINKING

eral other studies in the volume by Levinet a/ (34), which contains accounts oftwenty-two separate studies, most of themdealing with some form of perception. Inthe collection Levin has two other studiesof variable grapheme-phoneme correspond-ences and, in addition to the study men-tioned, Gibson deals with the perceptionof letters.

The analysis of some of the more com-plex perceptions of children as they readparagraphs or stories is currently not anactive area. Earlier studies by McKillop(39), Groff (21), and others illustrated thatperception of the meaning of paragraphsmay be affected by attitudes toward thesubject matter read. Reed (44) has tracedsome of the relationships between person-ality scores and reading choices in the sixthgrade. Studies are needed, for example, ofchildren's perceptions of different types offictional characters or of different kindsof poems.2. Associative Thinking. Associative think-ing is a broad term which includes suchtheories or constructs as conditioning, S-Rbonds, primacy, and reinforcement. Withthe exception of the study of reinforce-ment, research on this topic has declinedfrom the interest of the 1920's but thereseems little doubt that the label describesmuch thinking of a rather routine sort inwhich simple relationships are established.It may be the most accurate descriptionof children's learning names of letters or asight vocabulary.

Both "(Creary (38) and Otto (43) havestudied associative learning in relation toreading ability. A number of other studieswithin this framework have been reportedin the new publication The Journal ofVerbal Learning and Verbal Behavior.

In his book, Learning Theory and Per-sonality Dynamics, Mowrer (41) extendedthe concept of association in a two-factorexposition to include both contiguity theory

11

and drive-reduction theory. Current inter-est in the area is also evident in Skinner'sreinforcement theory and its application toteaching machines. In the Soviet Union,Luria (36) is continuing the Pavlov tradi-tion with studies of children's thinkingbased on theories of conditioning and asso-ciation.3. Concept Formation. Research in thisarea goes back at least to the 1890's andG. Stanley Hall. It has always been pur-sued with some diligence and has recentlyfluorished with even more prominence ascertain scientists, mathematicians, andscholars in structural linguistics have be-come interested in the concepts childrencan learn. One problem nagging today'sprimary teacher is whether young childrenunderstand more than they did a genera-tion ago as a result of television, travel, andother phases of modern life. Another prob-lem in curriculum planning is that ofselection of the most important conceptsin a discipline. A third one concerns thecurrent teadency to introduce corceptsearlier; children can learn them soonerthan we once thought but is the earliergain worth the extra effort?

Research on concept formation has beensummarized in general articles by Russell(46) and by Carroll (12) and in specificsubject-matter fields by research workerswith interests in particular areas. The re-search on concepts can be divided intothree categories: 1) concept discovery, 2)gradual concept attainment and enrich-ment, and 3) children's knowledge of con-cepts at various age levels. Carroll believesthe first phenomenon is the result of in-ductive thinking, the second of deductivethinking. The first is usually used in lab-oratory experiments; the second and thirdare closely related to the usual teachingand learning procedures in school.

The laboratory studies of concept dis-covery began with the work of Hull (27)


in 1920 on learning nonsense name's forpseudo-Chinese characters. This type ofstudy was continued in the 1940's by Heid-breder (25) and expanded in the book, AStudy of Thinking by Bruner, Goodnow,and Austin (8). Bruner hypothesizes thatthe subjects use different "selection strat-egies" and "reception strategies" in sortingout a sequence of events or group of ex-amples so they can categorize them. Heuses such terms as simultaneous-scanning,conservative-focusing, and focus-gamblingto describe ways the category may be estab-lished. In many of his more recent booksPiaget has been concerned with conceptdiscovery in simple science experiments.Although the relationship between conceptdiscovery and the reading process is notclear, Kress (32) has shown that there aredifferences between good and poor readers,who have been matched on general intel-ligence, in the ability to discover conceptsin some of the well-known, clinical type,non-verbal sorting tests. The retarded read-ers preferred concrete to functional orabstract methQds and scored lower on ver-satility and fleicibility in concept formation.

The second main area of concept acqui-sition seems to have much significance forreading instruction; Undoubtedly manychildren beyond the ages of seven or eightlearn many concepts, at least partially, byreading about them. The series of studiesby Welch and Long (58) suggest thatchildren can use a two-step hierarchy be-tween two and four years and that mostkindergartners can grasp a three-step hier-archy (people-man-soldier). In the spiralcurriculum or through reading, children mayadd "layers of meaning" to their concepts.The most important work in the area ofconcept learning is that of Piaget who isalso concerned with the third division ofconcepts typically known at various devel-opmental levels. His numerous experimentshave been summarized and evaluated in

part by Flavell (16) and by Hunt (28).There are scores, perhaps hundreds, of

respectable investigations of conceptsknown, or not known, at various stages.For example, Russell (48) has summarizedsome doctoral studies at the University ofCalifornia on the development of socialconcepts, conservation concepts, the selfconcept, the concepts of liberty and justice(as contained in the Pledge of Allegiance),concepts of God, and concepts understoodby middle class and culturally deprivedchildren. Among other things, in The Meas-urement of Meaning, Osgood (42) sug-gests the importance of connotative mean-ings and personality factors in any analysisof a store of concepts.

Such investigations raise theoretical ques-tions of interest. For example, if certainconcepts, as in mathematics or science, arenot typically grasped at some age level,should the teacher, forsaking all others,make strenuous efforts to have the childrenunderstand these concepts if they havebeen labelled important by the mathemati-cians or scientists? Since some conceptsseem to be harder than others, but alsomore fundamental than others, in whatsequence should concepts be studied? Athird question is whether children, adoles-cents, and adults think alike or differentlyin concept formation. There is considerableagreement in the literature that thinking issimilar at all levels. In The Process ofEducation, for example, Bruner (7) writesof a central conviction "that intellectualactivity anywhere is the same whether atthe frontier of knowledge or in a third-grade classroom" (p. 14). On the otherhand, Piaget believes that the preschoolchild relies on what he calls "intuitivethought" based largely on perceptual ex-perience, that the child of elementaryschool ages shifts into a stage of "concreteoperations" or ways of getting informationwhich begin with the objective, world but


are internalized and symbolized. It is notuntil eleven or twelve years, Piaget believes,

that the child becomes capable of "formaloperations," of understanding "reversibili-

ty," or grasping possibility as effectively as

reality. It is therefore not until this stage,Piaget believes, that the child can fullygrasp the abstractions of mathematics orphysics or other disciplines. These prob-lems are examples of some of the questionsabout concept formation which must bestudied in relation to the whole curriculum,including reading.4. Problem Solving. The psychologicalview of problem solving is one of a com-plex operation involving several specifictypes of thinking. Problems may exist inany field, including those on the printed

page, where there may be a question for

the child of deciphering strange words, of

grasping an author's argument, or of judg-

ing a fictional character. Modern psychol-ogy still accepts Dewey's (13) classic five

steps in problem solving but regards themas a general, somewhat idealized picturerather than an exact description of someof the frustrations and circumlocutions ofthe individual who cannot find an imme-diate solution. For example, today we use

many labels to describe the solver's be-havior. These include 1) relational think-

ing (Maier's combining the essentials of

two isolated experiences), 2) logical reas-oning (Guilford tests for this factor), 3)rigidity (Werner and Kaplan and Bloomand Broder find this a useful concept, but

some research indicates it is a specificrather than a general trait), and 4) anxiety(Fattu reports a negative relationship be-tween anxiety and number of problemssolved). A number of studies such as thatof McNemar (40) have found that goodproblem solvers excel poor problem solversin ability to overcome an induced set and

to do deductive thinking. Harootunian (23)found that reading ability, intelligence,

13

judgment, and problem recognition were

important predictors of problem solvingability; closure, word fluency, and idea-tional fluency made little independent con-tribution to variance in problem solving

ability.As indicated elsewhere, the research sug-

gests that problem solving behavior varieswith 1) the nature of the problem, 2) themethods of attack used, 3) the character-istics of the solver, and 4) the group orsocial factors in the situation. Problem solv-

ing has been studied most thoroughly in

science situations (14, 29) and with math-

ematical materials (Wertheimer, 59), buteach of these four areas may apply in the

reading situation. For example, the first(the nature of the problem), might include

the numbers of unknown words in the se-

lection, the second (methods of teaching),the pupil's ability to outline, the third, theattitudes of the reader to the content, andthe fourth (influence of the group), maybe of interest in terms of current viewsabout individualized and group reading.What we know about the dynamics of

groups (6, 31) has not been tested in

groups organized for reading instruction.

But a number of writers including Stauffer

(52) have shown that problem solving in

the areas of word recognition and simplecomprehension may be encouraged as early

as the first grade.5. Critical Thinking. From the psycholog-

ical point of view, critical thinking is the

most dubious of the six labels by which I

am attempting to summarize research inthinking. Usually it is part of some otherprocess, as in evaluating the kinds of evi-

dence collected in problem solving or judg-

ing the original result in creative thinking.

The nearest the psychologist comes toallowing the term is in his use of theword judgment. Educational writings, onthe other hand, are full of the two words,and the term "critical thinking" is especially


the darling of the social studies people.One trouble in educational writing hasbeen that critical thinking has had so manymeanings. It has been made synonymouswith the ability to abstract and organizeinformation, to draw inferences, to searchfor relevant materials, to evaluate data, tocompare sources, to employ a from-Missouriattitude, to distinguish fact from opinion,to detect propaganda, and to apply therules of logical reasoning (49). Perhapsthe time has .arrived when we should becritical of our use of the phrase "criticalthinking."

As a research area, the field of criticalthinking accordingly suffers from this lackof precision. The exploratory study ofGlaser (20) is still about the best at thehigh school level. A number of studies ofpropaganda analysis are closely related toreading. The bulletin published by theNational Conference on Research in Eng-lish and entitled Critical Reading (50) iscorrectly subtitled as an introduction. Someof the confusion in terms is shown in thebook from England by Abercrombie (1)entitled The Anatomy of Judgment andsubtitled "An Investigation into the Proces-ses of Perception and Reasoning." Thismay be one more bit of evidence thatpsychologists confuse terms, or the wholepicture may be interpreted to mean thatcritical thinking is not a separate processso much as part of other cognitive func-tioning. Like some psychologists in thiscountry, Abercrombie reasons that in re-ceiving information from a given stimuluspattern we select from the total amount ofinformation available and from our ownstore of information. Thus the perceptualprocess involves selection and judgmentwith the subject sometimes deliberate,sometimes unaware of what he is doing.Abercrombie used a tape recorded, groupdiscussion method with university studentsand found that some of the factors influ-

encing judgment became apparent and thejudgments improved. The group discussionmethod may be one way of getting atassumptions or preconceptions and thus ofimproving critical thinking.

Recent attempts to clarify the concept ofcritical thinking have been made by Ennis(15) who divides the activity into sometwelve operlapping categories along logical,critical, and pragmatic dimensions, and bySaadeh (49) who related his analysis tosome of the rules of logic. Saadeh taughtcritical skills to sixth graders with con-siderable success as did Lundsteen (35)in another investigation of the possibilityof teaching critical listening abilities.6. Creative Thinking. In these days of em-phasis upon intellectual attainment, cur-ricular rigor and the "pursuit of excellence,"creative thinking, and creativity are fash-ionable topics. In addition to individualresearches, well-supported team studies arebeing made in a half-dozen centers through-out the country. Guilford includes creativityin his studies using factor-analysis at theUniversity of Southern California. At Chi-cago, Getzels and Jackson (17) have dif-ferentiated between adolescents scoringhigh on intelligence tests and adolescentsrated as creative, but have not studiedcases where the two groups overlap. InBerkeley, MacKinnon and his associates(37) have a series of studies of personalityfactors related to creativity in various pro-fessions, and in Minnesota, Torrance (56,57) is heading work on a group of studiesmore closely related than most to creativebehavior in classroom settings. Such studiesassume that creative thinking is not theprovince of a gifted few but exists on somesort of continuum for much of the popula-tion. In addition to certain skills, produc-tion of originality in some of these studiesseems to involve three general factorswhich may be labelled perceptual, integ-rative, and emotional. MacKinnon finds


different amounts of these in artislic crea-tivity, scientific creativity, and what heterms "overlapping" creativity. Artistic cre-ativity involves externalization of an in-ternal state; emotion and personality maybe heavily involved. In scientific cre-ativity the scientist functions as a media-tor between an external problem and itssolution and is, presumably, less involvedemotionally. Creativity in the "overlapping'category includes performers, interpreters,and high-grade individuals in such pursuitsas architecture and engineering.

In some of their research memoranda,Torrance and his students report troublein establishing the reliability of his testsof creativity in elementary achool childrenbut have certain findings about the per-sonalities of children rated as creative. Assome of us might suspect, the so-calledcreative child is not well accepted by hispeers or his teachers in the first four grades;he is often rated as limelighty and bossy.By the sixth grade a better status has usual-ly been achieved.

On the positive side, the upsurge of in-terest in creativity may be documented bythe publication, within five years, of at leastsix substantial volumes collected by variouseditors and reporting research in variousaspects of creativity. Alphabetically byeditor or author, these include Anderson'sCreativity and Its Cultivation (3), MacKin-non's The Creative Person (37), Stein andHeinze's Creativity and the Individual (53),Taylor's Creativity: Progress and Potential(54), Taylor and Barron's Scientific Crea-tivity: Its Recognition and Development(55), and Torrance's Creativity: SecondConference on Gifted Children (56). Com-bined with scores of research articles, thebooks represent increasing interest in, in-stead of final conclusions from the em-pirical study of creativity.

In the above research, two importantproblems are unsolved: the unique charac-

15

teristics of creativity in childhood andyouth, and some valid and reliable meas-ures of creativity itself. As one reviewerput it, "creativity is a constnict in searchof a generally acceptable objective refer-ent." Most of the tests of creativity havebeen developed by Guilford and his asso-ciates or adapted from his work. Theselend themselves to factor analysis, whichmay be regarded as one step on the wayto complete understanding. but some ofthem do not seem to correlate highlyeither with retest scores, with teachers' orsupervisors' judgments of creativity, orwith rating of students' creative productsby independent judges.

The books and articles are samples ofwork in progress which suggest four do-mains of research in creativity: 1) thenature of the creative process, 2) the char-acteristics of the creative person, 3) thequalities of creative products, and 4) thesocial-cultural milieu, including classrooms,which block or foster creative responses.Tha whole area of creative thinking thusbristles with problems. Is there such athing as teaching creativeness? Does crea-tivity in play, rhythms, and language occurbefore creative thinking about social orscientific problerris and are they differentthings? What can teachers do to achievesome sort of balance between conformityand spontaneity in the classroom? How canwe get more "discovery" into a readinglesson? What are the places of productionversus appreciation in reading and in othercurricular areas?

The act of reading has usually beenregarded as a receptive process rather thana creative one. There seems to be somejustification, ht. wever, :For the use of theterm "creative reading" to sigr v behaviorwhich goes beyond word ide .Ifk.vtion orunderstanding of literal meaning to thereader's interpretation of the printed ma-terials (47). Such reading may be produc-

I


five of new ideas, critical of old ones, orappreciative of the art of literature. Re-search studies suggest that certain mechan-ics of reading must be well in hand beforethe child or adolescent achieves thesehigher levels of reading and that they, likethe skills, can be developed by the rightkinds of instruction.

Conclusion

The above examples suggest that lan-guage abilities should be defined not merelyas perceptual skills nor as the ability tograsp a communication, nor as competencein solving verbal problems. Probably alltypes of thinking are involved in the learn-ing and use of language and these I havesubsumed under six labels. Other generaldescriptive terms could be used and moreprecise designations of specific verbal be-havior are undoubtedly needed. Further-more, this account neglects such topics asemotional factors in thinking, the role ofmemory, and attempts to derive a com-prehensive theory or model of thinking asin the work of Burt (11) or Guilford (22).It is probably not too important in thestudy of the reading process, from whichthe examples are drawn, to distinguish be-tween the types of thinking, here calledperceptualizing and conceptualizing. Be-havior which involves the apprehension ofevents or objects such as printed symbolsmay be profitably conceived as a categoriz-ing, whether perceptual or conceptual. AsBruner has put it, 'There are examples inwhich it is almost impossible to differen-tiate perceptual and conceptual categoriz-ing, notably in language learning" (8). Inaddition to its use in the discovery of con-cepts, reading seems to be one of the bestways we have of deepening and enrichingconcepts.

Similarly, there is overlap of criticalthinking with the processes of problemsolving and creative thinking. A child or a

scientist must be critical of his proposedsolutions to a problem. An adolescent oran adult must sometimes be critical abouthis creative production, whether an originalstory or an interpretation of A. E. Hous-man. Despite this blending of critical think-ing into problem solving and creative think-ing, it is my bias that some aspects ofcritical thinking can be taught directly assuch. (See the article by Stauffer in thisseries.) Similarly, I believe the other fivetypes of thinking can, to some extent, beisolated and taught in relation to the schoolcurriculum, including reading.

In red ling instruction of the past, mostof a reading teacher's time and energyhave gone into perceptual aspects of wordidentification and conceptual responses toliteral meaning. These are necessary basesfor more sophisticated approaches to read-ing, but perhaps the time has come whenwe can use our psychological knowledgeof the processes of problem solving andcritical and creative thinking to help teach-ers develop a more demanding set of goalsfor reading instruction.

Bililiography1. Abercrombie, M. L Johnson, The Anatomy of

Judgment. London, England: Hutchinson,1900.

2. Ames, Adelbert, "Visual Perception and lieRotating Trapezoidal Window," PsychologicalMonographs 65: No. 324, 1951.

3. Anderson, Harold H., (ed.), Creativity and ftsCultivation. New York: Harper, 1959.

4. Barron, Frank, "The Psychology of Imagine-tion," Scientific American 199:151-166, Sep-tember, 1958.

5. Blake, Robert R. and G. V. Ramsey, (eds.),Perception: An Approach to Personality. NewYork: Ronald, 1951.

6. Blake, Roy, "Small Group Research and Co-operative Teaching Problems," National Ele-mentary Principe/ 43:31-36, February, 1964.

7. Bruner, Jerome S. and Leo Postman, "Sym-bolic Value as an Organizing Factor in Per-ception," Journal of Social Psychology 27:203-208, May, 1948.

8. Bruner, Jerome S., Jacqueline Coodnow, andG. A. Austin, A Study of Thinking. NewYork: Wiley, 1956.

ATV

RESEARCH ON THE PROCESSES OF THINKING 17

9. Bruner, Jerome S., The Process of Education. 27.Cambridge: Harvard University Press, 1900.

10. Brunswik, Egon, Perception and the Repre-sentative Design of Psychological Experi- 28.ments. Berkeley: University of CaliforniaPress, 1956. 29.

11. Burt, Cyril, "The Differentiation of Intellec-tual Ability," British Journal of EducationalPsychology, 24:76-90, June, 1954. 30.

12. Carroll, John B., "Words, Meanings and Con-cepts," Harvard Educational Review, 34:178-202, Sprint 1964. 31.

13. Dewey, John, How We Think. Boston: Heath,1910.

14. Duncker, Karl, "On Problem Solving," trans.by Lynne S. Leer, Psychological Monographs,58, No. 27, 1945. 32.

15. Ennis, Robert H., "A Concept of CriticalThinking," Harvard Educational Review, 32:81-111, Winter, 1962.

16. Flavell, John H., The Developmental Psychol- 33.ogy of Jean Piaget. Princeton, N. J.: Van Nos-trand, 1963.

17. Getzels, Jacob W. and Philip W. Jackson,Creativity and Intelligence. New York: Wiley. 34.1962.

18. Gibson, Eleanor N. and V. Olum, "Experi-mental Methods of Study in Perception inChildren," P. H. Mussen (ed.), Handbook ofResearch Methods in Child Psychology. New 35.York: Wiley, 1900.

19. Gibson, Eleanor, et 4 '"Ilie Role of Graph-enie-Phoneme Correspondences in the Percep-tion of Words," American Journal of Psychol-ogy, 75:554-570, December, 1962.

20. Glaser, Edward M., An Experiment in theDevelopment of Critical Thinking. Contribu- 37.tions to Education No. 843. New York:Teachers College, Columbia University, 1941.

21. Groff, Patrick J., Children's Attitudes TowardReading and Their Critical Reading Abilities 38.in Four Content-Type Materials, doctor's dis-settation. University of California, Berkeley,1955.

