DOCUMENT RESUME SP 006 737 Kinesiology Review 1971 ... · ED 084 219. TITLE INSTITUTION. PUB DATE....

ED 084 219

TITLEINSTITUTION

PUB DATENOTEAVAILABLE FROM

EDRS PRICEDESCRIPTORS

ABSTRACT

DOCUMENT RESUME

SP 006 737

Kinesiology Review 1971.American Association for Health, Physical Education,and Recreation, Washington, D.C.7170p.NEA Publication-Sales, 1201 Sixteenth St., N.W.,Washington, D.C. 20036 ($3.00)

MF-$0.65 HC Not Available from EDRS.*Anatomy; *Cybernetics; Exercise (Physiology) ; HealthEducation; *Human Body; *Mechanics (rhysics) ; Motion;*Physical Education

This report contains articles on research inkinesiology, the study of the principles of mechanics and anatomy inrelation to human movement. Research on sequential timing, somatotypemethodology, and linear measurement with cinematographical analysisare presented in the first section. Studies of the hip extensormuscles, kinetic energy, and individuality are included in the secondsection. A computer program for descriptive analysis of movementsused during repeated performance of skills is explained. Motionsimulation, using a three dimensional human form, is also presented.Each article presents figures and tables of data along withreferences. (BRB)

CrN) .--1rJ

COOO1.1-1

US DE PAR TMENT OF HEALTH,EDUCATION & WELFARENATIONAL INSTITUTE OF

EDUCATIONroS DOCU,IF NT PAS firEN REPPO

DuCt D ExAc TIV AS RECI,VED ; PON.1111 PE PSON OU 014,ANitATIONOF1010AI ING rf PO,N15 Of v1 E4 OP OPINIONS',1.411-D DO NO' NI TE-V,AP,I. I-PPE',ENT 011111(4 NAtiONAL ,NSIlioTE orI Dur p ,)0 1,0',1110N OP P01 IC V

Plif/!,SION TO PUPPOOLICUCOPv1,,r,1,1FD NN,./f.;41:f. BY MICROFICHE ONLY IIAS BEEN GIIANTEI1 BY

TO f- 1(- AN OR/7GANI/AnONS

OPERATING 1101)117 AC04E-IMF-NTS WITO 111) NATIONAL INS1 11U1C OF I. DUCATLONI 1114 1111 4 Al PPODUC 1 ION OU1500111,1 f pft M. wujoHei S Pt 10:1!,',ION 01 IIll COPY{,0(.1..T CANER

kinesiologyreview1971Published by the Council on Kinesiology

of the Physical Education Division

AMERICAN ASSOCIATION FOR HEALTH,

PHYSICAL EDUCATION, AND RECREATION

1201 Sixteenth St., N.W., Washington, D.C. 20036

FILMED FROM BEST AVAILABLE COPY

Copyright C) 1971

American Association for Health, PhysicalEducation, and Recreation

A National Affiliate of the National Educaticn Association1201 Sixteenth St., N.W., Washington, D.C. 20036

Order from: NEA Publications-Sales, 1201 16th St., N.W.Washington, D.C. 200.36

Contents

Sequential Timing of Three Overhand Patterns, Marlene J. Adrian 1

Somatotype Methodology and Kinesiology Research, J. E. Lindsay Carter andBarbara H. Heath 10

A Problem Solving Approach to the Study of Muscle Action, Ernest IV. Degutis 20

Errors in Linear Measurement with Cinematographical Analysis, T. L. Doolittle 32

A Differential Study of the Flip Extensor Muscles, M. Marilyn Flint 39

Kinetic Energy: A Measure of Movement Individuality, Gladys E. Garrett andCarol J. Widule 49

ANGLCALC: A Computer Program for Descriptive Analysis of Movements UsedDuring Repeated Performance of Skills, Shirt James Hoffman andKenneth Worsham 55

A Three-Dimensional Human Form and Motion Simulation,Walter S. Reed and Richard E. Garrett 60

iii

Council on Kinesiology1970-71 Officers

Chairman:

JOAN C. WATERLAND, Department of Physical Education for WomenUniversity of Wisconsin, Madison, Wisconsin

Chairman-Elect:

GLADYS E. GARRETT, Department of Physical Education for WomenPurdue University, W. Lafayette, Indiana

Past Chairman:

BENJAMIN RICCI, Laboratory of Applied PhysiologyUniversity of Massachusetts, Amherst, Massachusetts

Secretary:

WAYNE E. SINNING, Springfield College, Springfield, Massachusetts

iv

2Kinesiology ReviewBoard of Editors

Editor:

CAROL J. \VIDULE, Department of Physical Education for WomenPurdue University, W. Lafayette, Indiana

A,vsociate Editors:

LOUIS E. AL. L Y, Department of Physical Education for MenThe University of Iowa, Iowa City, Iowa

J. E. LINDSAY CARTER, Department of Physical EducationSan Diego State College, San Diego, California

DAVID H. CLARKE, Department of Physical EducationUniversity of Maryland, College Park, Maryland

H. HARRISON CLARKE, School of Health, Physical Education and RccrcationUniversity of Oregon, Eugene, Oregon

JOHN M. COOPER, School of Health, Physical Education and RecreationIndiana University, Bloomington, Indiana

BRYANT J. CRATTY, Department of Physical EducationUniversity of California, Los Angeles, California

RICHARD E. GARRETT, Department of Mechanical EngineeringPurdue University, W. Lafayette, Indiana

KATHRYN LUTTGENS, Department of Physical Education for WomenBoston-Bouv6 College of Northeastern University, Boston, Massachusetts

DALE 0. NELSON, Department of Health, Physical Education and RccrcationUtah State University, Logan, Utah

GEORGE 0. RICH, Human Performance LaboratorySan Fernando Valley State College, Northridge, California

JOSEPH ROYCE, Department of Physical EducationUniversity of California, Berkeley, California

LEON E. SMITH, Department of Physical Education for McnThe University of Iowa, Iowa City. Iowa

WAYNE D. VAN HUSS, Department of Health, Physical Education and RecreationMichigan Statc University, East Lansing, Michigan

Headquarters Stall:

WILLIAM L. COOPER, Director of Publications

JEAN WILSON, Research Editor

MARLENE J. ADRIANMARY LOU ENBERG

Sequential Timing of Three Overhand Patterns

ACCORDING TO SEVERAL AUTHORS in theareas of mechanics and kincsiology Broer(1). 13roer and Houtz (2). Cooper andGlassow (3). and Wells (6) there is sim-ilarity in the pattern of overhand move-ments. 13rocr states that throwing, a tennisserve, an overhead clear in badminton, andthe smash in tennis and badminton are thesame movement pattern in terms of me-chanics. Indeed, the model in the photo-graphic sequence in the Brocr work doesexhibit the same kind of pattern for the de-picted overhand skills.

One of the investigators in the currentstudy was interested in the mechanics ofperformance; the other was involved fromthe learning aspect. since it is a commonpractice in the teaching of new skills to relatethem to patterns previously learned. Withthe emphasis in the literature on similarity ofpatterns and the a:;sumed transfer of learn-ing from one skill to another, it was pro-jected that a comparison of several overhandpatterns might indicate intra- and inter-indi-vidual similarities and differences. One of theinvestigators. Enbcrg (4). noted in a limiteddescriptive study that interindividual differ-ences might be important considerations in

Marlene J. Adrian and Mary Lou Enbergare members of the faculty of WashingtonState University, Pullman, Washington.The authors extend their gratitude to sub-jects Donna Chun, Cathy Burquist, andCarolyn Jensen Hein, to Barbara Fechtwho contributed the drawings, and to Her-bert Howard, research photographer atWashington State University.

KINESIOLOGY REVIEW 1971

the choice of descriptions for introducingnew movement skills. Mechanically, since themusculoskeletal system limits the types ofmovement possible at each joint, similaritiesin patterning may result primarily fromstructure of the body rather than function fora given sport skill. More subtle differencesmay occur which are specific to the partic-ular sport skill. Furthermore, this structuremay be a reinforcer of one particular sportskill and serve to inhibit successful function-ing for another sport skill.