22. Guilford, Joy P., "Three Faces of Intellect," 39.American Psychologist 14:469-479, August,1959.

23. Harootunian, B. and M. B. Tate, "The Rela-tionship ot Certain Selected Variables toProblem Solving Ability," Journai of Edu- 40.cational Psychology, 51:326-33, December,1900.

24. Hebb, Donald 0., The Organization of Be-havior: A Neuropsychological Theory. New 41.York: Wiley, 1949.

25. Heidbreder, Edna, "The Attainment of Con-cepts: IV, The Process," Journal of Psychol- 42.ogy, 24:93-138, July, 1947.

26. Helson, Harry (ed. ) and others, TheoreticalFoundations of Psychology. New York: VanNostrand, 1951.

36.

Hull, Clark A., "Quantitative Aspects of theEvolution of Concepts," Psychological Mono-graphs, 28, No. 123, 1-86, 1920.Hunt, Joseph McV., Intelligence and Experi-ence. New York: Ronald, 1961.Inhelder, Barbel and Jean Piaget, The Growthof Logical Thinking from Childhood to Ado-lescence. New York: Basic Books, 1958.Johnson, Donald M., "A Modern Account ofProblem Solving," Psychological Bulletin, 41:201-229, April, 1944.Kelley, H. H. and J. W. Thebaut, "Experi-mental Studies of Group Problem Solving andProcess," Handbook of Social Psychology,Vol. II (Gardner Lindzey, ed.). Cambridge,Mass.: Addison-Wesley, 1954.Kress, Roy, "An Investigation of the Relation-ship Between Concept Formation and Achieve-ment in Reading." Abstract of Dissertation;private communication from author, 1960.Langman, Muriel P., "The Reading Process:A Descriptive, Interdisciplinary Approach,"Genetic Psychology Monographs, 62:3-40,August, 1900.Levin, Harry, Eleanor Gibson, and others, ABasic Research Program in Reading. Coop-erative Research Project No. 639. Ithaca,New York: Cornell University, 1963. (Mim-eo& )Lundsteen, Sara W., Teaching Abilities inCritical Listening in the Fifth and SixthGrades, doctoral dissertation, University ofCalifornia, Berkeley, 1963.Luria, Aleksandr R., Speech and the Devel-opment of Mental Processes in the Child.London, England: Staples Press, 1950.MacKinnon, Donald W. (ed.), The CreativePerson. Proceedings of Conference of IPARand University Extension. Berkeley: Univer-sity of California Extension Division, 1961.McCreary, Anne P., "A Study of Association,Reinforcement and Transfer in BeginningReading," Journal of Experimental Education,31:285-290, Spring, 1963.McKillop, Ann S., The Relationship Betweenthe Reader's Attitude and Certain Types ofReading Responses. New York: Bureau ofPublications, Teachers College, 0-lumbia Uni-versity, 1952.McNemar, Olga W., "An Attempt to Differ-entiate Between Individuals with High andLow Reasoning Ability," American Journal ofPsychology; 68:20-36, March, 1955.Mowrer, 0. Hobart, Learning Theory andPersonality Dynamics. New York: Ronald,1950.Osgood, Charles E., G. J. Suci, and P. H.Tannenbaum, The Measurement of Mean-ing. Urbana: University of Illinois Press,1958.


43. Otto, Wayne, "lir Acquisition and Retention 52.of Paired Associates of Good, Average andPoor Readers," Journal of Educational Psy-chology, 52:241-M October, 1961. 53.

44. Reed, Charles H., Relationships of Personalityand Reading Choices of Sixth-Grade Children,doctor's dissertation, University of California, 54.Berkeley, 1962.

45. Russell, David H., Children's Thinking. Bos- 55.ton: Ginn, 1956.

46. Russell, David H., "Concepts," Encyclopediaof Educational Research, 3rd edition, (C. W. 56.Harris, ed.). New York: Macmillan, 1960.

47. Russell, David H., Children Learn to Read(2nd ed.). Boston: Ginn, 1961.

48. Russell, David H., "Six Studies of Chffiren's 57.Understanding of Concepts," ElementarySchool Journal, 63:255-200, February, 1963.

49. Saadeh, Ibrahim Q., An Evaluation of theEffectiveness of Teaching for Critical Think- 58.ing in the Sixth Grade, doctor's dissertation,University of California, Berkeley, 1962.

50. Sochor, E. Elm (ed.), Critical Thinking:An Introduction. Bulletin of the National Con- 59.ference on Research in English. Champaign,Illinois: National Council of Teachers of 00.English, 1959.

51. Spearman, Chades E., The Abilities of Man.New York: Macmillan, 1927.

Stauffer, Russell G., "Children Can Read andThink Critically," Education 80:522-525,May, 1960.Stein, M. I. and S. J. Heinze, (eds.), Crea-tivity and the Individual. Glencoe, In.: TheFree Press, 1900.Taylor, Calvin W., Creativity: Progress andPotential. New York: McGraw-Hill, 1964.Taylor, Calvin W. and Frank Barron, (eds.),Scientific Creativity: Its Recognition and De-velopment. New York: Wiley, 1963.Torrance, E. Paul, (ed.), Creativity: Pro-ceedings of the Second Minnesota Conferenceon Gifted Children. Minneapolis: Universityof Minnesota Extension Division, 1959.Torrance, E. Paul, (ed.), Talent and Educa-tion: Present Status and Future Directions.Minneapolis: University of Minnesota Press,1900.Welch, Livingston and L. Long, "The HigherStructural Phases of Concept Formation inChildren," Journal of Psychology, 9:59-95,January, 1940.Wertheimer, Max, Productive Thinking. NewYork: Harper, 1945.Wohlwill, Joachim, "Developmental Studiesof Perception," Psychological Bulktin, 57:249-288, July. 1980.

JAMES IL SQUIRE

Form Consciousness, an ImportantVariable in Teaching Language,Literature, and Composition

Teachers of English have long consideredform and content to be complementary at-tributes of language, literature, and com-position. Much classroom time is devotedto studying interrelationships of the two,so much so that some teachers would evenargue that form and content are insepar-able and that the structure or form throughwhich ideas are communicated actuallydefines in large measure the nature andmeaning of the ideas themselves.

During recent years, an increasing num-ber of scholars have interested themselvesin the formal qualifies of language as dis-tinct from the meaning of language. Thusthe structural linguists separate the lin-guistic meaning of a single utterance fromthe semanfic meaning; the literary criticstrives to describe the aesthetic form of aliterary work; the educational psychologiststrives to identify individual variation inthe ability to perceive form in language.The concern with form, shape, and recur-rent pattern in communication is truly in-terdisciplinary. Psychologists, rhetoricians,and literary critics each have much to con-tribute. Although present knowledge aboutthe ways in which individuals perceive formin language is fragmentary in many re-spects, studies bearing on the topic may

Dr. Squire is Executive Secretary of The NationalCouncil of Teachers of English and Professor ofEnglish at the University of Illinois.Elementary English, XLII (April 1965), 379-390.

19

be found in several fields. This article con-siders selected findings with respect to theprocesses of perception and to what isknown about consciousness of form insimple spoken dialogue, in written dis-course, in literature, and in related artmedia.

The Perceptual ProcessAllport has reviewed the way in which

the perception of structure or form isviewed in thirteen separate theories of per-ception (2). He finds form perceptioncentral to every theory and advances num-erous generalizations concerning form per-ception basic to all positions. Hilgardfurther suggests that the emphasis uponorganization and perceptual patterning socharacteristic of Gestalt psychology ( andof importance in any discussion of indivi-dual perception of form ) has been em-braced by associational psychologists in theirwork with problem solving and creativethinking (35). Thus concern with the per-ception of form appears central to manyof today's perceptual theories. Some of therecent research bearing on perception hasbeen summarized from different points ofview by Rosenstein (68 ), Carroll (10),Fearing (21 ), Gibson (27), Suchman andAschner (85), and Wohlwill (89).

Reporting on a series of experimentstesting the effect of organization on per-ceptual recall of visual form, Glanzer andClark advance a hypothesis which may

1


have significant implications for the per-ception of form in literature and composi-tion (29). They suggest that in perception,restricted primarily to visual perception intheir work, subjects translate perceived in-formation into words before formulatingfinal responses. Any interference in thetranslation of the pattern perceived priorto verbalization interferes with the accu-racy of the final response. The hypothesissuggests the possibility that individuals donot finally comprehend total pattern, evenvisual patterns, until they identify and de-velop their percepts in words. It further sug-gests the complexity of assisting individualsto perceive form in literature in view of theinterference likely to occur during theprocess of reading Squire's discovery thatadolescent subjects reading short storiestend to delay judgment of overall aestheticform until their reading is completed offerssome evidence to support Glanzer andClark (81). The hypothesis offers re-searchers the further suggestion that thelength of verbalization elicited by anystimulus offers a measurable, objective in-dex of form perception.

Several studies suggest the ways inwhich concept development occurs in child-ren. Concepts of form, like generalizationsassociated with meaning, emerge slowlyfrom both deductive and inductive pro-cesses out of the percepts, memories, andimages resulting from each individual'sreorganization of experience (11 ). Indeedsimple percepts of form appear to be amongthe earliest percepts to develop (70). Dur-ing childhood, the ability to perceive form,like other abilities of perception and ab-straction, probably varies more with theexperiences of children than with theirintelligence (85). As Rosenstein says,

A child may see, feel, or experience anobject appropriately, more in accordancewith some habit rather than in accordancewith the actuality of the situation. The pro-

cess of inductive thinking, which involves asearch for common attributes in a givengroup, may easily go astray because theprocess of generalization is neither under-stood nor consciously controlled by thechild. Even more, a wrong generalizationmay easily be accepted by the child be-cause he has no related experiences bywhich to check his conclusions (69).

Planned experiences with perceiving formin language may thus influence the child'sawareness of form, but additional researchis needed to suggest whether such plannedexperiences in language in sentences,paragraphs, and literary selections, for ex-ample ultimately assist the individualin developing greater consciousness ofform. Dienes reports that extensive andvaried experiences with related abstractionsin mathematics contribute to strengthenoverall generalizations ( 18), but evidencedoes not yet indicate clearly that such find-ings apply equally to the language field.Certainly supporting such an hypothesis isFearing's belief that the structuring pro-cess involves isolating, organizing, under-standing, and ultimately retaining overallcognitive structures or "schema" which canbe drawn upon later to organize additionalpercepts (21). Gibson says, however, thatthis theory of the overall concept or imageas prototype for recognition conflicts withan equally widely held view that percep-tion results from the child's increasinglydifferentiated response to distinctive stim-uli, such as even to aulividual phonemesor graphemes or to minute combinationsof the two (27)

Russell suggests that conscious effortsto encourage the perception of form inlanguage may begin too early. In sum-marizing what is known about the analysisand synthesis of aesthetic form in music,painting, literature he reports that child-ren's aesthetic concepts tend not to reachtoward the abstract principle of unity inart (70). However, Carroll writes that

FORM CONSCIOUSNESS, AN IMPORTANT VMUABLE DI TEACHING 21

after age six, children appear to learnlittle in language but vocabulary andschool-imposed standards of language us-age (10).

The context in which individuals are re-quired to perceive form also appears tobe important. The total framework of thelearning situation appears to affect percep-tion (77). So does the time required forperception. Objects, events, or ideas thatneed to be associated and fitted into anystructure must occur close together and bepresented in such a way as to accentuatetheir similarities (21). Thus learners shouldhave less difficulty in perceiving the formof a well-written paragraph than the over-all pattern of an essay involving severalparagraphs. Of even greater significancein the classroom situation is the suggestionthat any irterruption or lengthening of thetime sequence required for perceptionthreatens to interfere with the learner'sability to perceive overall pattern. A childpresumably can better perceive form in apoem or short story if these are read with-out interruption. Time becomes increas-ingly important, states Hartman, whenoverall pattern is achieved and mademeaningful by metaphorical or illustrativestimuli (34), such as apparently occurswhen the overall structure of a poem likeShelley's "Ode to the West Wind" is per-ceived through its metaphorical unity.

Clearcut studies of form consciousness inlanguage are not easily identified. Russelland Fea recently observed that despitethe multi-faceted nature of abstraction andgeneralization involved in concept forma-tion, much research has been single-faceted (72). Also, many experiments inconcept formation have been concernedless with the teaching of concepts of anykind than with the determination of theconcepts children have already acquired.Yet existing studies suggest that conceptsof form, like other concepts, can be ac-

quired in two possible ways: they can betaught consciously by the teacher pre-senting appropriate stimuli and assistingchildren in their generalizing; and theycan be discovered by the children them-selves through "hit and miss" experimenta-tion (68).

Form in Linguistic Utterance

For several years researchers in linguis-tics and in psychology have concernedthemselves with the study of the singleutterance or the single statement. Brownemphasizes how language provides cate-gories for ordering and classifying experi-ence (7, 8). Miller and Isard demonstratethat both syntactic and semantic rules areinvolved in the perception of language(56), but it is with the syntactic, formalelements that this paper is concerned. Psy-chological studies indicate clearly thatlanguage is more easily recalled when itoccurs in formal unitsstatements, phrases,clauses, or single utterances, i.e., in thecontext of some identifiable form (28, 55,57). Linguists like Sapir have demonstratedthat within the English statement, certainwords, traditionally called verbs and nounsand their modifiers, carry essential mean-ing; that certain other words, sometimescalled articles or intensifiers, offer structuralclues to the meaning (73). The study ofthe structure of the English statement, un-encumbered by attention to the sense, cansometimes help pupils to develop under-standings about the ways in which ideasare presented. Glanzer points out that the"unit of communication" in English is notthe single word but combination of multi-word signals which include both meaningwords, such as nouns and verbs, and signalwords (28). Most linguists agree that in-dividual words function only in combina-tions and that language is perceived andstudied in clusters. Loban used bothphonological units, identified by inflection,

1


stress, and pause in his subjects' voices;and communication units, identified by se-mantic elements, in his study of languagedevelopment (44, 45).

In a discussion of recent educationalresearch involving the insights of genera-tive grammar, Postal suggests that thestructure of the English sentence involvesa highly complex, abstract series of ele-ments which can be related to actual ut-terances only by a set of highly structuredrules. He finds that views of languagelearning which restrict attention on theone hand to gross phonetic properties oron the other to such gross psychologicaltheories as generalization or stimulus-responses fail to account for "the fantasticfeat by which a child with almost no directinstruction learns that enormously exten-sive and complicated system which is natu-ral language" (65). Postal assumes thatgenerative grammarians will ultimately finda basic linguistic system underlying alllanguage performance.

Structural studies of language review theimportance of the formal structure of theEnglish sentence. Hook and Matthews, forexample, report that 96 percent of all Eng-lish sentences, spoken or written, may bereduced to four basic sentence patterns(38). Fries revealed that the ability tomanipulate these basic patterns and achievevariation in sentence structure distinguishesverbally competent adults (25). In hisstudy of language development of childrenduring elementary years, Loban substan-tiated his finding with respect to the abilityof children to manipulate basic sentencepatterns. "Although differences in struc-tural patterns are not notable with theexception of partials and linking verbsvery important differences do show up inthe dexterity with which subjects use ele-ments within these structures... . Not pat-tern but what is done to achieve flexibilitywithin the pattern proves to be a measure

of the effectiveness and control of languageat this level of development" (44).

Awareness of the form of English sen-tences is thus increasingly seen as a foun-dation for effective communication, butawareness of sentence form, not an aware-ness of grammatical generalizations. Strick-land, who presented evidence to indicatethat preschool children use in their speechall of the complex sentence forms whichthey will ultimately use in their writing,notes that, "Children's concepts of many, ifnot most of the things and ideas they encoun-ter, develop without benefit of definition . . ."(83). Definitions of linguistic terms maybe more confusing than helpful and domore harm than good. Rather than relyon definitional approaches in developingconsciousness of sentence form, Stricklandadvises teachers to build on children's in-terest in manipulating language, beginningwith a sentence the child has used, strip-ping it down to its basic subject and predi-cate ( the "irreducible minimum which canbe called a kernel, nucleus, or core"), andencouraging children to add elements ofexpansion the movable puts of the sen-tence which can influence emphasis de-pending upon their placement.

Studies of the language development ofchildren summarized by McCarthy indi-cate that mean sentence length has beenthe most reliable and objective measureof linguistic maturity (51). Such studieshave concentrated on the formal qualitiesof written language rather than on speech,which according to recent findings reflectschildren use far more complicated linguis-tic patterns than does their writing (45,83). LaBrant discovered that the clauseswritten by grade school children becomelonger as writers mature (43). Hillocksrecently reported ability to subordinateand compound were characteristic of ma-turity (36). Utilizing transformational ap-proaches, Hunt analyzed the prose of

1FORM CONSCIOUSNESS, AN IMPORTANT VARIABLE IN TFACHING

fourth, eighth, and twelfth graders and re-ported that major growth occurs in thenominal structures rather than in adverbialmodifiers. He found older students ableto incorporate more varied grammaticalstructures within a single grammaticallyindependent unit; i.e., able to incorporatea larger body of thought into a single in-tricately related sentence organization. Be-cause Hunt discovered that verb comple-ments, unlike nominals, apparently do notincrease with age in sentence-combiningtransformations, he hypothesized that hisstudy revealed the "growth buds" for sen-tence transformation. His work demon-strates the potential value for research inlanguage learning of modern transforma-tional theories of English grammar.

Recent linguistic scholarship also revealsattempts to identify basic spelling patternswhich may be taught to children. Friesreports modern English spelling to be pri-marily a system of a comparatively fewarbitrary contrastive sets of spelling pat-terns which efficient readers need to learnto master language signals (26). Hall alsosuggests that phonemic patterns may beidentified, and Strickland sees a programof spelling instruction based on identifica-tion of such patterns to offer a fruitful ap-proach to teaching spelling (33, 83). Quitepossibly these new attempts to improvechildren's recognition of spelling patternsbased on modern phonological analysismay prove more successful than earlierattempts to teach classifications of spellingwords.

This brief review of selected studies inlinguistic analysis and psychology suggestsonly the possibilities for research. Ls in-creasing numbers of psychologists informthemselves about grammatical anPlysis andas more grammarians develop interest inempirical studies, the potential value ofsuch scholarship is manifest. Structuralanalysis of the English sentence already

23

has yielded important insights with impli-cations for teaching. Perceptual studiesutilizing the insights of transformationaltheory may prove even more basic.

Form in Composition

The ability to perceive form is basic togood composition. Indeed the traditionalprinciples of effective writing unity, co-herence, and emphasis all point to thesignificance of imposing an overall stiuc-ture or form on the ideas communicatedthrough the paragraphs or longer units ofdiscourse.

Research indicates that the ability toorganize is an important variable in dis-tinguishing between good and poor writ-ers. The good writer is one who worriesabout organization, about not being speci-fic, about not having a clearcut purpose,about not being direct and to the point;whereas the poor writer concerns himselfwith such organization as the slant andalignment of his handwriting, and withmechanical matters not directly related tothe problem of organizing the expressionof ideas (4, 16, 80). At one major univer-sity some years ago a faculty committeestudied the writing of college undergradu-ates and concluded that the ability to or-ganize was basic in all good writing, thatmastery of the techniques of grammaticalusage seemed to be a corollary of abilityto write well (41).

Russell and Fea report that purpose willgovern whether a child sees the overallidea or form of a paragraph (72). Studiesare needed to ascertain more clearlywhether training in such awareness helpsstudents as they compose.

Research suggests that those who areable to observe form in their own writingare best able to perceive it in the writingof others (4). One effective way to im-prove a writer's control of paragraph formis to provide a study of good models (74),

i


although the effect of such instruction hasnot yet been carefully studied by research.Parker does suggest that awareness of topicsentences and concluding summaries, i.e.,awareness of basic paragraph form, in-creases the comprehension of a student'sreading of expository prose (62). Milesanalyzed the writing of thirty college fresh-men capable of reading Thoreau, Jung,Freud, Plato, and Reisman, and concludedthat many bright students fail to relatetheir basic ideas to any pattern of organi-zation (53); that is, they fail to predicatetheir ideas by expressing them in state-ments which can be expanded. For manystudents, ideas are equated with abstractwords or phrases rather than with predi-cated statements. Such students fail toassociate the concept of an idea with theconcept of the statement, thus are unableto see the essential connection between thepredication made, and the arrangement ofthe supporting illustration which follows.The importance of form and predication inparagraph development has been carefullydiscussed by Evans who argues that with-out the imposition of formal pattern ofexpression, students learn to write but notto compose (19).

Form is one of several variables whichappears to influence writing ability. In astudy of factors influencing success inwriting, Diedeush, French, and Carltonstudied evaluations of selected specialistson writing teachers, professors, editors,and writers (17). They discovered fiveoverall factors which influence judgment:content, form, diction, flavor (or style),and mechanics. Moreover, they reportedthat readers vary in the emphasis whichthey place on various factors. Many teach-ers, regrettably, were found to confinetheir corrections and annotations largelyto mechanical matters, rather than to con-sider other factors as well. Inasmuch as areasonably balanced attention to all factors

is as clearly desirable in annotating papersas in organizing instructions, the studyoffers important leads for researchers andcurriculum specialists concerned with theproblem of balance. In a statewide surveyof practice in teaching composition, Meckel,Squire, and Leonard reported the teachersmost successful in teaching composition tobe those who emphasize the clarificationof ideas. This clearly suggests the import-ance of form and structure in effectivewriting and the need for providing formalanalysis as part of the instructional pro-gram. In an interdisciplinary seminar rep-resenting scholars in several fields, Sebeokand others studied the nature of style inlanguage and identified a number offormal qualities (75). An interesting newapproach is that of Christensen who foundfour qualities to determine the effective-ness of expression in writing: the principleof addition, the direction of modification,the level of generality or abstraction, andthe texture or density of any sentence interms of its stylistic qualities (13, 14).Christensen's work has been called the be-ginning of a generative rhetoric of Englishprose, because it stems from the concernwith generating ideas first expressed bytransformational grammarians. If extendedby additional study, his work may providenot only overall guidelines for effectiveinstruction, but methods of identifying andisolating selected characteristics of Englishprose style for separate study. For example,what differences can we find in the textureor density of children's writing at differentage levels? By applying Christensen's anal-ysis to selected samples of writing, re-searchers may provide helpful answers.

Concerned primarily with the findings ofempirical research, this paper necessarilyavoids any extensive discussion of severalother recent attempts to provide rhetoricalanalyses of English prose (5, 6, 32, 47, 66).However, any researcher concerned with

voftwr... ...~0...1.0kowers.S.mrvelWerAntrarowyroerimprn.A.werw.v.,....nr.r.rone-mys,,,,,,....--

FORM CONSCIOUSNESS, AN IMPORTANT VARIAME IN TEACHING 25

stylistic matters needs to be aware of suchwork.

Form in Literature

To a considerable extent, the form ofany literary work determines its aestheticeffectiveness. On other factors perhaps,such as the universality of the idea, maydepend ultimately our assessment of thevalue of the work, but it is form morethan any other quality which determineswhether a work is literature at all. Andit is awareness of the importance of form,perhaps more than any other factor, thathas led curriculum specialists to recom-mend against permitting the study of frag-ments, snippets, or adapted versions ofliterature in the classroom (15). Consci-ousness of form in reading any literarywork is important.

Russell's review of research on the im-pact of literature shows how fragmentaryis all knowledge in this area, includingknowledge about perception of form inliterature (71). Smith has also summarizedsome studies of appreciation (79).

Studies of responses to literature haveyielded some insights. Richards analyzedmisinterpretations of college readers ofthirteen poems of unknown authorship, anddiscovered not only stereotyped responsesand difficulties in comprehension, but alsopossible effect of general critical precon-ceptions and technical prejudgments, someof which reflected excessive concern witharbitrary notions of form (67).

Meckel found unfavorable student com-ments to Hugh Walpole's "Fortitude" re-sulted more from ideas in the literaryselection than from reactions to form orstyle (49). His findings tend to be sub-stantiated in part by the Wilson reviewof college readers to three novels (88).

Squire analyzed responses of fifty ado-lescent readers to four short subjects andreported covariation of literary judgments

(which deal with formal qualities) andemotional self-involvement responses.Fewer literary judgments occur while ado-lescents read the central portion of a storythan occur before involvement or at theend of reading (81).

Burt discovered an age differential inthe awareness of children of such formaiqualities of poetry as poetic expression,rhythm, melody, and vowel music (9).He reports that younger children respondto poems largely in terms of story factorsand that appreciation of other, more formalelements does not emerge until later years.

A series of studies indicated that al-though general factors may determineoverall literary preference, a secondaryfactor related to consciousness of form inliterature also influences responses.

Williams, Winter, and Woods appliedfive tests of literary appreciation to morethan 200 children and adolescents andfound that a general factor of literary ap-preciation, correlated with intelligence, ac-counted for 50 percent of variation inresponses, whereas a second bipolar factor,accounting for 20 percent of the variance,separated readers preferring the objective,form-conscious styles of classicists fromthose who preferred subjective approachesof romantic school of writers (87).

Gunn subsequently identified a generalaesthetic factor associated with such quali-ties as liking, emotional effect, mode of ex-pression, appeal of the subject, and a bi-polar factor distinguishing readers con-cerned with rhyme, word music, andrhythm, as distinct from those concernedwith emotional effect, appeal of subject,and mental imagery (31).

Forman found free responses morehelpful than responses elicited by specificquestions in attempting to measure appre-ciation. He utilized three scales elabora-tion of details to measure visualization,

*continuity and purpose to indicate com-


prehension of meaning and sequence, andcharacter vitalization to measure responseto literary figures (23). The correlationbetween scales was so high that he con-cluded that they measured a general ap-preciative factor.

Carroll attempted to quantify aspects ofliterary style by analyzing 150 passages ofEnglish prose in accordance with 29 ad-jectival scales covering major qualities andtraits. His discovery of a factor calledGeneral Stylistic Evaluation indicatingoverall positive or negative evaluation ap-pears to substantiate the earlier studies.Six independent dimensions of style werestudied by Carroll: General Stylistic Eval-uation, Personal Affect, Ornamentation,Abstractness, Seriousness, and Characteri-zation versus Narration (12).

Recent studies of preferences for litera-ture have been described by Peel whotested responses to thirty-one abstract pat-terns in terms of such criteria as rhythm,symmetry, lightness of tone, and lack ofcontrast (63). Peel's method for determin-ing preference by relating specific deter-miners of preference to the liking choicesof those investigated suggests a way ofstudying form consciousness in judgingworks or arts.

Peel, also, has described a use of Os-good's semantic differential to analyzepreferences in terms of qualities in a workof art by applying a set of twenty scalesincluding measures of vividness, depth, andclarity to selections from twelve majornovelists (60, 61).

Peel notes that in applying methods usedto analyze preferences for paintings toanalyzing preferences for literary works,researchers must note that paintings evokemore or less instantaneous attitudes,whereas responses to prose and poetryoften require longer reflection. Thus Peel'smethods probably are most appropriate instudying reactions to short passages.

Not all studies of literary form of inter-est to teachers stem from empirical re-search. Contemporary literary criticismespecially has yielded many valuable stud-ies. For example, Goodman's analysis ofthe structure of literature focuses on formalqualities (30), as does Ohmann's attempt toidentify recurring patterns of expression inthe language of Bernard Shaw (59). Somecritical writing combines concern for theformal qualities of literature; Fagan, forexample, reviews field theory in the sci-ences and suggests extensions to the studyof literature. His detailed analysis of afield approach to the novel Storm suggestsa teaching dimension to literary studieswhich emphasizes symbolic unity (20).

After reviewing much modern criticismas well as the recommendations of special-ists on the teaching of English, Walkeridentifies nine points of agreement con-cerning teaching the structure of literature:

1. The study of literature should includea careful analysis of the woe- to seewhat structural relationships exist.

2. The investigation of structure shouldbe inductive (not with formula orlist of structural elements in hand).

3. The structure of literature is whatgives it unity or enables it to have aunifying effect.

4. Structure is the overall pattern of re-lationships that holds parts together.

5. The study of structure involves thestudy of form and content.

6. The study of structure should beginwith a view of the whole.

7. The study of structure facilitates un-derstanding and interpretation of lit-erature.

8. The understanding of structure broad-ens the base of literary appreciationby enabling readers to appreciateliterature not only for its appeal ormeaning to them but also for thecraftsmanship involved in its creation.

FORM CONSCIOUSNESS, AN ImPORTANT VARIABIE DT TEACHING 27

9. The study of structure facilitatestransfer of ephemeral principles per-tinent to literature, composition, anda process of investigation (86).

Walker found that specialists disagreeon whether the purpose for studying struc-ture in literature is to judge or to under-stand a work. He believes those influencedby the New Criticism of The Thirties andForties see structure as an effective meansof evaluation; others feel that understand-ing is the greatest value to be gained fromstudy of a work's structure.

Form in Other Media

Perception of form is basic to the appre-ciation of all artistic expression. Whatevercan be done to encourage perception ofform in one medium of expression may re-inforce and extend such awareness inothers (39). Burt's research indicates, also,that an individual who is a good judge inone field is likely to be a good judge inanother (9), although Nordberg warns thatany visual stimuli have unique characterof their own (58). The bulk of researchsuggests that young people who learn todistinguish beauty of form in one mediummay transfer some of their awareness toexpression in a related medium.

Studies of reactions to pictorial formhave probably been more extensive thanstudies of such perception in literature andcomposition. The findings reported in suchresearch suggest some interesting possibil-ities concerning form consciousness of other

kinds.Peel (1945), for example, found that

addiszcents prefer well-drawn designs, re-flecting naturalistic principles, symmetricaldesigns, balanced patterns, and brilliantcolors (64).

Siegel and Siegel (1950) found childrenabout ten years of age able to conceptualizebeauty in tenns of material possessions and

moral values, rather than aesthetic ideas(78).

French (1951) demonstrated up to thethird grade a positive relationship betweendegree of structural complexity shown inchildren's own art and the structural com-plexity in pictures they prefer. In uppergrades children prefer more complex pat-

terns (24).Morrison (1935) found that boys and

girls of the same age do not differ inprefe:ence for structural organZation ofpictures but do in terms of subject matter(57).

Hochberg reports that pictorial commu-nication of shape and form is not a simplelearning shell but a process which occursearly in life and is related to how the in-dividual learns to see life (37).

After analyzing the reactions of 77 nur-sery school children to ambiguous pictorialmaterial, Amen reported that the directionof form perception in children moves fromdetails to a larger whole (3).

Because such findings have been achievedwith visual rather than verbal stimuli, theymay or may not apply to perception ofform or other media. However, researchersinterested in the problems of form con-sciousness will consider reviewing and rep-licating some of these studies, substitutingverbal stimuli for the pictorial.

Form may also be studied in relation toart form related to literaturethe film, thetelevision play, and the theatre. In a newstudy of the motion picture, Sheridan,

Owen, Macrorie, and Marcus apply thearproaches of textual criticism to the filmand identify important formal qualities forclassroom study (76). Attempts have beenmade to organize institutions in the hu-manities which relate expression in severalaesthetic media. For example, the State ofMissouri recently published a course ofstudy on "the allied arts" which providesfor instruction in the basic principles un-


derlying and relating literature, music,painting, sculpture, and architecture (1).Among the elements common to most artmedia studied in the course are line, color,value, volume, texture, shape, and perspec-tive for the space arts; and pitch, duration,timbre, and intensity for the time arts.Thus direct attention is being devoted tothe perception of aesthetic form in somesecondary school classes (37).

The Status of Research

Form consciousness then appears to bean important variable in teaching litera-ture, language, and composition, eventhough present knowledge about its de-velopment is limited. This review of re-search and contemporary thinking on se-lected aspects of form perception only in- 14.

troduces an exceedingly complex subject.Much experimentation bearing on the per-ception of form occurs in research which 15.

serves other primary purposes. Classifica-tions of empirical studies rarely list formperception as a distinct category. The re-searcher interested in exploring such prob-lems is advised to read widely and deeplyin the several disciplines mentioned in this 17.

article.

7.

8.

9.

10.

11.

12.

13.

16.

Bibliography

1. The Allied Arts: A High School HumanitiesGuide for Missouri. Publication No. 128 G.Columbia, Missouri: State Department of Ed-ucation, 1963.

2. Allport, Floyd H, Theories of Perception andthe Concept of Structure. London: John WCeyand Sons, 1955.

3. Amen, E. W., "Individual Differences in Ap-perceptive Reaction: A Study of the Responseof Preschool Children to Pictures," GeneticPsychology Monographs, 23 (1923), 319-85

4. Barth, A. L and Robert L Wright, 'TheBackground and Self Picture of Good andPoor Writers," Journal of Communication, 7(Winter 1957-58), 192-3.

5. Boulton, Marjorie, The Anatomy of Prose.London: Rout ledge and Paul, 1954.

6. Brooks, Cleanth and Robert Penn Warren,Modern Rhetoric: Shorter Edition. New York: 24.Harcourt, Brace & World, Inc., 1961.

18.

19.

20.

Brown, Roger W., "Language and Categories,"J. S. Bruner and others, A Study of Thinking.New York: John Wiley and Sons, (1956),247-312.Brown, Roger W., "Linguistic Determinismand the Parts of Speech," Jour 7 :1 of Ab-normal and Social Psychology, 55 (1957 ),1-5.Burt, Cyril, "The Appreciation of Beauty,"British Journal of Educational Psychology, 33(June, 1963), 194-201.Carroll, John, "Language Development inChildren," Psycholinguists, A Book of Read-ings. New York: Rinehart and Winston,(1961), 331-425.Carroll, John B., "Words, Meanings and Con-cepts," Harvard Educational Review, 34(Spring, 1964), 178-202.Carroll, John B., "Vectors of Prose Style,"Thomas Sebeok (ed.), Style in Language.New York: Wiley, 1960, 283-92.Christensen, Francis, "A Generative Rhetoricof the Sentence," Colkge Composition andCommunication, 14 (October 1963), 155-61.Christensen, Francis, "A New Rhetoric: Sen-tence Openers," and "A Lesson from Heming-way," College English, 25 (October 1963),7-17.Commission on English, Preparation in Eng-lish for College-Bound Students. New York:College Entrance Examination Board, 1960.Culpepper-Hagen, Lessie, "An Investigationof Factors Relating to the Writing Effective-ness of Freshmen at the University of Den-ver," unpublished doctoral dissertation, Uni-versity of Denver, 1950.Diederich, Paul, John French, and Sydell T.Carlton, Factors in Judgment of Writing Abil-ity. Princeton, New Jersey: Educational Test-ing Service, 1961.Dienes, Z. P., "On Abstraction and Generali-zation," Harvard Educational Review, 31(Summer 1961), 281-301.Evans, Bertrand, "Writing and Composing,"English Journal, 48 ( January 1959), 12-20.Fagan, Edward R., Field: A Process forTeaching Literature. University Park, Penn-sylvania: University of Pennsylvania, 1964.

21. Fearing, Franklin, "Human Communication,"A V Communication Review, 10 ( September1962), 78-108.

22.

23.

Fleming, Malcolm, "Perceptual Principles andAudiovisual Practices," rv CommunicationReview, 12 ( Spring 1964), 75-87.Forman, Earl, "An Instrument to Evaluatethe Literary Appreciation of Adolescents,"unpublished doctoral dissertation, Universityof Illinois, 1951.French, J. E., "Children's Preferences for Pic-tures of Varied Complexity of Pictorial Pat-

FORM CONSCIOUSNESS, AN hIPOIrrANT VARIABLE IN TEACHING 29

tern," unpublished doctoral dissertation, Uni- 41.versity of California, Berkeley, 1951.

25. Fries, C. C., American English Grammar. NewYork: Appleton-Century-Crofts, 1940. 42.

26. Fries, C. C., Linguistics and Reading. NewYork: Holt, Rinehart, and Winston, 1963, pp.159-85. 43.

27. Gibson, Eleanor J., "Perceptual Develop-ment," Child Psychology, Sixty-second Year-book, Part I, National Society for the Studyof Education, University of Chicago Press, 44.1963, 144-95.

28. Glanzer, M., "Grammatical Category: A RateLearning and Word Association Analysis,"Journal of Verbal Learning and Verbal Be- 45.havior, 1 ( 1962), 31-41.

29. Glanzer, Murray and William H. Clark, "Ac-curacy of Perceptual Recall: An Analysis ofOrganization," Journal of Verbal Learningand Verbal Behavior, 1 (January 1963), 289-99. 46.

30. Goodman, Paul, The Structure of Literature.Chicago: University of Chicago Press, 1954.

31. Gunn, Douglas G., "Factors in the Apprecia-tion of Poetry," British Journal of Educational 47.Psychology, 21 (June 1951 ), 96-104.

32. Guth, Hans, English Today and Tomorrow.New Ymic: Prentice-Hall, 1964.

33. Hall, Robert A., Jr., Sound and Spelling in 48.English. Philadelphia: Chilton Company,1961.

34. Hartman, Frank, "A Behavioristic Approachto Communication: A Selective Review of 49.Learning Theory and a Derivation of Postu-lates," A V Communication Review, 11 (Sep-tember-October 1963), 155-90.

35. Hilgard, Ernest R., "The Place of Gestalt 50.Psychology and Field Theories in Contem-porary Learning Theory," Theories of Learn-ing and Instruction. Sixty-third Yearbook,Part I, National Society for the Study of Ed-ucation, University of Chicago Press (1964), 51.54-77.

36. Hillocks, George Jr., "An Analysis of SomeSyntactic Patterns in Ninth Grade Themes,"Journal of Educational Psychology, 57 (April 52.1964), 417-20.

37. Hockberg, Julian, "The Psychophysics of Pic-torial Perception," A V Communication Re-view, 10 ( September 1962), 22-54.

38. Hook, J. N. and E. G. Matthews, ModernAmerican Grammar and Usage. New York:The Ronald Press, 76-94.

39. Hovland, Carl I., "Effects of Mass Media ofCommunication," Handbook of Social Psy-chology, II. Reading, Massachusetts: Addison-Wesley Publishing Co., (1954), 1062-103.

40. Hunt, Kellogg W., Differences in GrammaticalStructures Written at Three Grade Levels.Report on Cooperative Research Project 1998, 56.U. S. Office of Education, 1964.

53.

54.

55.

Jones, C. W., "Report of the Board of Ad-missions and Relations with Schools," Aca-demic Senate Record, 4 (April 8, 1958), 1-3.Karel, Leon C., "The Allied Arts: Implica-tions for Literature Teachers," The MissouriEnglish Bulletin ( March 1964), 3-6.LaBrant, Lou L., "Study of Certain LanguageDevelopments of Children in Grades Four toTwelve, Inclusive," Genetic Psychology Mono-graphs, 14 ( November 1933), 387-491.Loban, Walter, Language Ability in the Mid-dk Grades. Report on Contract Research SAE7297, Washington, D. C.: U. S. Office of Ed-ucation, 1961.Loban, Walter, The Language of ElementarySchool Children: A Study of the Use andControl of Language and the Rektions amongSpeaking, Reading, Writing and Listening.Champaign, Illinois: National Council ofTeachers of English, 1963.Loban, Walter, Margaret Ryan, and James R.Squire, Teaching Language and Literature.New York: Harcourt, Brace, and Company,1961.Martin, Harold G. and Richard M. Ohmann,The Logic and Rhetoric of Exposition ( Re-vised). New York: Holt, Rinehart & Winston,Inc., 1963.Meckel, Henry Christian, "An ExploratoryStudy of the Responses of Adolescent Pupllsto Situations in a Novel," unpublished doc-toral dissertation, University of Chicago, 1946.Meckel, Henry C., "Research on TeachingComposition and Literature," Handbook forResearch on Teaching. Chicago: Rand Mc-Nally Co., 1963, 966-1006.Meckel, Henry, James R. Squire, and VincentT. Leonard, Practices in the Teaching ofComposition in California Public HighSchools. Sacramento: California State Depart-ment of Education, 1958.McCarthy, Dorothea, "Language Develop-ment in Children," A Manual of Child Psy-chology, Second Edition. New York: Whitey(1954), 492-630.McGurk, Anne, "NSSC News," Colkge Com-position and Communication, 9 (May 1958),103-4.Miles, Josephine, "Essay in Reason," Educa-tional Leadership, 19 (February 1962), 311-3.Miller, George A., "Some Psychological Stud-ies of Grammar," American Psychologist, 17( November 1962), 748-62.Miller, George, G. A. Heise, and W. Lichten,"The Intelligibility of Speech as a Functionof the Context of the Test Materials," Journalof Experimental Psychology, 41 (1951), 329-35.Miller, George A. and Stephen Isard, "SomePerceptual Consequenct of Linguistic Rules,"

I


Journal of Verbal Learning and Verbal Be-havior, 2 ( September 1963), 217-28.