The purpose of the present study was tomake a comparison of intra-individual andinterindividual performances on three over-hand sports skills: the volleyball serve, thebadminton smash, and the tennis serve. Theportion reported here is part of a largerstudy now in progress.

ProcedureThe subjects were three highly skilled col-

lege women: D.C. had competed on theintercollegiate level in all three sports, C.B.was a member of the intercollegiate teams intennis and volleyball and an intermediatelevel player in badminton, and C.H. was abadminton player of international caliber, amember of the varsity volleyball team, anda novice at tennis.

The subjects were clad in leotard andtights, marked with standard tape markingsat the joints, and taped with two spines, oneat the upper spinal level, and the other atthe sacral level.

Subjects were filmed on the same day,with a HyCam camera, front a distance of70 feet, and at speeds of 730 to 775 framesper second. A timing device which operated

1

at a constant 400 rpm was also included inthe filming to aid in the identification of theframes. Each skill was repeated severaltimes to insure that the subject was ac-quainted with the filming procedure andcould achieve consistency in performance.The skill which each performer judged tohe her best, i.e., badminton smash, volley-ball or tennis serve, was filmed twice toallow for a comparison and to insure thatthe skill analyzed was not an isolated eventbut represented a typical performance. Thefilms were analyzed while projected by aRecordak MPE-1 film reader. Tracings weremade for the purpose of comparing thetime involved and the sequential positioningin the various skills. It was felt that therewould be disadvantages from the use of onlyone camera, but the advantages of the highspeed of filming could provide an importantaddition to the literature in the descriptionof the sequences of movement.

FindingsFrom Figure 1, (a five-tracing sequence

of subject D.C. for each of three skills), adescriptive comparison may be made ofsegmental positioning from the moment ofcontact, backward in time at approximately.05 second intervals to .189 seconds beforecontact. In Figure 1A, at .189 seconds be-fore contact, differences will he noted in thethree stances, the amount of knee flexion,truhk inclination, and resultant angles ofshoulder inclination. The differences instance evolved from previous foot actionwhich was: in the badminton smash, a de-finite step forward with the left foot duringthe stroke; in the volleyball serve, a plant-ing of the left foot and continucu foravardmovement of the unweighted right foot; andin the tennis serve, no step, but rather adorsiflexion followed by planter flexion ofboth ankles. The trunk inclination in thevolleyball and tennis serves appeared to oc-cur as a result of the action of both kneesand, to a lesser extent, both ankles. How-ever in the badminton smash there were de-finite trunk positioning movements up to thispoint in the execution of the skill.

2

Figure 1 B, at .136 seconds before con-tact, indicates the extent of shoulder, elbow,and wrist positioning. The body position forbadmintel and tennis was more upright inspace; the volleyball position was lower be-cause of prominent knee flexion. The shoul-der inclinations for the tennis and hadmin.ton skills were similar. However the twoskills that showed similar striking arm posi-tion were the badminton and volleyball. Inthe tennis stroke there appeared to begreater humeral outward rotation, resultingin a drop of the forearm and racket. Thesimilarity of the tennis and badminton stancewill also be noted. At this point in the se-quences hip rotation for only the tennisstroke had progressed far enough into thebackswing to remove the sacral spine fromview.

Figure 1C, at .0S3 seconds before con-tact, indicates the extent to which knee andankle action had altered the height of thebody in space. Using the rectangular co-ordinate system, one would label the posi-tions as a negative trunk inclination in thevolleyball skill but a positive one for boththe badminton and tennis. A noticeablefeature was the elbow lead in badminton andtennis which was not yet present in the vol-leyball skill, the one with the shortest leverarm. Hip rotation had progressed evenly inthe three skills, but the upper spinal rota-tion appeared to be less pronounced in vol-leyball than in the other two sport skills.