57. Morrison, Jeanette G., Children's Preferences 75.for Pictures Commonly Used in Art Apprecia-tion Courses. Chicago: University of Chicago 76.Press, 1935.

58. Nordberg, Kenneth, "Introduction to Sym-posium on Perception Theory," A V Com-munication Review, 10 (September 1962), 77.1-9.

59. Ohmann, Richard M., Shaw, the Style andthe Man. Middletown, Connecticut: Wesleyan 78.University Press, 1962.

00. Osgood, C. E., "Studies on the Generality ofAffective Meaning Systems," American Psy-chologist, 18 (January 1962), 11-2E.

61. Osgood, C. S., George J. Suci, and Percy H. 79.Tannenbaum. The Measurement of Meaning.Urbana, Illinois: University of Illinois Press,1957. 80.

62. Parker, John P., "Some Organization Var-iables and Their Effect Upon Comprehen-sion," journal of Communication, 12 (March1962), 27-32.

63. Peel, E. A., David H. Russell, Margaret Early,and E. Farrell (eds.), "The Analysis of Pref-erences," Research Design and the Teachingof English. Champaign, Illinois: NationalCouncil of Teachers of English, 1964.

64. Peel, E. A., "A Statistical Study of thePreferences of a Group of Children andAdults as Shown by Certain Tests of AestheticAppreciation," unpublished doctoral thesis,University of London, 1945. (Suinmarized inDavid H. Russell's, Children's Thinking, 147.)

85. Postal, Paul M., "Underlying and SuperficialLinguistic Structure," Harvard EducationalReview, 24 (Spring 1964), 246-66.

66. Read, Herbert, English Prose Style. Boston:Beacon Press, 1952.

67. Richards, I. A., Practical Criticism: A Studyin Literary judgment. New York: Harcouit,Brace, and World, Inc., 1930.

68. Rosenstein, Joseph, "C.oncept Developmentand Language Instruction," Exceptional Chil-dren, 30 (April 1964), 337-43.

09. Rosenstein, J., "Perception, Cognition andLanguage in Deaf Children," ExceptionalChildren, 27 (January 1961), 276-84.

70. Russell, David H., Children's Thinking. Bos-ton: Ginn and Company, 1956.

71. Russell, David H., ;ome Research on theImpact of Reaching," The English journal,47 (October 1958), 398-412.

72. Russell, David H. and Henry R. Fea, "Re-search on Teaching Reading," Handbook ofResearch on Teaching. Chicago: Rand Mc-Nally Co. (1963), 865-928.

73. Sapir, Edward, Language. New York: Har- 89.court, Brace and Co., 1921.

74. Sauer, Edwin H., "The Positive Approach to

Language Success," Maryland English jour-nal, 2 ( Fall 1963), 1-6.Sebeok, Thomas A. (ed.), Style in Language.New York: John Wiley & Sons, Inc., 1900.Sheridan, Marion C., Harold Owen, Ken Mac-rorie, and Fred Marcus. The Motion Pictureand the Teaching of English. New York: Ap-pleton-Century-Crofts, 1965.Siegel, Laurence, "The Instructional Gestalt:A Conceptual Framework," Teachers ColkgeRecord, 62 (1900), 202-13.Siegel, Laurence and Lila Corkland Siegel."The Instructional Gestalt: A ConceptualFramework and Design of Educational Re-search," AV Communication Review, 12(Spring 1964), 16-45.Smith, Nila B. (ed.), Developing Taste inLiterature. Champaign, Illinois: NationalCouncil of Teachers of English, 1963.Squire, James R., "New Directions in Lan-guage Learning," Elementary English, 39(October 1962), 535-44.

81. Squire, James R., The Responses of Adoles-cents Reading Four Short Stories. Champaign,Illinois: National Council of Teachers of Eng-lish, 1964.

82.

83.

84.

85.

86.

87.

88.

Stevens, Peter D., "Linguistics in LanguageTeaching Again: A British Point of View,"The English Leaflet, 41 (Winter 1962), 30-8.Strickland, Ruth G., 'Ilse Contributions ofStructural Linguistics to the Teaching ofReading, Writing, and Grammar in the Ele-mentary School," Bulletin of the School ofEducation, Indiana University, 00 (January1964), 1-44.Strickland, Ruth G., "The Language of Ele-mentary School Children: Its Relationship tothe Language of Selected Children," Bulletinof the School of Education, Indiana Univer-sity, 38 (July 1962), 1-131.Suchman, J. Richard and Mary Jane McCueAschner. "Perceptual and Cognitive Develop-ment," Review of Educational Research, 31(December 1961), 451-62.Walker, Jerry L., "An Investigation into In-dividual Differences and the Structure of Lit-erature," unpublished doctoral dissertation,Wayne State University, 1964.Williams, E. D., L. Winter, and J. M. Woods,"Tests of Literary Appreciation," British kur-nal of Educational Psychology, 8 ( November1938), 265-84.Wilson, James R., "Responses of CollegeFreshmen to Three Novels," unpublished doc-toral dissertation, University of California,Berkeley, 1962.Wohlwill, J. F., "Developmental Studies ofPerception," Psychological Bulktin, 57 (1960),249-88.

i

The Teaching of ThinkingEducators have long said to themselves

and to others that the proper business ofschool is to teach students to think. Yetthis objective has remained a pious hopeinstead of becoming a tangible reality.A variety of factors have militated againstdeveloping a serious and well thought outstrategy for helping students to becomeautonomous, creative, and productivethinkers.

Perhaps the most serious inhibiting fac-tor has been the hazy conceptualizationboth of what is meant by teaching andwhat thinking consists of. Thinking hasbeen treated as a global process whichseemingly encompasses anything that goeson in the head, from daydreaming to con-stmcting a concept of relativity. Conse-quently, the problem of defining thinkingis still before us. The distinctions betweenthe various types of thinking have beendefective also. Even the more serious edu-cational thinkers fail to distinguish thestrategies of thinking, such as problem solv-ing, from the basic cognitive process andskills, such as generalizing, differentiating,and forming concepts. These processes arethe necessary ingredients of problem solv-ing if this strategy is to amount to anythingbeyond s'aeer formality.

Implementation of thinking as an educa-tional objective also has been handicappedby several questionable assumptions. Onerather widely accepted assumption is thatreflective thinking cannot take place untila sufficient body of factual information isaccumulated. Teaching, which follows thisDr. Taba is a Professor of Education at San Fran-cisco State College.Elementary English, XLII (May 1965), 534-542.

HIIDA TABA

assumption, stresses factual coverage andburdens the memory with unorganized and,therefore, rather perishable information.

An equally unproductive assumption isthat thought is an automatic by-product ofstudying certain subjects and assimilatingthe end products of someone else's disci-plined thought. Some subjects are assumedto have this power independently of howthey learn or are taught. Inherently, mem-orizing mathematical formulae or the stepsin mathematical processes is assumed to bebetter training than memorizing cake rec-ipes, even though both may be learned inthe same manner and call for the samemental processrote memory (15).

The analysis of teaching suffers fromsimilar difficulties. Teaching is still viewedlargely as communication of knowledge,and often knowledge is equated with de-scriptive informationthe "what," "who,"and "when" questions are the main diet ofclassroom instruction. As a consequencethe current methods of teaching tend tobe shaped by this emphasis. Research onteaching has skirted the actual process ofteaching and has concentrated instead onsuch matters as personal characteristics ofgood teachers and a priori criteria for ratingeffective teaching (6).

It is no wonder, then, that despite thewidespread acceptance of thinking as aneducational objective little considerationhas been given to the ways in which learn-ing. to think differ from the ways in whichstudents learn knowledge or content ofvarious sorts.

Recent research is producing changes inboth of these areas. Studies of cognitionare under way, which promise a more

31


precise analysis of the processes and of thepsychological dynamics of the mental ac-tivity we call thinking. Some of these stud-ies are concerned with styles of labeling(12), others with strategies of conceptformation ( 2), and still others with whatamounts to the styles in strategies of think-ing ( 7, 10) Important as these studies are,as yet their results cannot be easily trans-lated into the methods for modifying theways of thinking. But at least they are open-ing up the possibility of a scientific ap-proach to the analysis of thinking.

The development of thinking has re-ceived renewed attention also, as exempli-fied by the recent interest in the work ofPiaget and his followers. Piaget's theoriesregarding the nature of thought and thesequences in the transformation of thepatterns or modes of thinking have influ-enced such enterprises as Bruner's (3)analysis of the process of education andSuchman's (13) experiments with inquirytraining.

Some progress is being made in the studyof the teaching process also. Recent studiesof teaching have focused on teaching as itoccurs in the classroom instead of inferringits effectiveness either from a priori notionsof good teaching or from the characteristicsof good teachers. Studies by Hughes (8 ),Flanders (5), and Bellack (1 ) focus ondescribing and cataloguing the teachingacts and on inferring from these descrip-tions their impact on learning in general,on classroom climate, and on achievement.

This article is a description of a study ofclassroom interaction designed to examinethe relationship between teaching strate-gies and the development of cognitive proc-esses (16). The study, conducted under agrant from the Cooperative Research Branchof the US. Office of Education, focused onseveral hypotheses. The central hypothesiswas that it is possible to train students inthe processes of thinking, provided that the

trainable cognitive skills could be identified.The studies of thinking cited above

seemed to have one difficulty in commonas far as the application of their findingsto instruction in the classroom is concerned.The findings regarding the styles of thoughtfail to shed light on the processes by whichthese styles are acquired or to describe theskills on which these styles are founded.

Another hypothesis was that under op-timal conditions this training would resultin an acceleration of the usual develop-mental sequence, such as the appearanceof abstract or formal thought. The studiesof the development of thought and intel-ligence by Piaget and the Geneva school(9*, 11, 14" ) suggest that the evolutionof thought takes place in three stages, es-sentially: 1 ) the sensory-motor stage or thepreverbal intelligence; 2) the stage of con-crete operations or thinking with objectsand concrete events, which stage lasts fromaround two to eleven years of age; and 3)the stage of conceptual or formal thoughtwhich is established between eleven yearsof age and adolescence. There is a ques-tion, however, whether training would alterthese age placements since the availabledata recorded the performance of untrainedchildren, or those with only a minimum oftraining, such as in the study by Ervin (4).It seemed reasonable to assume that if boththe curriculum and teaching strategieswere addressed to the development ofthought, formal thought could appearearlier.

The third hypothesis was that with ad-equate teaching strategies the possibility ofabstract thought would be opened to stu-dents who are now considered to have toolow an IQ to be capable of higher levelsof mental activity.

The study was conducted in elementaryclasses which were using a curriculum in

'Chapter 6.Pp. 107-112.

1

I

I

WrOMPPN

THE TEACHING OF THINKING 33

social studies that systematically stressedthe development of an ability to generalizeand to use generalizations productively.What remained to be done was to specifythe necessary teaching strategies and totrain the teachers in their use, in order tobecome adept at these processes them-selves, and to learn how to induct childrenin the mastery of the required cognitiveskills.

The concept of cognitive tasks.

In an effort to arrive at teachable andlearnable aspects of thought, three cogni-tive tasks were identified: 1) concept for-mation, 2) the development of generaliza-tions and inferences through interpretationof raw data, and 3) the explanation andpredicition of new phenomena by apply-ing known principles and facts.

Concept formation. In its simplest form,concept development may be described asconsisting of three processes or operations.One is the differentiation of the propertiesor characteristics of objects and events,such as differentiating the materials ofwhich houses are built from other charac-teristics of houses. This differentiating in-volves analysis in the sense of breakingdown global wholes into specific propertiesand elements.

The second process is that of grouping.This process calls for abstracting certaincommon characteristics in an array of dis-similar objects or events and for groupingthese on the basis of this similar property,such as grouping together hospitals, doc-tors, and medicine as something to do withhealth care or according to their availabil-ity as an index to the standard of living.Naturally, the same objects and events canbe grouped in several different ways. Forexample, hospitals, X-rays, and surgicalequipment can be grouped together ashealth facilities, as type of services, or as

indices of standard of living, depending onthe purpose of the grouping.

The third process is that of categorizingand labeling. This process calls for thediscovery of categories or labe:s which en-compass and organize diverse objects andevents, such as evolving the concept of aunit measurement from measuring with acup, a yardstick, a plain stick, and a rubberband. It also involves the process of super-and subordination; that is, deciding whichitems can be subsumed under which cat-egory.

In classrooms this cognitive task occurs inthe form of enumerating or listing, such asidentifying a series of specific items notedin a film or reported by a research com-mittee, then grouping similar things, and,finally, labeling the groups.

Interpretation of data and inference. Es-sentially this cognitive task consists ofevolving generalizations and principles froman analysis of concrete data. Several sub-processes are involved. The first and thesimplest is that of identifying specific pointsin the data. This process is somewhat anal-ogous to the listing or enumeration preced-ing grouping. The second process is that ofexplaining specific items or events, such aswhy ocean currents affect temperature,why Mexico employs the "each one teachone" system in eradicating illiteracy, orwhy the way of life in California changedwhen its harbors were opened for freetrade. This process also involves relatingthe points of information to each other toenlarge their meaning and to establish re-lationships.

The third operation is that of forming in-ferences which go beyond that which isdirectly given, such as inferring, from thecomparison of the data on population com-position with data on standards of living incertain Latin American states, that coun-tries with predominantly white populationstend to have a higher standard of living.

1

1


Interpretation of data and formulation ofinferences takes place in the classroomwhenever the students must cope with rawdata of one sort or another, such as com-paring the imports and exports of severalcountries or analyzing and synthesizing thefactors which determine the level of tech-nological development in a given cultureby examining the tools and techniques usedin the production of goods.

Application of principles. A third cogni-tive task is that of applying known prin-ciples and facts to explain new phenomenaor to predict consequences from knownconditions. For example, if one knows whata desert is like, what way of life it permits,and how water affects the productivity ofthe soil, one can predict what might hap-pen to the desert way of life if water be-came available.

This cognitive task requires essentiallytwo different operations. One is that ofpredicting and hypothesizing. This processrequires an analysis of the problem and ofthe conditions in order to determine whichfacts and principles are relevant and whichare not. Second is that of developing infor-mational or logical parameters which con-stitute the causal links between the con-ditions and the prediction and, in fact,make a rational prediction or explanationpossible. For example, if one predicts thatthe presence of water in the desert willcause cities to be built, one needs also tomake explicit the chain of causal links thatleads from the availability of water to thebuilding of cities. These chains may consistof logical conditions, such as that the pres-ence of water is the only condition to makethe soil productive, or from factual condi-tions, such as whether the desert soil con-tains salt or not.

These predictions and explanations areof different orders of generality and com-plexity: for example, the prediction thatcities will be built as a consequence of a

water supply represents a greater leap thandoes the prediction that grass will grow.

In order to develop criteria for effectiveteaching strategies it was necessary toevolve a theoretical construct. In the lightof this construct these processes and theirdevelopment were viewed.

Space permits the description of only afew principles in this theoretical construct.First, the learning of thinking was viewedas essentially an active transaction betweenthe individual and his environment. Thenature of this transaction is only partlycontrolled by the nature of the immediatestimulus. Partly, it is controlled by what-ever mediation is available either in theform of models offered or of guidance thatis available. Chiefly, however, the individ-ual must develop for himself both the con-ceptual schemes and the processes of usingthem. In other words, the environment andtraining become available to the individualonly to the extent that he performs certainoperations cn what he receives. These op-erations cannot be "given" in the ordinarysense of the word. An individual may, forexample, imitate a model of the "if-then"reasoning. But this model remains unpro-ductive unless he internalizes and elabo-rates this process himself.

Second, the development of thought fol-lows a sequence in which the simpler andthe more concrete operations must precedeand prepare for the more complex and theabstract. The elementary school child, forexample, must work out the idea of causeand consequence on concrete material be-fore he can evolve an abstract concept ofcauses and consequences. It appears alsothat the elementary school years are theperiod during which the concrete thinking,or thinking with concrete objects andevents, is being transformed into formalthinking or thinking with symbols. For thisreason an emphasis on the development of

THE TEACHING OF THINKING

certain basic cognitive skills on this levelis crucial.

The idea of a sequential order appliesalso to the mastery of the skills involvedin the cognitive tasks described above. Asa matter of fact, the skills as describedabove could be seen as a series of sequen-tial steps in which each preceding one is

a prerequisite for the success in masteringthe next one. For example, in interpretingdata the differentiation of specific points is

a prerequisite to comparing and contrastingthese points or to seeing relationships be-tween them. The latter is, in turn, a pre-requisite for making inferences, and so on.

Finally, the conceptual schema undergoa constant reorganization. The dynamics of

this reorganization can be visualized as arotation of intake of information into theexisting conceptual scheme and the exten-sion or reorganization of the scheme when-ever the problem or the information re-ceived creates a dissonance because it does

not fit the scheme. For example, a childwhose concept of relationship of altitudeand temperature is that the higher one goesthe colder it gets is jarred into modifyingthis concept when faced with the fact ofhigh altitude combined with high temper-ature. He now needs to extend this conceptto include the concept of geographic zones.

Piaget (11) calls these two processes "as-

similation" and "accommodation," and theseterms will be used in the discussion thatfollows. This rotation of assimilation andaccommodation seems to describe the psy-chological dynamics or mechanism for thegradual maturation of thought, and, assuch, is extremely important in the strat-egy of training.

Hunt (9) points out, in addition, thatthis rotation requires a proper match be-tween the existing conceptual scheme andthat which is required by the new informa-

tion or task. When the requirements of theaccommodation are too far beyond the ex-

isting conceptual scheme it is impossiblefor the child to make a leap. When it istoo close there is no challenge for reor-ganization.

Teaching strategies for cognitive growth.

The concepts of the cognitive tasks to-gether with the principles which governthe development of the cognitive skillshave interesting implications for the for-mulation of teaching strategies.

First, the concept of thinking as essen-tially an active process, in the sense that itcan be learned only by doing, sets the

process of teaching into a new perspective.If students are to develop a cognitive struc-

ture by their own efforts, the usual role ofteaching and of the teacher has to be re-versed. Instead of teaching consisting pri-

marily of communication of information,with the role of the teacher as a fount ofthat information, he needs to become anadroit guide of the heuristic process. In this

kind of teaching strategy the art of askingquestions assumes a crucial role. Questions,furthermore, need a double focus: on thesubstance of what is being discussed andon the cognitive operations. A quistionsuch as, "What materials do we use inbuilding houses?" focuses on the materials

and excludes other characteristics of build-

ing houses such as tools and labor. Thisquestion also asks for enumeration of these

materials rather than explanations of why

these materials are used. Other questions

are addressed to explanation, such as why

women in certain primitive tribes carrythings on their heads or why some coun-tries fail to use the natural resources they

have.The concept of sequence and of the ro-

tation of assimilation and accommodationsuggests, further, that teaching acts, such

as the questions, need to be programmed

to foster an appropriate sequence of learn-

ing. If the learning to apply knowledge to


explaining new phenomena involves mas-tering certain modes of thinking in a cer-tain order, then the questions the teacherasks and the remarks she makes need tofollow that order. If there is to be rotationof intake of new information with tasksthat require changing the conceptual struc-ture, then the teaching acts need to beorganized to stimulate such a rotation. Iftime and pacing of transitions from onemode or level of thinking into another isessential, then the teaching strategy mustmanage this pacing. In other words, teach-ing needs to be addressed first to the ob-jective of thinking; second, seen as a seriesof acts, each of which has a specific peda-gogical function; and, finally, viewed as astrategy or organization of these functions.

In the study described above, Thinkingin Elementary School Children (16), twogroups of teaching functions were identi-fied which seemed to affect the develop-ment of cognitive skills, either positivelyor negatively. First are questions or state-ments made by the teacher or the studentswhich are psychological or managerial intheir function and unrelated to the logic ofthe content. Statements of this type in-cluded approval, disagreement, disapproval,management and reiteration. Second, areteacher questions or statements which givedirection to discussions and are related tothe logic of the content and of the cogni-tive operations sought. This group of func-tions included focusing, refocusing, changeof focus, deviating from focus, extendingthought on the same level, lifting thoughtto a higher level, and controlling thought(16' ).

Focusing questions or remarks establishboth the content topic under considerationand the cognitive operations to be per-formed. They set the cognitive task. Forexample, a question by the teacher such as,

'Chapter 7.

i

"If the desert had all the water it neededwhat would happen to the desert way oflife?" establishes the central content topicfor discussion and calls for prediction ofconsequences. However, to prevent stu-dents from indulging in associative think-ing which follows a single line and opens upnew dimensions, a change of focus may beneeded. Refocusing may be necessary tobring the discussion back to the originaltopic.

Extending thought on the same level ful-fi"s the requirement of allowing a sufficientamount of assimilation before thought islifted to another level, such as making atransition from description of specific pointsnoted in a film, to explaining why certainevents took place in the film or from pre-diction to establishing its validity. This isessentially a strategy in which a number ofstudents are induced to respond to thesame question instead of proceeding froman answer by one student to a question tothe same one, as is usual. Extension ofthought on the same level also assures theparticipation of the slower students. Thisengages them in the initial step of the proc-ess and thus prepares them for participa-tion in the next step.

Lifting of the level of thought occurswhen the teacher or child either gives orseeks information that shifts the thought toa level higher than the previously estab-lished one. Thus, making a transition fromenumeration to grouping and from group-ing to labeling represents lifting of thought.However, pursuing each of these steps byengaging more students or by seeking clar-ification and elaboration would representextension.

Controlling thought occurs when theteacher gives what the students should dofor themselves, such as suggesting a cate-gory or classification or giving explanationsof phenomena observed instead of seekingexplanation from the children.

THE TEACHING OF THINKING

The examples below illustrate the func-tion of focusing, extending, and liftingthought:

(1) C Malobi took the money homewith her. (Child gives specificinformation.)

(2) T What did Malobi do with themoney? (Teacher -zeks specificinformation.)

(3) C She saved it. (Child extendsthought on the level of specificinformation.)

(4) C She put it underground. (Childextends thought on the level ofspecific information.)

(5) C She put sticks and tin over it.(Child extends thought on thelevel of specific information.)

(6) C Before she did that she put it in alittle pot. (Child extends thoughton the level of specific informa-tion.)

In the following example the teacher at-tempts to lift the level of thought from thegiving of information to explanation:

(1) C They carried things in baskets ontheir heads. (Child gives specificinformation.)

(2) T Explain why. (Teacher liftsthought to the level of explana-tion.)

(3) C I suppose they can carry morethings that way. (Child gives anexplanation.)

The combination of these functions to-gether with the pacing of assimilation andthe timing of lifting thought to a new levelis what constitutes the teaching strategy.This strategy is determined by recognizingthat it takes time to learn the skills involvedin these cognitive tasks. They are not inthe class of instantaneous learning. Fur-thermore, presumably there are individualdifferences in the speed with which theseskills can be mastered. Some students maymake a clear distinction after a few at-

I

37

tempts at enumeration, while others n-edto "mess around" for a longer time to dis-cover what is at stake and what the modelof differentiation is. Teaching strategy, tobe effective, must allow variation in pacingeach step, determining how long to con-tinue on the plateau of each step, and whento make a transition to the next one.

In order to assess the effectiveness ofthese pedagogical functions, the verbal re-marks of students were rated as to level ofthought in each of the three cognitivetasks. In effect, these ratings described thesuccessive cognitive operations involved ineach of the tasks described previously.Presumably the process of making infer-ences is a more complex one and of ahigher order than is identification of thepoints in the information presented, thelatter being a prerequisite to the former.In the task calling for inferring from data,a teacher may seek, first, specific informa-tion. She may then attempt to lift the levelof thought to that of explanation, and fol-low with questions designed to elicit infer-ence, etc. The success in eliciting appro-priate responses constitutes the measure ofthe effectiveness of the teaching strategy.

The charting of this flow of teaching actsand of the level of students' responses de-scribes visually the relationship of the two.For example, when the teacher attempts toraise the level of thought too early in thediscussion, this typically results in the chil-dren's returning to a lower level and intheir inability to sustain discussion at thehigher levels of thought. On the other hand,an effective strategy of focusing, extending,and lifting thought, combined with appro-priate pacing of extensions and properlymatched lifts, will result in a gradual move-ment toward higher levels of mental opera-tion by the majority of the students. Afrequent change of focus produces an al-ternation between several levels, a lack ofsustained thought at any level, and a grad-


ual return to the most primitive one. Thesame result occurs when the teacher insertscontrols of thought by giving students what

LEVELS OF THOUGHT

70 Pattern A

60

50

40

30

20

10

F

Pattern B

70so50

40

30

20

10

Pattern C

they should be doing for themselves. Thefigure below illustrates some of these strat-egies:

Ti.F FC FC FC

FFocusingFCChangs of Focus

ru

FC

11FC FC FC FC

Class Discussion Patterns

Pattern A represents a strategy in whichthe transitions are paced appropriately,with the result that the class follows thetransitions from one level of thought to thenext and sustains the thought on each. InPattern B the lifting --.E thought occurs tooearly, with the result that when the fewstudents who could follow it have ex-hausted their ideas the class settles down

1

to the lowest level. Pattern C illustrates adiscussion in which the focus is lost, andthe teacher is forced to keep the discussionalive by constantly changing the topic,without being able to sustain thought onany.

What, then, can be said about the meritsof this approach to teaching thinking?First, the specification of thinking as an

THE TEACHING OF THINKING 39

object of educational effort permits a clear-er analysis of the appropriate pedagogicalfunctions necessary to make this objectiveboth more realistic and attainable. A moreclearly focused target together with morearticulated pedagogical functions may alsopermit a more effective training of teacherFthan is possible when both the nature otcognitive processes and of the appropriateteaching strategies for them are vague andobscure.

Second, it seems that a similar analysisof other educational objectives, such as theenhancement of the ego concept, thegrowth in affective domain, and the de-velopment of a creative approach to lit-erature and art, might eventuate in thekinds of description of instructional proces-ses which may provide the material for thedevelopment of a generic and a functionaltheory of learning and teaching.

Finally, such an approach to teachingthinking may reach students who are nowrelatively untouched by instruction. Theresults of the study described here indi-cated a lack of correlation between theperformance on the test of Inference inSocial Studies and the students' IQ. Anal-ysis of tapescripts suggested that a carefulstructuring of the sequential steps in mas-tering the basic cognitive skills and an ap-propriate timing and pacing of the transi-tions from one level of thought to anotherare the chief ingredents to opening the pos-sibility for a highc.. /evel of mental func-tioning to students of low ability (asmeasured by tests of intelligence). Anal-ysis of a few individual cases indicatedthe possibility that among the so-calledslow students are many who are only slowabsorbers. Evidently, when the amount ofi tformation to be assimilated is reducedand opportunity is provided for systematicprocessing of that information, such stu-dents can function on abstract levels ofthought.

Bibliography

1. Bellack, A., J. R. Davitz, et al., The Languageel the Classroom. New York: Institute of Psy-chological Research, Teachers College, Co-lumbia University, 1963.

2. Bruner, J. S., Jacqueline J. Goodnow, and C.A. Austin, A Study of Thinking. New York:Wiley, 1956.

3. Bruner, J. S., The Process of Education.Cambridge: Harvard University Press, 1960.

4. Ervin, Susan M , "Training and Logical Op-eration of Children," Child Development,31 (1960) 555-563.

5. Flanders, N. A., Teacher Influence, Pupil At-titudes, and Achievement. Prepublication man-uscript of a proposed research monograph forthe U.S. Office of Education, Cooperative Re-search Branch, Washington, D.C. 1900.

6. Gage, N. L. (Editor), Handbook of Researchon Teaching. A Project of The AmericanEducational Research Association. Chicago:Rand McNally and Company, 1963, Chap. 11.

7. Guilford, J. P., "Basic Conceptual Problemsin the Psychology of Thinking," Ann& NewYork Academy of Science, 91 (1961 ) 9-19.

8. Hughes, Mane, et al., Development of theMeans for the Assessment of the Quality ofTeaching in Elementary School. (Mimeo.)Salt Lake City: University of Utah, 1959.

9. Hunt, J. McV., Experience and Intelligence.New York: Ronald Press, 1961, Ch. 5-9.

10. Peel, E. A., The Pupil's Thinking. London:Oldbourne, 1960.

11. Piaget, J., The Psychology of Intelligence.New York: Harcourt, 1950.

12. Sigel, I., Cognitive Style and Personality Dy-namics. Interim report. Merrill-Palmer Insti-tute, 1961.

13. Suchman, J. R., The Elementary School Train-ing Program in Scientific Inquiry. U.S. Officeof Education, Title VII, Project 216. Urbana:University of Illinois, 1964.

14. Taba, Hilda, Curriculum Development. Theoryand Practice. New York: Harcourt, Brace &World, 1962.

15.. Taba, Hikia and F. F. Elzey, "Teaching Strat-egiw and Thought Processes," Teachers Col-lege Record, Vol. 65, No. 6, March, 1964.

16. Taba, Hilda, S. Levine, and F. F. Elzey,Thinking in Elementary School Children, U.S.Office of Education, Cooperative ResearchBranch Project No. 1574. San Francisco: SanFrancisco State College, 1964.

1

HARatm Amami

Concept Formation in ChildrenDiverse routes have been taken toward

understanding how children form Cj...ceptsand a multiplicity of methods have beenutilized. The leading investigator of chil-dren's concepts, ,i m Piaget, has describedspecific concepts employed by childrenand, from these, has inferred certain formalcharacteristics of their thinking processes.His approach is a generic one in which heattempts to distinguish and describe stagesof thinking and to categorize the psycho-logical functions within these stages. Incontrast, the main emphasis of this paper ison how concepts are actually learned.

Many investigators have attempted tostudy how concepts are attained understandard conditions by studying what thechi:d does in the course of attaining a newconcept. Some of these investigators inferthe psychological processes of the childfrom the concept he learns during the ex-periment while others study introspectivereports as more direct reflections of theprocesses they employ. Researchers alsovary greatly in their definitions of conceptformation. Even among the investigatorswho have attempted to study the psychol-ogy of concept acquisition experimentally,there are vast differences in the types ofmaterials employed and the nature of thebehavior which is measured. I shall discussthree intriguing problems within the area

Dr. Amster is an Assistant Research Psychologist inthe Institute of Human Learning at the Universityof California, Berkek y, which is supported in partby a grant from the Nitional Science Foundation.Her own research in this area, some of which iscited herein, is supported by the U. S. Office ofEducation, Project #2243.Elementary English, XLII (May 1965), 543-552.

and present some research which is for themost part, illustrative of the newer psycho-logical approaches to concept formation inchildren.

Processes of Concept Learningin Children

There must be innumerable ways bywhich children and adults form concepts.However, two types of responses whichseem basic to the many possible ways offorming concepts are associative processesand deductive processes. Associative proc-esses are tbought to be intuirive in thatconcepts are arrived at on the basis ofboth verbal and nonverbal associations tothe presented exemplars. Despite the factthat the outcome of such association mayseem to be intuitive, I th k that the asso-ciations which "emerge" automatically havebeen formed through a basic type of rotelearning. By contrast, the deductive proces-ses are thought to involve logical reasoningand hypothesis testing on the basis of strat-egies. I think that these do not occur spon-taneously without prior learning. Rather,they are learned and could also have beenacquired originally through a basic form ofrote learning. Processes related to thesehave been described by Woodworth (36)as alternative processes by which conceptsmay be formed, but it is likely that theycan occur simultaneously and interact ( 23,24, 32). More specifically, I hypothesizedthat the associative processes would bemore universal than the deductive proces-ses. Therefore, when adults form conceptswithout deliberately attempting to do so,the associative processes should be domi-nant, but when they make a deliberate

40

CONCEIT FORMATION IN CHILDREN

attempt to form concepts, the deductiveprocesses might be dominant. The evidencewas consistent with the hypothesis.

I had expected that efficiency in conceptformation would improve with the use ofhypothesis testing, and that this would bedemonstrated by greater efficiency of in-tentional concept formation than inciden-tal concept formation among children oldenough to use lcgical reasoning. In severalexperiments (1, 25) involving detectionand recall of features common to a set ofspatial forms, kindergarten children andfourth graders excelled under the inten-tional set. The trends of the means by nomeans confirmed the hypothesis that theintentional set should be more beneficial tothe older children than to the younger chil-dren. These results strongly suggest thatinstructions are as important a factoramong kindergarten children as they areamong older ones, and, therefore intentionalinstructions should be used to increaselearning in that age group. However, itdoes not necessarily follow that the kinder-garten children are henefitted in the sameway as the older children. This is illustratedindirectly by the fact that kindergarten chil-dren improved with practice under the in-tentional instructions to a greater degreethan under incidental instructions, while incontrast, the older children showed greaterimprovement under the incidental instruc-tions. Quite possibly the first test of con-cept recall alerted the older children to theneed to engage in intentional learningwhile the younger children may have con-tinued to respond on an incidental basis.

As pointed out by Vinacke (33 ), deduc-tive processes of concept learning may be-come increasingly evident as children growolder. Various findings, including those ofPiaget (22) and Inhelder and Piaget (9),are consistent with this concept of theemerging dominance of logical reasoning.Although his analyses involve processes of

41

thinking much more specific than the twotypes we have described, the stages he dis-tinguishes involve an increase in the use ofmental operations with age. Moreover, thetypes of operations employed seem tochange with age, culminating in the ap-pearance of logical operations about age11 to 14. Similarly, much of the vast lit-erature surveying the types of conceptsformed by children cf varying ages is gen-eraly consistent with this position. For ex-ample, Olver (19 ) studied how seeminglydifferent objects are grouped into equiva-lence classes. In addition to studying thesemantic bases of grouping, e.g., the typesof features which children use in classify-ing objects, she studied the syntacticalstructure of the equivalence sets. With re-spect to the former, she found, in agree-ment with other studies, that the tendencyto categorize on a perceptual basis declinedfrom age 6 to 9 while the tendency to cat-egorize on the basis of function increased.With respect to the syntactical structure ofthe groups, she singled out three types ofstructures: heaps or arbitrary collections,described earlier by Vygotsky (34); com-plexes in which the various members areincluded in accordance with a rule thatdoes not account uniformly for the inclu-sion of members in the set; and superordi-nate groups in which one rule accounts forall the objects in the set. Among the varioustypes of complexes were those character-ized by edge matching in which an objectis included on the basis of some similaritywith a neighboring member. For example,a six-year-old would say that a banana,peach, and potato are alike because bananaand peach are yellow and peach and potatoare round. Others involved assorted objectsrelated by their relevance to a story or theirrelation to one object although the objectsmay not have been related to each other.As expected, the occurrence of superordi-nate groups increased from age 6 to 19

1


while a parallel decrease in complexivegroupings occurred. Needless to say, theclassification of objects into superordinategroups involved the ability to test hypoth-eses to determine whether each object isor is not a member of the set. Consequent-ly, these results are consistent with thepostulation of the increasing dominanceand effectiveness of logical processes withage. Tnis has been shown among adoles-cents who have been found to improve inefficiency of concept attainment and in theuse of strategies. Yudin and Kates (37)found that 14- and 16-year-olds were notdifferent from each other but were superiorto 12-year-olds in this regard.

Teachers may be especially interested tolearn that children of superior intelligencemay attain concepts by testing hypotheseswhile children of average intelligence do sothrough more of a rote process. Os ler andTrautman (21 ) have found corroborativeevidence for this hypothesis. The ability ofchildren aged 6, 10, and 14 to identify theconcept two was studied. In one group thechildren were given pairs of stimuli ofwhich one was two solid black circles andthe other was 1, 3, 4, or 5 solid blackcircles. In the other group the pairs ofstimuli again consisted of one card withtwo figures and another card with somenumber of figures other than two. How-ever, in this case colored pictures of com-mon objects were used, and different ob-jects were used to represent the positiveand negative instances within a pair. Thecrucial difference between the two sets ofstimuli was believed to lie in the amount ofinformation which was irrelevant to theconcept two. Half the children were giventhe simple stimuli and the other half weregiven the more complex stimuli. It wasexpected that more hypotheses could begenerated on the basis of the complexstimuli than on the basis of the simpleones; consequently, children who were test-

1

ing hypotheses should do more poorly onthe basis of the complex stimuli than thosewho were learning by a rote process. Asexpected, the highly intelligent children inall three age groups made more errors inachieving the concept from the complexinstances than from the simple instances.The complex instances did not, however,slow down the average children; it sloweddown the superior children who were un-doubtedly testing diverse hypotheses toarrive at the concept. I believe that thisexperiment also supports the idea that theprocesses interact. Bright children must useassociative processes as do average chil-dren, but they would have more differentassociations to any two stimuli than aver-age children do. Therefore, the many in-teresting and varied associations interferewith the production of the simple associa-tion two which emerges more readily inthe average child. Both nrocesses couldthus contribute to explaining the results.It is interesting to note that in this experi-ment the use of hypothesis testing wasfound to be a relatively inefficient strategyin contrast to what we assume the usualcase must be.

Dramatic confirmation of the role of in-telligence in concept formation was ob-tained by Osler and Fivel (20) using a tasksimilar to the above. Among children ofsuperior intelligence there were more sud-den learners than gradual learners whilethe reverse was true for children of averageintelligence. 'This suggests that the brighterchildren use deductive reasoning in con-firming the solution.

The Acquisition of Word Meanings

Words, meanings, and concepts may beconsidered aspects of verbal conceptswhich may be considered as separate whena teacher tries to communicate a new ideaor when concept formation is analyzed bya researcher. The need for and a discussion

CONCEPT FORMATION IN CHILDREN 43

of distinctions among these terms appear inan interesting paper by Carroll (6). In gen-eral, in learning a word the child mustperceive the various instances of a givenword as similar and eventually differentiateamong the contexts in which a given soundor sound pattern is used. A concept, on theother hand, involves a response to the sim-ilarities among a series of experiences andto the range of variations in such instances.This response to the range of variationsserves to permit differentiation among con-cepts.

A "meaning" can be thought of as a con-cept which is standardized by society and"when we say a word stands for or namesa concept it is understood that we arespeaking of concepts that are shared amongthe members of a speech community" (6).Psychologically speaking, the processes ofacquisition of concepts and meanings mustbe identical. However, there may be con-siderable variation depending on whetherthe formation of a concept involves thelearning of new responses to a set of experi-ences or merely regrouping concepts whichhave been learned in the past.

There are undoubtedly many ways bywhich words and concepts come to acquiremeaning to the child or the adult. It isquite possible that during the early years,processes which are of prime importancerequire the physical presence of the refer-ent or concrete examples of the conceptand some manner of reinforcement contin-gent upon the response. This could taketwo forms ( 28), i.e., when a child sayswater," water is then brought, and as a

consequence the child learns the meaningof water, and when a child is praised forsaying water whenever he sees or experi-ences water. However, as the child's knowl-edge of the world increases, it is quitelikely that he acquires the meaning ofwords and concepts from other words. Forexample, he may do so either from the

presentation of synonyms or roughly equiv-alent words or verbal definitions. He is toldthat joy means happiness or is given amore formal definition in terms of genusand species. But between these two typesof situations lies what may be the mostwidespread manner by which verbal mean-ings are acquired. It involves the acquisi-tion of verbal concepts from their use inverbal contexts.

The influence of the verbal context onthe way in which a meaningless word ac-quires connotative meaning has been ex-tensively studied by Staats and Staats andtheir associates (29, 30, 31). When mean-ingless words are merely paired with mean-ingful words they acquire the connotationof the meaningful word in accordance withthe expectation based on conditioning prin-ciples. These studies are outstanding inthat they involve a finegrain analysis of theacquisition of meaning.

It is probably only with very young chil-dren that the acquisition of new conceptscan be studied, although the extent to whicha concept is new is undoubtedly a matterof degree. The acquisition of language andconcepts has probably received more at-tention among Russian investigators thanAmericans in recent years. Kollsova (16)investigated the role of the breadth of ex-perience with objects on the formation ofconcepts of them. The concept doll was in-vestigated in one experiment. Two groupsof children about 20 months of age wereshown a doll 1,500 times in the course ofseveral months. For one group the experi-menter employed only three statements,"Here is a doll," "Take the doll," and "Giveme the doll." For the other group he em-ployed thirty different statements such as"Look for the doll," "Rock the doll," "Seatthe doll," etc. The total amount of verbal-ization by the experimenter was the samefor both groups, but the group for whichthe presentation of the doll was accom-


panied by the greater variety of verbali7a-lions was superior in their learning of theconcept, doll. This was shown by the abil-ity of the children to select dolls fromamong other toys. On appropriate instruc-tion the group receiving greater variety ofverbalization restricted their selections todolls whereas the other group dld not. Ina similar experiment using children of 19months of age, two factors were varied:the number of different examples and thevariety of verbalizations. There were threegroups: in one group, one book was shownonce with one verbalization; in a secondgroup, twenty different books were shownwith only one verbalization; in a thirdgroup, one book was shown with twentyverbalizations. The group with twenty dif-ferent verbalizations was superior to theothers, while the group with only onebook and one verbalization learned theleast about the meaning of book. Theycould not separate books from other ob-jects and could not carry out commandsusing the word book.

Werner and Kaplan (35) attempted tostudy the course of acquisition of wordmeanings in children who ranged in agefrom 8% to 13% years. They devised aningenious method, the Word Context Test,which involved the presentation of severalseries of sentences each containing anartificial word. For example, the followingsentences were among those from whichthe concept gather was to be abstracted:

All the children will lidber at Mary'sparty.

The police did not allow the people tolidber on the street.

The people lidbered about the speakerwhen he finished his talk.

The children were asked to define theconcept as each succeeding sentence waspresented. The first sentence remained inview when the second was shown. Theauthors attempted to analyze developmen-

tally the reported changes in verbal mean-ing and found many changes with age, e.g.,a striking drop in the occurrence of con-cepts which consist of parts of a sentence(sentence-contextual concepts ) at about10% years. They believe that word meaningis derived from the verbal context anddescribe many processes by which this oc-curs. The most important among these isdecontextualization, a process of acquisi-tion involving a gradual refinement inmeaning in which the specific features ofthe contexts are dropped, until all that re-mains are the features of meaning whichcharacterize the word in all contexts. How-ever, this final stage of decontextualizedmeaning is only to be found in relativelymature individuals. Bruner (3) analyzedthe process of acquisition in this situationin a different way. He claims that the solu-tion to each series of sentences is the wordthat has the highest probability across theseries. This suggests that there is some as-sociative process whereby the context prob-ability of the new word increases fromsentence to sentence. I think there is alsoa strong possibility that acquisition is atwo-stage process involving not only asso-ciation, but also deductive reasoning thoutwhether the associations to the sentences"fit" the context.

It seems quite reasonable to assume thatthe formation of verbal concepts must of-ten be unintentional and that unintentionalconcept formation should be largely asso-ciative. As mentioned in the discussion ofthe research by Osler and Trautman (21)above, consideration of concept formationunder a large as compared with a smallvariety of instances suggests that the largevariety should, in the case where manyhypotheses are desirable, make for superiorconcept learning for both associative anddeductive processes. However, the differ-ence in effectiveness between a large and asmall variety of instances should be greater

CONCEPT FORMATION IN CHILDREN 45

for children employing both processes, i.e.,older children. A small variety might bemore effective if memory for the instancesor attention to the task is a problem. Thus,it would be expected that a large varietyshould be clearly more effective than asmall variety for older children who woulduse both deductive and associative proces-ses, but that the superiority should be lessclear for the child who does not employlogical reasoning and/or the more distract-ible child who may tend to be younger orless capable.

A child may not employ logical reason-ing either because he has not yet acquiredthe tendency or ability to do so effectivelyor because he is forming concepts uninten-tionally. In one study (25) I compared thenumber of concepts correctly elicited onthe basis of a large variety and a smallvariety of different sentences. A large vari-ety was usually six different sentences anda small variety was either three or two dif-ferent sentences repeated twice or threetimes, respectively. A general superiorityof a large variety of sentences over a smallvariety would be expected on the basis ofthe theory that concept elicitation is amatter of association. The more sentenceswhich are added, the greater the probabil-ity that one of the sentences will elicit theconcept and the greater the probabilitythat the set of sentences will do so. Whenlogical reasoning is used, children shouldbe checking to see whether hypothesizedconcepts 'fit" the sentences. Increasing thenumber of different sentences increases theoccasions on which a hypothesized conceptmight be rejected and, therefore, on thisbasis a large variety should be more effi-cient than a small variety. In general, chil-dren about six years old and about elevenyears old tended to form more correct con-cepts on the basis of the larger variety ofcontexts and to apply them more appro-priately in a new situation. However, as

anticipated, there was a reverse trendamong two groups: older children whowere forming concepts incidentally andolder children who were relatively ineffi-cient at the task. I believe both processesare employed by older children who formconcepts intentionally, but when they doso unintentionally, the associative processis dominant.

On the basis of our limited evidence, itseems possible that the associative processis primary not only among young children,but also among older children who havelimited memory or attention span andamong younger children.

In a new study, I am investigating fur-ther the process by which the concept isacquired. Fourth graders were each givenonly one sentence corresponding to eachconcept. The same sentences and conceptswere used which had been used in someof the experiments mentioned above. Nostatistically reliable difference was foundbetween the frequency with which the con-cept was given in response to the "best"sentence of the set and the frequency withwhich it was given in response to the setsof 2, 3, or 6 sentences. Certainly one wouldexpect it to be better in response to thegroup of sentences than to any single sen-tence if there is some kind of summationof associations across the series. Furtherevidence of absence of summation was ob-tained by comparing the response to thegroup of sentences with the expected re-sponse which would be predicted on thebasis of the responses to the individualsentences. It was found that the expectedresponse to the group of 3 or 6 sentenceswas significantly greater than the obtainedresponse to the group. However, it can atleast be said that the occurrence of thecorrect response to the group is strikinglygreater than the occurrence of this responseto the average single sentence that com-prises it.


The Role of Verbalization in theAcquisition of Concepts

Russian psychologists (17, 18) have pio-neered in the investigation of the role oflanguage in the regulation and control ofbehavior. These investigators have postu-lated and also found increasing linguisticcontrol of behavior with age. Americanstudies on the role of verbal mediationgenerally suggest that the critical age forthe occurrence of mediation may be dif-ferent for different situations (27). Amongthe foremost American investigators of therole of semantic mediation in children'sconceptual learning are the Kendlers whohave intensively studied reversal shift be-havior. In a simple version of their proce-dure, the nursery school child is shown alarge black (white) square and a smallblack (white) square for the initial dis-crimination task. He is rewarded everytime he selects the small square, regardlessof whether it is on the left or the right.After they have learned to select the smallsquare consistently, the children are givena similar task in which the objects differin two dimensions. Some are then rewardedfor shifting their choice to the other ob-ject, some are rewarded for learning torespond to a cue from the new dimension,and for comparison purposes, some con-tinue to be rewarded for the same choiceas before. Other reversal shift proceduresinvolve presentation of pairs of objectswhich differ in only two dimensions for theinitial discrimination learning task. For thesecond task, the pair may differ in onlyone dimension. This task measures theability to shift to the opposite value of theinitial dimension and also the ability tolearn to respond to a value of the newlyvaried dimension. The ability to shift isbelieved to involve the same mediated re-sponse as the initial task although the overtresponse is changed. The new response re-

quires a new mediated response as well asa new overt response. Consequently, ifmediation is employed, the reversal shiftshould be learned faster than the new re-sponse. Several studies support this hy-pothesis. The reversal shift behavior of col-lege students (12) supported the hypoth-esis being in direct contrast to that of rats(11 ). For kindergarten children as a whole,no difference between reversal shift andnew learning was found, but when thegroup was divided into fast and slowlearners a dramatic difference appeared(13 ). The fast learners acted like collegestudents and the slow learners tended toact like the rats. It was expected thatnursery school children would resemble theslow learning children of kindergarten age,but that requiring them to verbalize mightenable them to behave in a more maturefashion. However, compared with an unin-structed group, instructing the children toverbalize both the right and the wrongvalue of the dimension during the initialtask did not improve their ability to shift(15). In another experiment one-third ofthe nursery school children were instructedto verbalize the right value, e.g., blackwhen black was rewarded, one-third wereinstructed to verbalize the wrong value,e.g., black when white was rewarded, andone-third were not instructed to verbalizeIn this case, verbalization of the right valueof the dimension did facilitate the abilityof nursery school children to shift and italso facilitated this ability in second graders(14 ). For children of both ages, irrelevantverbalizations retarded the ability to shift.There was also a trend for the relevant ver-balizations to be more beneficial than noverbalization for the younger group, butfor the older children there was no differ-ence. This suggests that the seven-year-oldsspontaneously used verbal mediation andwPre, therefore, not additionally assisted bythe instruction to do so.

CONCEPT FORMATION IN CHILDREN

Despite the strong evidence for the in-creasing use of verbal mediation with age,there is a curious fact which runs throughmuch of the literature on children's learn-ing. It has been noted in a wide varietyof investigations and also among childrenof varying ages (4, 8, 15, 18) that thespontaneous verbalizations of children donot necessarily or even consistently cor-respond to the behavior which they exhibit.For example, the Kendlers (15) mentionthat some nursery school children namethe correct value but continue to make thewrong choice for many trials. The kinds ofcorrespondence between language and per-formance require a good deal more inves-tigation before this puzzle can be illumi-nated further, and some of this work is nowgoing on in Russian laboratories and atHarvard by Bruner and his associates.

Despite the fact that subjects in conceptformation experiments can often learn toclassify objects correctly without beingable to define the class, giving an instruc-tion to verbalize distinguishing character-istics of the class has been found to ben-efit 11- to 13-year-old children (10). In thisexperiment one group was asked to classifycolored pictures of birds into gunkle birdsand bunkle birds. They were then askedto define the concepts. The other groupperformed the same task, but during theclassification task they were asked to guesshow to tell a gunkle bird from a bunklebird. Although no evaluation of the answerwas given by the experimenter, there werealmost twice as many good definitionsgiven in this group. Although it is impos-sible to tell whether the instructions tonotice similarities and differences or thenecessity to verbalize them produced thefacilitation, the results certainly suggestthat older children can benefit from suchinstructions.

Using a task in which children between5 and 7 years of age were to arrange nine

47

plastic glasses in a 3 x 3 matrix, Brunerand Kenney (5) studied the linguistic de-scriptions of children's behavior. The glas-ses varied in diameter and height, andwere initially shown to the children, cor-rectly ordered in the matrix. The childrenwere acquainted with the matrix and askedhow the glasses in the columns and rowsare alike and how they differ. Followingthis, the glasses were scrambled and thechild was asked to reproduce the matrix.Finally, a transposed situation is presented.The shortest, thinnest glass is placed in adifferent corner from the original and thechr_l is asked to make a matrix, leavingthis glass in the new position. Almost allthe children succeed in the matrix-matchingalthough, as expected, the older childrenperform more quickly. On the other hand,most seven year olds succeeded in thetransposition test while hardly any of theyounger children did. However, the mostinteresting analysis concerned the relation-ship between the type of verbal descriptionmade by the children and their ability toperform the transposition test. There werethree types of verbalizations: Dimensionalin which the child distinguished both se-mantically by naming both ends of eachscale, e.g., "that one is higher and that oneis shorter; Global in which the languagedid not refer specifically to one dimension,e.g., "that one is bigger and that one islittle"; Confounded in which a dimensionalterm would be used for one end of a con-tinuum and a global term for the other,e.g., "that one is tall and that one is little."Interestingly enough, the children who usedthe confounded descriptions were far morelikely to fail on the transposition task thaneither of the others. Also, the investigatorswere unable to find a relationship betweenthe language the children used and theirability to perform the simple task of re-producing the matrix.

The widespread occurrence of this lack

-


of direct correspondence between languageand action certainly compels us to conclude

that although language assumes an increas-

ingly important role in the processes ofconcept acquisition, one cannot assumethat thought and action are directly cor-

related or that the relationship betweenthem is a simple one.

There is an additional body of experi-mental literature on the role of verbaliza-tion in the acquisition of concepts whichdeals with the relative efficacy of varioustypes of verbalization about objects on theability to manipulate those objects as exem-plars for the formation of a concept. In astudy by Dietze (7 ) four- and five-year-old children learned names for similar-appearing spatial forms. One group learnedsimilar-sounding names and theother grouplearned different-sounding names, and thosewho learned the different-sounding namesexcelled on all measures of concept forma-

tion employing the spatial forms. Rasmus-sen and Archer ( 26) found thai comparedwith making aesthetic judgments, merelylearning nonsense names for objects did notaid college students in using the objects toform concepts in a later situation. Thegroup which made artistic judgments aboutthese objects must have noted and verbal-ized more similarities and differences amongthe objects than the group which merelylearned names for the objects. Therefore, inthe subsequent situation, those who hadmade the aesthetic judgments excelled overthose who learned the name. With someprocedural modifications, this experimentwas repeated on eleven-year-old childrenwith basically the same result (2 ). Thecrucial question must be not whether ver-balization is employed, but to what andhow the verbalization applies. It surelydemonstrates the relevance of the type ofverbal label to the ability to form conceptson the basis of the labelled objects.

References1. Amster, H., "The Effect of Instructional Set

and Variety of Instances in Children's Learn-ing," unpublished manuscript, 1964.

2. Amster, H. and L. Marascuilo, "The Effect ofType of Pretraining and Variety of Instanceson Children's Concept Learning," Journal ofExperimental Child Psychology, 1965 ( in

press).3. Bruner, J. S., J. J. Goodnow, and R. A. Austin,

A Stut'y of Thinking. New York: Wiley, 1956.

4. Bruner, J. S., "The Course of CognitiveGrowth," American Psychologist, 19, (1964 ),1-15.

5. Bruner, J. S. and H. Kenney, "The Develop-ment of the Concepts of Order and Proportionin Children." Reported in J. S. Bnmer, "TheCourse of Cognitive Growth," American Psy-chologist, 19 (1964 ), 1-15.

6. Carroll, J. B., "Words, Meanings and Con-cepts," Harvard Educational Review, 34 (1964),178-202.

7. Dietze, D., "The Facilitating Effect of Wordson Discrimination and Generalization," Jour-nal of Experimental Psychology, 50 (1955),255-260.

8. Furth, H. G., "Conceptual Discovery andControl on a Pictorial Part-Whole Task as aFunction of Age, Intelligence, and Language,"Journal of Educational Psychology, 54 (1963),

191-196.9. Inhelder, B., and J. Piaget. The Growth of

Logical Thinking from Childhood to Ado-lescence. New York: Basic Books, 1958.

10. Johnson, D. M. and C. A. O'Reilly, "ConceptAttainment in Children: Classifying and De-fining," Journal of Educational Psychology,55 (1964). 71-74.

11. Kelleher, R. T., "Discrimination Learning asa Function of Reversal and Nonreversal Shiftsin Human Concept Formation Behavior,"Journal of Experimental Psychology, 49 (1955),

165-174.12. Kendler, H. H. and M. F. D'Amato, "A Com-

parison of Reversal Shifts and NonreversalShifts in Human Concept Formation Behav-ior," Journal of Experimental Psychology, 49(1955 ), 165-174.

13. Kendler, T. S. and H. H. Kendler, "Reversaland Nonreversal Shifts in Kindergarten Chil-dren," Journal of Experimental Psychology,(1959 ), 56-60.

14. Kcndler, H. H. and T. S. Kendler, "Effect ofVerbalization on Reversal Shifts in Children,"Science., 134 (1961 ), 1617-1620.

15. Kendler, T. S., H. H. Kendler and D. Wells,"Reversal and Nonreversal Shifts in NurserySchool Childen," Journal of Comparative andPhysiological Psychology, 53 (1960).

t

CONCEPT FORMATION IN CHILDREN

16. Kortsova, N. M., The Formation of Higher 26.Nervous Activity of the Child, in G. Razran,"The Observable Unconscious and the Infer-able Conscious in Current Soviet Psychophys-iology: Interoceptive Conditioning, Semantic 27.Conditioning, and the Orienting Reflex." Mos-cow: Medgiz, 1958. Psychological Review, 69(1962 ) 344-354.

17. Liublinskaya, A. A., "The Development ofChildren's Speech and Thought," Psychologyin the Soviet Union, Simon, B., Ed., StanfordUniversity Press, 1957, 197-204.

18. Luria, A. R., "The Role of Language in theFormation of Temporary Connections," Psy-chology in the Soviet Union, Simon, B., Ed.,Stanford University Press, 1957, 115-129.

19. Olver, R. R., "A Developmental Study ofCognitive Equivalence," reported in J. S.Bruner, "The Course of Cognitive Growth."American Psychologist, 19 (1964 ), 1-15.

20. Osler, S. F. and M. W. Fivel, "Concept Attain-ment: 1. The Role of Age and Intelligence inConcept Attainment by Induction," Journal ofExperimental Psychology, 62 (1Q61 ), 1-8.

21. Osler, S. F., and G. E. Trautman, "ConceptAttainment: II. Effect of Stimulus Complexityupon Concept Attaimnent at Two Levels ofIntelligence," Journal of Experimental Psy-chology, 1 (1961 ), 9-13.

22. Piaget, J., The Psychology of Intelligence.London: Routledge, 1952.

23. Podell, H. A., "Two Processes of Concept For-mation," Psychological Monographs, 72 (1958).

24. Podell, H. A., "How Are Concepts Learned?"American Psychologist, 1962 ( abstract).

25. Podell, H. A., "The Effect of the Variety ofInstances on the Production of Verbal Con-cepts," American Psychologist, 1963.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

49

Rasmussen, E. A., and E. J. Archer, "ConceptIdentification as a Function of Language Pre-training and Task Complexity," Journal ofExperimental Psychology, 61 (1961 ), 437-44111-,ese, H. W., "Verbal Mediation as a Func-tion of Age Lever Psychological Bulletin, 59(1962), 502-509.Skinner, B. F., Verbal Behavior. New York:Appleton-Century-Crofts, 1957.Staats, A. W., C. K. Staats, W. G. Heard,and L. P. Nims, "Replication Report: Mean-ing Established by Classical Conditioning,"Journal Experimental Psychology, 57 (1959),64.Staats, A. W., C. K. Staats, and W. G. Heard,"Denotative Meaning Established by ClassicalConditioning," Journal of Experimental Psy-chology, 61 (1961 ), 300-303.Staats, C. K. and A. W. Staats, "MeaningEstablished by Classical Conditioning," Jour-nal of Experimental Psychology, 54 (1957 ),74-80.Vinacke, W. E., The Psychology of Thinking.New York: McGraw Hill, 1952.Vinacke, W. E., "Concept Formation in Chil-dren of School Ages," Education, 74 (1954),527-534.Vygotsky, L. S., Thought and Language, (Ed.and trans. by E. Hanfmann and G. Vakar).New York: Wiley, 1962.Werner, H., and E. Kaplan, "The Acquisitionof Word Meaning," Child Development Mon-ographs, 1951.Woodworth, R. S., Experimental Psychology.New York: Holt, 1938.Yudin, L. and S. L. Kates, "Concept Attain-ment and Adolescent Development," Journalof Educational Psychology, 54 (1963), 177-182.

1

!

MURRAY S. MIRON

The Quality of Qualification

A n:tural language such as English isthe result of a series of historic accidents.Such languages contrast sharply with con-trived languages such as mathematics, com-puter iastructions ( Fortran, Algo 60), orEsperanto in that the ecological processesof selective survival of the former are eithertruncated or completely absent in the lat-ter. It is the fact that the language is livingwhich is important, despite the circum-stances that its breath often may mist themirror reflecting the essence of wisdom.The outcome of the struggle to be under-stood by lazy people, while still adequatelycharacterizing what impresses our sensesand what doesn't, is distilled and ongoingin the present state of the language, evenin the noises we make with our eating ap-paratus merely to show sentience. Thus wemay study natural languages for clues asto the psycholinguistic characteristics ofthe people who speak them.

Such has been the goal of the group of

researchers at the University of IllinoisInstitute of Communication Research. Ourmain attack on the psychological functionsof language has been through the use ofan instrument called the semantic differen-tial. This instrument, developed by CharlesOsgood (6, 8), uses the affective or con-notative terms of a language to character-ize the meaningful differences among theconcepts employed by language users. Everylanguage has developed some means orother of specifying the intensity and qual-

ity of the experiences impinging upon our

Dr. Miron is a Co-Director of the Center for Com-parative Psycholinguistics at the University of Illi-nois, Urbana.Elementary English, XL1I (May 1965), 553-558.

human senses and, by metaphorical exten-sion, the intensity and quality of thoseconcepts not open to sensory confirmation.

In English such qualification is largely con-trolled by the class of words ordinarilycalled adjectives. The semantic differentialtechnique asks subjects to rate the degreeto which they feel that the concepts to be

differentiated are characterized by an ef-ficiently small set of contrastive adjectival

attributes.

Form of the Task

A typical administration form of thesemantic differential would contain any-where from 10 to 70 adjectives with theiropposites, separated by seven step divi-

sions, e.s.,

50

Fathergood_._:_:_:_:_:_bad

strong_:_:_:_:_:_:_weakThe subjects are instructed to indicate with

a check mark the degree to which they

feel that the adjectival "scale" is expressive

of the concept displayed at the top of thelist of these scales. The closer the checkmark is placed to either one side or theother of the contrastive attributes expressed

by the scale, the more extreme the related-

ness. Subjects are instructed to proceed

rapidly through each of Ae scale items,judging for each the degree of relatedness

of the attributes for the same conceptterm. After a short practice period, a sub-ject typicaPv can be expected to complete

from 20 to 30 such judgments in one min-

ute's time.

THE QUALITY OF QUALIFICATION

Theonj and Standardizationof the Technique

It should be obvious that the degree ofprecision of the characterization of a con-cept in terms of adjectival attributes shouldbe dependent on the choice of the partic-ular scale and concept items chosen. Moregenerally, the hope that such a techniquecould be at all useful is dependent uponthe demonstration of a stable and parsi-monious structure in the ways in whichnatural languages organize the domain ofattribute terms.

Although it is theoretically possible toenumerate all of the qualifiers in any na-tural language, in practice their number istoo large to make such enumeration prac-ticable. Thus, although we might theoretic-ally hope to find optimal discriminationamong concept terms by having subjectsrate their degree of relatedness with everyadjective, such a technique would hardlybe parsimonious. Even granting that wecan achieve neax optimal discriminationamong concepts with something less thanthe total list of adjectives, it is still possiblethat the meanings of these adjectiveschange so drastically when used as attributeterms for different concepts that no dis-crimination would be possible. Hence therequirement that the natural language musthave a stable structure of attribute terms.

The hope of the discovery of such astructure in natural languages is not asoptimistic as might appear to be the casefor anyone sensitive to the nuances ofqualification language makes available to itsspeakers. We are all well aware that colorscan con i a bewildering variety ofshades; the on-artists might not be aware,however, that all of these colors may beexactly characterized by reference to onlythree fundamental attributes: hue, satura-tion, and brightness. Assuming 50 degreesof variation in each of these three dimen-sions, their combinations would produce in

51

excess of 100,000 different colors. Alterna-tively, we all know that cardboard boxesare manufactured in a variety of shapes;yet, if we wish to order a particular boxfrom New jersey or California, we needonly specify the three dimensions of height,width, and depth in order to pick that onebox, from the many, which we desire.'

Fortunately, there is a statistical tech-nique which can be employed to discoverthe dimensional structure of a contentdomain for which the structure is unknown.This technique is known as factor analysis.From the patterns of the intercorrelationsbetween a finite but large number of char-acteristics of the domain of interest, weinfer the existence of a basic set of dimen-sions whose combinations could give riseto all of the observed variations. For ex-ample, on intuitive grounds alone, it iseasy to see that if two characteristics ofthe items of a content domain always yieldthe same quantity, as wou!d be the casefor the measurement of characteristics Aand B of the diagram, we may safely dis-card one of the characteristics as irrelevant

atarackvislic 11

II 1

I 1

I 1

I 1

as.....aroas.as.as.as.low

CAaracte.nbc A C .-triensbc C

'Note that we assume that we will be unconcernedakaut thickness of cardboard, finish, etc. We as-sume that vo'urne or color is the characteristicwhich is essential to our decision. We shall returnto the problem of ig-ored characteristics later inthe paper.

1


to the unique specification of the domain.2lf, however, we were to observe a charac-teristic which never gave us any knowledgeof any other characteristic, if the charac-teristic never systematically varied withany other characteristic, as the C elarac-teristic of the diagram theoretically be-haves, it should be equally clear that sucha characteristic supplies us with valuablenew information in the specification of thecontent items.

The Content Domain of Adjectives

In many respects, the adjectives of Eng-lish behave like colors. The combination ofan adjeciive and its opposite has the effectof canceling the import of the tenns takenseparately, just as the additive mixture ofcomplementary hues produces neutral gray.We may identircy adjective opposites bythe fact that they are mutually substitut-able in utterances. For every statementwhich contains one of a pair of adjectivalterms there is at least a theoretically appo-site utterance which may contain its oppo-site. For every strong man, building, drink,and point, a weak counterpart is madeavailable by the language. Not so with apair like good and puce, as can readily beseen by the semantic anomalies which re-sult when thes- two terms are substitutedas qualifications of the above list.

Although il might be argued that colorsexist in continuously variable gradationsand adjectives are finite in number, thelimitations of our visual apparatus limit thelexicon of colors far more than is the theo-retical limit placed upon the number ofqualifiers a language may contain None-theless, it still may be the case that a small

'The techmque does not give us any clue as towhich characteristic shodd be discarded. Much ofthe misunderstanding about factor analysis stemsfrom the belief that we will retain the -true" char-acteristic. That we would somehow maintain A ofthe diagram rather than B. The hope is unwar-ranted.

number of basic dimensional attributes ofthese adjectives can serve to specify theirmultitudinous diversity, just as we saw tobe the case with colors.

Let us assume an n dimensional mean-ing space built on the analogy of the threedimensional color space. Each adjective ofthe language specifies some specific locusof the possible variations in meaning en-compassed by the entire space, just as eachcolor specifies a particular variation of thechromatic visual stimulus. The closer twoadjectives lie within this space, the morenearly similar their meanings. Just as redand orange would be expected to lie closerin the color space than red and blue, sowe should expect that good and nice wouldbe in closer proximity than good and heavy.But how are we to systematize our intui-tions about such meaningful similarities?The answer is deceptively simple. We en-quire of the speakers of the languagewhether or not they would judge a largeset of concepts to be appropriately quali-fied by the same set of adjective terms.If many speakers agree that mothers mayappropriately be said to be good, nice, andgentle and that football players (by some,e.g., high school coeds) are said to be good,nice, and rough, we are well on the way toa specification of the fact that the languageexhibits more similarity between good andnice than between each of these terms andgentle. If we extend ti.,is approach to alarge list of now- concepts and adjectivesand allow gradations of degree of qualifica-tion such as extremely good, slightly nice,we have what has been called the semanticdiffetential technique. The degree to whichsubjects choose and rate the same adjectiveterms as appropriate to the list of conceptsestablishes the degree of intercorrelationbetween those terms in the language thesubject employs.3

31 have called such intercorrelations indices of dis-tributional synonymy for want of a better term.

1

I

I

;

THE QUALITY OF QUALIFICATION 53

The Dimensional Structure of AdjectivesIn a number of independent replications

of these procedures we have consistentlyround that the adjectives of English exhibita stable and relatively simple structure ofessentially three dimensions. The first andmost important of these dimensions is whatwe have called evaluation. This dimensionrepresents those qualification terms whichhave reference to the worth or value ofconcepts and is best represented by suchadjective pairs as good-bad, nice-awful,happy-sad, and pleasant-unpleasant. Thesecond dimension we have named potencyand is best characterized by such terms asstrong-weak, big-little, powerful-powerlessand heavy-light. The third dimension is ac-tivity, represented by such terms as fast-slow, active-passive, and noisy-quiet. Inour studies these three dimensions takentogether account for approximately 50% ofall the meaningful variation in the se-mantic space encompassing numerous largesamples of the qualifiers of English. Otherand different dimensions do exist, but theircontribution to the total variation in mean-ing becomes progressively smaller and morespecific.

The earlier coarmation of this structurefor English encouraged the group of in-vestigators at the University of Illinois toattempt to extend the generality of ourfindings to all natural languages. In thesummer of 1963 we established the Centerfor Comparative Psycho linguistics undergrants from the National Science Founda-tion and the Institutes of Mental Health.With the help of some 40 social scientistsin 20 countries around the world we beganto collect semantic differential data in 16different languages.

To avoid the potential bias of translation,and resultant ethnocentric bias, the proce-dures for selecting qualifiers that wouldeventually serve as the dimensions of judg-ment in semant;c differential tasks had to

be entirely intracultural; each language/culture group must determine its own de-scriptive scales. However, the overall meth-odology of these intraculturally independ-ent samplings had to be standardized inorder to make possible the interculturalcomparisons required for testing the prima-ry hypothesis of structural equivalence.Additionally, it was clear that testing thelimits of generality demanded as hetero-geneous a sample of both languages andcultures as could be obtained practicably.Our solutions to these requirements wereto employ a standard word associationmethodology in each of a wide range ofselected languages. Native speakers of eachlanguage are asked to provide qualificationassociates for a standard list of translationequivalent noun concepts. These elicitedqualifiers represent the pool of items fromwhich we draw the descriptive scales as-sumptively representing the infringe quali-fication dimensions in each language. Byhaving recourse to the actual speakers ofthe language in order to derive our quali-fiers, we may claim that the kinds of quali-fiers obtained are being determined solelyby those atfributes considered importantby the native speakers and not some out-side agent.

Using an appropriately diverse set ofstimulus terms, we obtain 10,000 qualifierresponses in each language. From theseitems we wish to select a small number ofrepresentative qualifiers displaying the fol-lowing characteristics: (1 ) The qualifiersmust have high overall frequency of usage,(2) the qualifiers must display high diver-sity of usage, i.e., the qualifiers must beapplicable to a broad range of substantivecontexts, and (3) the qualifiers must ex-hibit relative independence of usage, i.e.,they must represent uncorrelated attributesof the substantive concepts.

Indices of these requirements are com-puted from the responses obtained from


the subjects. All responses are ordered byfrequency of occurrence to satisfy require-ment one: the number of noun stimuli giv-ing rise to a particular adjective responseis counted in order to satisfy requirementtwo. Both tabulations are then combinedinto a single index known as the infor-mation statistic H. This statistic gives high-est weight to those responses which haveboth a high frequency of occurrence anda large diversity of eliciting stimuli. Thethird requirement is indexed by the inter-correlations between the patterns of dis-tribution of the responses to the stimulusitems. Responses occurring to the samestimuli are said to be distributionally syn-onymous, and if of lower H value are dis-carded from tNe list.

The remarkable thing is that even at thisstage, similarities in the H ordering of theadjectival responses are clearly evi 4.entacross all of the languages. Translationequivalent qualifier types are found occu-pying the same H ranks in the variouslanguages well beyond any chance expecta-tions. As to the structure of these qualifiers,we have been able to identify in everyinstance the existence of the three dimen-sions of evahaation, potency, and activityin that general order of prominence.

The Qualification Structure of Children

The fact that the basic structure of qual-ification can be characterized as an explicitdimensional system which transcends lan-guage differences suggests that we aredealing with an attribute which is pnmi-lively human. In fact, we have found thatthe qualifications used by children displayprecisely the same dimensional stnictureas that found for adults. DiVesta (1), em-ploying the same procedures used in ourcross-linguistic investigations, has testedchildren in grades 2 through 6 and foundessentially identical factors even in theyoungest children. Other studies, notably

one by Ware (9), have shown that thestructure is not modified when subjects ofdivergent intelligence are employed. In asense, none of this is surprising; children,and even the very dull, differ from thelower organisms in the one crucial regardthat they speak the language.

Interpretive Summary

The object of concern in these studieshas been homo loquens. Man is the onlytalking animal. He has been taught to talkby other talking men in an historical re-gress that extends back in time to ourearliest inferences about man. Being a char-acteristic of the human species, one shouldexpect to find massive similarities trans-cending language boundaries; yet the ob-vious differences between languages havetended to obscure the more subtle, butmore fundamental, similarities from theview of language scientists and scholars.

The researches reported here have ledfrom the standard American college soph-omore, to Americans differing in age, sex,and other characteristics, and even to peo-ples differing widely in both the structureof their language and the nature of theircultures. Nevertheless, the modes of qual-ifying used by these people are structuredin the same fashion with remarkable reg-ularity.

We are left with the question of why thedimensions of primitive affect should beevaluation, potency, and activity. The or-ganism's earliest, most persistent and mostsignificant interactions with its environ-ment can be characterized as varying be-tween the dichotomous states of physio-logical satisfaction and dissatisfaction. Inthe earliest stages, these states are induceddirectly by the properties of the stimuliencountered, but later they become asso-ciated with signs of such originally ade-quate stimuli. Still later, and most signifi-cantly for man, these states come under the

1

THE QUALITY OF QUALIFICATION 55

control of social signs produced by others,particularly linguistic signs. In addition tothis satisfaction-dissatisfaction dichotomy( which we may identify with the evalua-tive dimension), stimuli, and therefore theirsigns, ce -. vary in at least two other re-gards: tbe amount of work or effort theydemand from the organism (potency di-mension ), and the degree to which theyarouse the organism (activity dimension).

Constructed as we are as human organ-isms, we have, affective reaction systemsdesigned to respond grossly in a limitednumber of ways, but to a very wide rangeof itimuli. These reactions by no meansexhaust the ways in which we are capableof finely representing the minute distinc-tions in our environment, but they are nowand always have been fundamental forsurvival. While the dimensions of affectivemeaning appear to be innate and hence arecommon to all humans, the way in whichhumans will react to particular signs de-pends upon learning. Thus, the particularmeaning of particular signs will differ fromone organism to another; but the dimen-sions which can specify the location of thismeaning in the semantic space of the or-

ganism will be common to all organisms.With some clues as to what these dimen-sions may be, we can proceed to studythese differences in meaning for differentindividuals with a common metric.

References1. DiVesta, F. J., "The Qualification Structure of

Children," unpublished report. Syracuse Uni-versity, 1964.

2. Fries, C. C., The Structure of English. NewYork: Harcourt, Brace and Co., 1962.

3. Greenberg, J. H., (Ed.), Universals of Lan-guage. Cambridge: MIT Press, 1963.

4. Hannan, H. H., Modern Factor Analysis.Chicago: University of Chicago Press, 1960.

5. Miron, M. S., and C. E. Osgood, "The Multi-variate Structure of Qualification," in R. A.Cattell and S. Hammond (Eds.) Handbookof Multivariate Methods in Experimental Psy-chology. Chicago: Rand-McNally, 1965(spring ).

6. Osgood, C. E., "The Nature of Measurementof Meaning," Psychological Bu//etin, 49(1952), 197-237.

7. Osgood, C. E., M. S. Miron, and W. IC.Archer, "The Cross-Cultural Generality ofMeaning Systems," unpublished report, Uni-versity of Illinois, 1962.

8. Osgood, C. E., G. J. Suci, and P. Tannen-baum, The Measurement of Meaning. Urbana,Illinois: University of Illinois Press, 1957.

9. Ware, E. E., "Relationships of Intelligenceand Sex to Diversity of Individual SemanticMeaning Spaces," unpublished doctoral dis-sertation, University of Illinois, 1958.

JAMES J. CALLAGIIER

Expressive Thought by GiftedChildren in the Classroom

For many years there has been an un-fortunate gulf between those professionswhose purposes are to create or teach ex-pressive thought, and phychologists whosegoals are to attempt to understand moreabout man's thinking processes and devel-oping intelligence. The psychologist hasbeen fascinated with how a person caninternally process information in such away as to generate a product different fromthe information received. This interest inproductive thought and the more specificsubject area of creative thinking, has waxedand waned over the past few decades butnow seems to be enjoying an importantrevival.

A review of the research in this area( Stein and Heinze, 1960) has presented afine picture of the two main avenues of ap-proach to studying this problem, investi-gating the cre...tive thinker and investigatingthe creative process. While the study ofthe creative person has resulted in manyinteresting hypotheses, it is also fraughtwith difficulty. Many of the great creativegeniuses that we would wish to study havelong since passed from the scene and biog-raphers often tell us more about themselvesthan about their subjects. It is also difficultto weigh the life of a Wilde against thestandards of modern society, or to try anddetermine whether Wagner's financial ir-

Dr. Gallagher is a Professor in the Institute forResearch on Exceptional Children at the Univer-sity of ninois, Urbana.

The research reported herein was made possibleby Cooperative Research Project Grant #965 fromthe U. S Office of Education.

Elementary English, XLII (May 1965), 559-56841

responsibility was an essential or irrelevantpart of the total creative person.

Recent work by MacKinnon (8) and byBarron ( 1) in their studies of creative vsnoncreative persons have suggested thatthere are certain personality characteristicsthat seem closely related to the creativeperson. Those judged most creative devi-ated more from the norm on measures ofmaladjustment but also possessed more egostrength than the less creative. They havean openness to experience and intellectualrisk taking that permits them to take theatypical view.

The educator cannot be satisfied merelywith dealing with the once-in-a-lifetimestudent who is destined to place his markon history or to create enduring literature.His duty is a broader one which entailsincreasing the productive potential in allstudents.

Many investigators felt the key to im-provement of thinking abilities was thestudy of the process by which the indi-vidual generates new thoughts. In order todetermine what factors might influence stu-dent productivity or non-productivity inthought expression, one particular focus ofthis research interest has been the class-room environment. To study such a widerange of expressiveness as occurs in thisspecial environment, one needs a methodof cataloguing the classroom content anda theory of thinking ability that can aid ininterpretation of the scene.

In this study, the theoretical model ofGuilford (5, 6) was used as a basis for

. analyzing the expressive behavior. Guil-

56

-

EXPRESSIVE THOUGHT BY GIFTED CHILDREN IN THE CLASSROOM 57

ford's structure of intellect was developedthrough a decade of research studies usingfactor analytic methodology. The pal ame-ters of this theoretical structure consist inthe operations of thinking, the contentwithin which these operations are per-formed, and the products which result fromthe performance of these operations uponthe content.

Other investigators with similar ap-proaches but different theoretical modelsare Smith and, Meux ( 11); Flanders (3);Taba, Levine and Elzey (12); Medley andMitzel (7); and Bloom (2), to mentiononly a few.

The purposes of the present study wereto describe the kinds of thinking opera-tions taking place in this variety of class-rooms and to determine factors or variablesthat seemed to have influence on the ex-pressive abilities of gifted students.

Subjects

The subjects in the present study were118 boys and 117 girls, in junior and seniorhigh school, placed in ten classes for giftedstudents. Each student was chosen formembership in these class groups on thebasis of IQ scores and proven academicattainment. The mean verbal IQ was 131for boys and 130 for girls; the mean non-verbal IQ score was 131 for the boys and128 for the girls. Since group IQ scoreswere used in this instance and these scoresoften are found to be lower on group thanindividual tests, it was assumed that thegroups chosen here represented at least thetop five percent of their age group on thisdimension. Gifted children were chosen forparticular study due to their propensity toexpress themselves in the classroom setting.

Classification System

The present category system was con-structed primarily on the operations of in-tellect as Guilford has described them. Five

primary categories have been developed.These are: cognitive memory (C-M ), con-vergent thinking (CT ), divergent thinking(DT ), evaluative thinking (ET ), and rou-tine ( R ). The routine category consists ofthe familiar and conventional interpersonalmaneuverings of speakers in the manage-ment activities of the classroom setting, andin a number of categories defining behav-iorsverbal and otherwiseexpressing af-fect and feeling tone. In order that thereader have some idea of the dimensionsof each of these areas of cognitive behaviorin the classroom, a brief description is givenbelow:

Cognitive-memory operations representthe simple reproduction of facts, formulae,or other items of remembered contentthrough use of such processes as recogni-tion, rote memory, and selective recall.Examples of cognitive-memory perform-ance can be seen in the following:

T: What were some of the main pointscovered in our discussion about mer-cantilism?

Mary: One of the things we learned wasthat there was an attempt to keepa favorable balance of trade.

T: What is a conjunction?The above examples of teacher-student

interchanges do not require the student tointegrate or associate facts; the questionscan be handled by direct reference to thememory bank. The sole duty of the stu-dent is to select the appropriate responsefrom his store of remembered items.While factual information is clearly in-dispensable to the development of higherthought processes, it is also obvious thatit would be a sterile and uninteresting classthat dealt exclusively with this type ofquestion, never moving into the challengeand excitement of more complex opera-tions.

Convergent thinking represents the anal-ysis and integration of given or remem-bered data. It leads to one expected end-


result or answer because of the tightlystructured framework through which theindividual must respond. Examples of con-vergent thinking are as follows:

T: If I were going to town A 170 milesaway and drove at 50 milts an hour,how long would it take me to getthere?

Bob: Three hours and twenty-four minutes.T: Can you sum up in one sentence what

you think was the main idea in Paton'snovel, Cry the Beloved Country?

Pete: That the problem of the blacks andthe whites in Africa can only be solvedby brotherly love; there is no otherway.

Thus, convergent thinking may be in-volved in the solving of a problem, in thesummarizing of a body of material, or inthe establishment of a logical sequence ofideas or premisesas, for example, in re-porting the way in which a machine works,or in describing the sequence of stepAywhich the passage of a bill through Con-gress is accomplished.

Divergent thinking represents intellec-tual operations wherein the individual isfree to generate independently his owndata within a data-poor situation, or to takea new direction or perspective on a giventopic. Exakiples of divergent thinking are:

T: Suppose Spain had no '. been defeatedwhen the Armada was destroyed in1588, but that instead, Spain hadconquered England. What would theworld be like today if that had hap-pened?

Sam: Well, we would all be speakingSpanish.

Peg: We might have fought a revolution-ary war against Spain instead ofEngland.

Tom: We might have a state religion in thiscountry.

These examples represent teacher-stimu-lated divergent thinking, but it need notalways be teacher-generated. In a regulardiscussion of the "spoils system," a studentmay come up with the following:

Well, sure, the spoils system might be

a good thing when a political party isgetting started, but what about whenthere's no party systemlike in the UnitedNations?H :re the student reveals his ability to

take off from an established fact or factsand see further implications or unique asso-ciations that have not been requested orperhaps even thought of by the teacher.Instances of this type of self-initiated stu-dent behavior would also fall under thegeneral category of divergent thinking.

Evaluative thinking deals with matters ofjudgment, value, and choice, and is charac-terized by its judgmental quality. For ex-ample:

T: What do you think of Captain Ahabas a heroic figure in Moby Dick?

Bob: Well, he sure was brave, but I thinkhe was kind of mean the way hedrove the men just because he hadthis crazy notion of getting back atMoby Dick.

T: Is it likely that we will have a hardwinter?

Mary: Well, I think that the pattern of highpressure area suggests that we will.

T: Who was the stronger PreE'dent,Jackson or Adams?

Mike: A dams.In the first of the above examples, the

student is asked to construct a value di-mension of his own in terms of what heconsiders "heroic," and then to make ajudgment as to where on this value dimen-sion he would place Captain Ahab. In thesecond response, the student is asked tomake an estimate or to give a speculativeopinion or assessment of probability. Athird possibility involves entering a quali-fication or disagreement, wherein the re-spondent would offer a modification of aprior judgment of another student; or hemay state a counter-judgment, in which hedeclares direct opposition to the statementof the previous speaker.

The final category, Routine, contains alarge number of miscellaneous classroom


activities. Included here are the attitudinaldimensions of praise and censure of othersand of self. Also present are dimensionsof structuring, a kind of prefatory remark,telling in advance what the speaker in-tends to say or do, or what he expectssomeone else to say or do. Other charac-teristic occurrences, such as humor, as wellas the ordinary "routine" classroom man-agement behaviorseven to requests to

100

90

80

70

60

50

40

30

20

10

o

100

40

10

0

2

70

60

30

0

50

90

so

I 2

HAT

close the door or asking what time it is

are included in this primary category.Procedure

Each of the ten classes were tape re-corded for five consecutive hour sessions(two classes were taped in the fall andagain in the spring to check on classroomconsistency). In addition I. the tape re-cordings, two observers were present inthe classroom and took extensive notes on

Figure 1

PROPORTION OF THOUGHT PROCESSESASKED FOR BY THE TEACHER

3

JACK

m

SOCIAL STUDIES

II

s

5

100

90

80

70

60

50

40

30

20

10

0

ENGLISH

-100

90

80

70 .60 .

50

40

30

20

10 -

0

IS

2

IDEA

KING

COGNITIVE-MEMORY III DIVERGEN1

CONVERGENT 1:3 EVALUATIVE 0

1


classroom activities such as demonstration,charts, blackboard material, etc. In addi-tion, they noted the more obvious attitudinalrelationships in the classroom such as cen-

sure, praise, frustration, humor, etc. Eachtranscribed classroom session was classified

statement by statement by trained judgesworking with the scoring manual describedabove. The results of this analysis werethen compared with test findings and

HAT

teacher ratings on the dimensions of cog-nition (IQ tests and tests of divergentthinking), attitude (a semantic differentialscale and sentence completion test), andsociability (teacher ratings).

Results

Figures 1 and 2 indicate the proportionsof each major thought operation producedby teacher and by the boys in class over

Figure 2

PROPORTION OF THOUGHT PROCESSESGIVEN BY STUDENTS

SOCIAL STUDIES

JACK

90

80

70

60 -

50

40

30

20

BOYS

100

90

80

70

60

50

40 .

30 -20

10 -0 1

ENGLISH

100

90

80

70

60

50

40

30

20

10

o

--

-

-

1 2

IDEA

MN;.'.....1.'...1-;,%:.....

3

KING

4

il:''. '....:.

'....'.,.;'

'f.;.t.:.:...4.

s

1 2 3 4

COGNITIVE-MEMORY". DIVERGENT

CONVERGENT ME EVALUATIVE E3

5

'/ A

EXPRESSIVE THOUGHT BY GIFTED CHILDREN IN 'ME CIASSROOM 61

five consecutive class sessions in two sec-tions of social studies and two sections ofEnglish. Figure 2 represents only the boys'responses since the pattern of the responseswere the same for the two sexes, althoughthe total output was often different.

It will be noted that the majority of bothteacher questions and student responsesfell in the Cognitive-Memory area. Therewould seem to be an inescapable baselineof factual and memorized material thatforms the basis of any classroom discus-sion. The more the class tends to a lecturetype of operation, as opposed to discussion,the more the total responses lean in thisdirection.

The second most frequent category interms of usage was Convergent Thinking.Much of the Convergent Thinking centeredaround explanations and conclusion draw-ing along one given line of thought andforms a recognizable part of most class-room discussions. It appears that class ses-sions can operate on these two thoughtoperations alone, without much evaluativeor divergent thinking. Indeed, in Jack 3session the entire class hour was devotedto teacher questions only in these twocategories. The evaluative and divergentcategories appeared to be particularly de-pendent upon teacher stimulation. The per-centage of responses in these two areaswere low unless stimulated by deliberateteacher action.

A visual comparison of Figures 1 and 2will indicate the close relationship betweenthe patterns of teacher questions and thestudent responses. It is not difficult to un-deistand why. A question such as, "VVhen

did Columbus discover America?* is hardlycalculated to bring forth divergent orevaluative thinking by the student, althoaghit is not impossible"Columbus didn't dis-cover America, Leif Erickson did." A ques-tion such as 'Compare the performances ofOlivier and Burton in Hamlet" can hardly

fail to evoke student evaluative responses.In short, the teacher controls the expressivethought patterns of the class to a largeextent. It will be noted that the Englishsections, jack and King, were almost devoid

of divergent thinking while being heavilyloaded with convergent thinking. This wasthe result of this particular teacher's stylerather than a direct result of the subjectmatter. Censure shouldn't be automaticallyapplied for the absence of certain thoughtcategories either. No standard exists tosuggest that one teacher pattern is 'better"than another. The definition of abetter" inthis case leads to very complex argumentsrelated to the desirable outcome of teach-ing.

If divergent thinking is considered desir-able, it can be inserted in any subjectmatter by changing the type of questionasked. For example, the teacher in askingthe students to outline a short passage orspeech bad one best way in mind. Thestudents wor1-,41 their way towards thatbest way though successive approxima-tions (convergent thinking). If divergentthinking had been a goal of the teacher hecould have posed the problemHow manyways could you reorganize this passage,keeping the same content, so as to mean-ingfully change the outline?

Other analyses suggest, as is obvio-s inthe Hat series, that a teacher will modifyhis pattern from one class session to an-other in the same series and that oneteacher's style is distinctively different fromanother teacher's. It is not possible to ob-

serve and catalogue one day's performanceand expect to have a typical sample ofo le teacher's style. There also is the sug-gestion that the same teacher may show adifferent style or teaching depending upondo group he is working with (observe in

Figure 1 the differences between Hat andIdea groups with the same teacher teaching

1


the same content on the same day to twodifferent sections).

In each of the classroom groups studied,boys were significantly more expressive insome of the thought dimensions used inthis studyctlian the girls, although the di-mension and intensity of difference variedwith the class group and subject matter.In no instance was there a significant dif-ference in favor of the girls. The generalconclusion then was that boys tended tobe more expressive in the classroom situa-tion. The male sex role of expected aggres-siveness being learned at this age level wasone of the possible explanations offered b3account for this result. No sex differenceswere found on written expressiveness!

Although originally it had been expectedthat different students would show differ-ent expressive patterns (i.e., one specializ-ing in divergent thinking, another concen-trating on evaluation), this was not found

to be the case. Instead, the high intercor-relations of all of the categories in class-room expressiveness, mostly in the .80's,made it clear that there was one generalfactor of expressiveness that was operatingand that it a student was expressive indivergent thimcing the chan,es were veryhigh that 11,, scored high on the other cat-egories also. There remained the interest-ing question as to what other variablesmight relate to this expressiveness factor.

An adjusted score for classroom expres-siveness was obtained for each student bydividing his production for the five daysby the total production for the whole classfor those class sessions. For example, ifSam gave 10 divergent responses while hisclass gave 100 responses in this category,his adjusted score would be 010 for diver-gent thinking responses. If Mary gave 20such responses while her class was givinga total of 200 responses in this category,

Table IThe Comparison of Expressive and Nonexpressive Gifted Students on

Cognifive Factors and Teacher Ratings

Expressive NonexpreuiveN Mean a N Mean 0'

Verbal IQ B 27 133.00t 9.56 27 128.44 7.93G 27 128.22 10.46 29 127.24 9.36

Non-verbal IQ B 27 129.33 7.62 27 129.89 8.45G 27 128.81 11.14 29 127.69 11.21

Uses Breadth B 26 21.58 6.46 27 1927 5.53G 27 19.96 8.03 29 17.48 5.67

Consequences B 26 15.00 3.90 27 14.11 3.47Breadth G 27 1&78 2.65 29 13A8 2.89

Consequences B 26 39.88 18.81 27 31.78 19.22Solutions G 27 41.04 20.05 29 47.52 17.82

Teacher Rating B 26 8.65" 3.99 27 11.74 t 11Cognition G 27 10.56 3.77 29 12.86 4.57

Teacher Rating B 26 2.31 .79 27 2.41 .93

Sociability G 27 2.00 .02 29 2.00 .70

tSialifeara a .10 ind of cosfsaksce Ssidiam at .05 ind of ce--Ammugi- 5-osifkmg- at .01 level of -`www0c.w---

.... . _....._.

EXPRESSIVE THOUGHT BY GIFTED CHILDREN IN ME CLASSROOM 63

her adjusted score would be 010, or thesame as Sam's. In this way, it was possibleto compare performance across classgroups.

In order to compare the expressive vs.nonexpressive groups, the top third of thetotal sample recorded was chosen on thebasis of their adjusted class scores andcompared with the bottom third of thegroup on adjusted class scores on expres-siveness. A part of this comparison can befound in Table I. The information for boysand girls was considered separately sinceinteresting sex differences in patterns ofperformance were noted throughout thecurrent study.

It might be expected that students scor-ing higher on verbal IQ would be the mostexpressive students in the classroom butthis expectation held, in the present study,only for the boys. Table I shows that themean I.Q. of the expressive boys was 133 asopposed to 128 for the nonexpressive boysand this difference exceeded only a prob-ability level of <.10, but the differencesbetween expressive and nonexpressive girlswas almost nonexistent. No differences werefound between the two groups, on non-verbal intelligence scores.

The Uses and Consequences tests weredeveloped by Guilford and have been usedas one of the key measures of divergentthinking in a number of other studies (4,13). The students are asked questions suchas, "How many different uses can you thinkof for a brick?" or "What would happen ifeverybody were born with three fingersand no thumb?" The answers are then con-sidered in terms of the number of differentcategories the student can produce in hisanswers and in the style of answer given.In Table I it can be seen that none of thedifferences between these two groups foreither sex reached a level of statistical sig-nificance. Thus, a test measure which wasspecifically designed for measuring written

expressiveness does not seem to predictoral expressiveness in thc classroom. Onthe Consequence Solutions score ( on itemslike the "Three fingers" the student willanswer, "Cars would have new types ofsteering wheels" instead of "You couldn'tdrive cars"), no statistically significant dif-ferences were obtained, but it can be no-ticed that the trends are in opposite direc-tions for the two sexes. The most expressiveboys give more solutions but the more ex-pressive girls give less solutions to theConsequences questions.

The teachers were given the task ofrating the students along two dimensions,cognitive abilities and sociability. In TableI, the teachers significantly distinguished intheir ratings between the expressive andnonexpressive students of both sexes. Sincea low score means favorable rating, thismeant that the teachers significantly ratedhigher on cognitive abilities those studentswho were the most expressive. It can beargued that it is this very expressivenessin class that might be influencing anddirecting the teachers' judgments in thisregard.

This favorable rating on cognitive areadid not carry over into the social dimen-sion. No statistically significant differenceswere found on ratings of social successbetween the two groups. In summary, inthis general area, the expressive boys dif-fered from the nonexpressive boys on ameasure of verbal IQ and on teacher ratingsof cognitive abilities. The only differencebetween the expressive and nonexpressivegirls was found on teacher ratings of cog-nitive abilities. These differences were notas great as might be expected and led tothe further supposition that other thanstricfii cognitive abilities were detenniningthe amount of thought expressiveness inthe classroom.

Each of the students in the present studywas administered a semantic differential

1


Table IIThe Comparison of Expressive and Nonexpress:ve Gifted Students on

Semantic Differential ScaleExpressive Nonexpressive

Concept N Mean a N Mean a

Mother B 25 .84 6.18 26 74.26** 5.41

G 25 72.32 5.93 28 72.17 5.18

Father B 25 66.36 9.40 26 70.50 7.93

G 25 70.16 8.49 28 70.25 7.30

Student B 25 63.88 9.71 26 65.34 7.51

G 24 67.12 7.75 28 67.89 7.26

Work B 25 63.60 6.54 26 67.73t 8.20

G 25 66.56 7.87 23 64.92 8.06

Competition B 25 66.28 9.32 26 67.61 13.77

G 25 09.04 6.57 28 64.53 8.85

Success B 25 59.72 11.76 28 70.84" 8.07

G 24 66.83 9.20 28 67.50 8.58

Love B 25 66.76 9.81 26 70.65 9.02

G 25 09.80 7.07 .42 7.18

Faith B 25 67.72 9.14 26 72.11 9.06

G 25 76.66 8.29 28 73.17 7.00

Imagination B 25 66.00 7.31 26 62.15 14.72

G 25 .96 9.63 28 65.85 7.89

School B 25 65.72 10.40 26 71.09' 7.20

G 25 70.24 7.10 28 68.46 7.52

tSisnificant at .11. lewd of confidence *Significant at .05 knd of confidence *Significant at .01 level of confidence

scale which previously had proven success-ful in distinguishing between gifted achiev-ers and nonachievers (10). The semanticdifferential scale was developed from workby Osgood (9). Students are presentedwith a series of concepts and asked torespond to the concepts in terms of anumber of adjective pairs, such as cold-hot,good-bad, active-passive, etc. The conceptsin the present study are listed in Table II.Fourteen adjective pairs were used and theconcept score for a student was the sum ofhis ratings on the fourteen pairs for thatconcept

Table II reveals that differences werefound between expressive and nonexpres-sive gifted boys on a number of the con-cepts. All of the differences were consistentin the same direction. On the concepts ofMother, Success, School, and Work the non-expressive group were consistently higherthan the expressive group in their ratings.The expressive group could not be said tobe rating these concepts negatively, sincea neutral score would be 49. It would bemore accurate to say they are merely lesspositive in their ratings. One possible inter-pretation of these differences would be to


suggest that the expressive boys possessmore self-confidence ai,-.I more assurance inthemselves and do not have to give extremepositive ratings on these concepts.

Students who are less sure of themselvesand less inclined to independent judgment,and thus less inclined to contribute to classdiscussions, might tend to overdo theirpositive ratings on such socially acceptableconcepts (for gifted students ) as those onwhich differences were obtained.

Consistent with other results in this study,the pattern noted for gifted boys is notrepeated for gifted girls. Only one differ-ence was found and that indicated a higherlevel of positive feeling for the concept ofCompetition for the expressive girls. Onmost of the other concepts the expressivegirls were, on the average, more positiveon these concepts than the nonexpressivegirls. One suggested explanation would bethat girls who are expressive accept theconcept of Competition as a desirable fac-tor and are willing to compete on equalterms for classroom recognition with theboys. If these results are confirmed byother studies it is clear that there are dif-ferent patterns of motives and abilities lyingbehind the classroom expressiveness of boysand girls.

Smnmary

The expressive behavior of gifted chil-dren at the junior and senior high schoollevel in ten different classrooms wasstudied through analyzing the tape record-ings of five consecutive class sessions. Aspecial classification system was developedas part of the project to allow the categori-zation of each teacher and student state-ment. It was found that 'certain types ofthought operations were more commonthan others in all of the classrooms regard-less of subject matter orientation.

Expressive behavior in the classroom inboth kind and amount of thought output

seemed dependent on the teacher's style ofqpestion asking, the sex of the student, thegoals of the teacher in a given lesson, thecomposition of the class group, and thepattern of attitudinal and personality char-acteristics of the student. Consistent sexdifferences suggested that different patternsof attitudes and cognitive skills underlieexpressive behavior in gifted boys and girls.

Bibliography1. Barron, F., Creativity and Psycholugical

Health. Princeton, New Jersey: Van Nostrand,1963.

2. Bloom, B. S. (Ed.), Taxonomy of EducationalObjectives. Handbook I: Cognitive Domain.New York: Longmans, Green, 1956.

3. Flanders, N. A., "Intent, Action and Feed-back: A Preparation for Teaching," Journal ofTeacher Education, 14 (1963), 251-200.

4. Cetzels, J. W. and P. W. Jackson, Creativityand Intelligence. New York: John Wiley,1962.

5. Guilford, J. P., "The Structure of Intellect,"Psychological Bulktin, 53 (1956), 267-293.

6. Guilford, J. P., "Three Faces of Intellect,"American Psychologist, 14 (1959), 469-479.

7. Medley, D. M. and H. E. Mitzel, "MeasuringClassroom Behavior by Systematic Observa-tion," in N. Gage (Ed.) Handbook of Re-search on Teaching. Chicago: Rand McNally,1963.

8. MacKinnon, Donald W., "Fostering Creativityin Students of Engineering," Journal of En-gineering Education, 52 (1961), 129-142.

9. Osgood, C. E., "Studies on the Generality ofAffective Meaning Systems," American Psy-chologist, 17 (1962), 10-28.

10. Pierce, J. W. and P. Bowman, "MotivationPatterns of Superior High School Students,"in The Gifted Student. Washington, D.C.:U. S. Office of Education, Cooperative Re-search Monograph, 1960, 33-66.

11. Smith, B. O. and M. O. Meux, A Study of theLogic of Teaching. U. S. Office of EducationCooperative Research Project #258. Urbana,Ill.: University of Illinois, Bureau of Educa-tional Research, 1962.

12. Taba, Hilda, S. Levine, and F. Elzey, Thi k-ing in Elementary School Children. U. S.Office of Education Cooperative RftearchProject #1574. San Francisco, California: SanFrancisco State College, 1964.

13. Torrance, E. P., Education and the CreativePotential. Minneapolis, Minn.: University ofMinnesota Press, 1963.

articles are: (I) languageand the Habit of - ERIC

Documents