Figure ID, at .034 seconds before contact,indicates a strong similarity in striking andbalancing arm positions but also a differencein the angle of the body, and hence in theangle of projection of force. An analogy maybe made between the position of the racketas a lever in the two implement sports andthe position of the forearm as a lever in thevolleyball skill. At this time the badmintonracket had a greater distance to travel beforecontact than did the tennis racket. Theremay he implications here for timing of thesequence in relation to weight, length, orabsence of a siriking implement.

Figure 1E, at the moment of contact, in-dicates a similarity in badminton and tennisbody positions but a difference in the angleof and point of contact with the projectile.


A

BAD.

VB.

TEN.

B C D E

Figure 1. A film-tracing sequence of the badminton smash and the volleyball and tennis serves forsubject D.C. A, .189 sec before contact (bc); B, .136 sec bc; C, .083 sec bc; D, .034 sec bc;E, contact.

KINESIOLOGY REVIEW 1971 3

BAD. (D C)

wristflexion

backward move.right shoulder h.o.r. I backward move. - elbow forward move.

forward move shoulder shoulder 31bow

P. left shoulder extension

upper spinal rotation

hip rotation torward

knee extension

left stepI 1

forwardi ii I ill i III I

.375 .275 .175

TIME /N .025 SECOND

D75 C

Figure 2. Segmental movements for the badminton smash from .375 seconds until contact forsubject D.C. P, pause in movement; h.o.r., humeral outward rotation; c, contact.

VB. fa c.)

backward move.right elbow 3

1 extension

backward move.right shoulder

forwardi h.o.r. move.

pause left shoulder extension


hip rotation forward

knee flexion pause knee extension

right stepforward

.375 275 .175

TIME IN .025 SECOND

.075 C

Figure 3. Segmental movements for the volleyball serve from .375 seconds until contact for subjectD.C. H.o.r., humeral outward rotation; c, contact; ( V ) , approximate time.

4 KINESIOLOGY REVIEW 1971

TEN. (DC)

wristflexion

backwardrt. elbow

move. I i elbow flexionIP forearm pronation extension

backward-mowrt. shoulder h.o.r. forward move.

left shoulder extension


hip rotationforward

hip rotationflexion

I '

kneeI I

flexion PI t I 1

,knee extension

i iI

375 275 .175

TIME IN .025 SECOND

.075 C

Figure 4. Segmental movements for the tennis serve from .375 seconds until contact for subjectD.C. P, pause; h.o.r., humeral outward rotation; c, contact; ( ), approximate time.

The volleyball position was higher (no trunkflexion as in badminton and tennis) andmore erect, and the point of contact wasmuch closer to the gravitational line of thebody than in the other two skills. Again,there may be implications about the influ-ence of the purpose of the skill andstate of learning of the performer, as well asthe influence of the striking implement, onsegmental timing. The movement patterns ofD. C., and particularly her body position atcontact, were individualized for each of theskills.

In Figures 2. 3, and 4, the time sequencesfor the segmental movements are graphedfor each of the three skills executed by D.C.These movements include only the force-producing phase of the skills. The time isfrom -.375 seconds until contact and rep-resents an analysis often blurred in 64 fpsfilming and indiscernible to the human eye.In all three skills the - .375 - second point intime is one in which the nonstriking arm ap-proached the termination of the pause be-tween flexing and extending.

Figure 2, the badminton time sequence,indicated that the establishment of the base


of support (left foot forward) occurred justprior to pelvic rotation, typical of classicalsummation of force patterns. A combinationof striking arm movement (right elbow andhumeral outward rotation) was detected at.227 seconds, or about .009 seconds be-fore completion of the establishment of thebase of support at 2.18 seconds. Bothmovements were followed closely by a hiprotation, detected at almost .210 secondsbefore contact. Upper spinal rotation beganshortly thereafter at .185 seconds and wasfollowed by knee extension at .163 sec-onds. The onset of the final striking motionof arm and forearm at .085 seconds andwrist at .033 seconds were indicative ofthe velocity these segments must have at-tained prior to contact. It will be noted thatthe single camera led to some confoundingof movements, but the number of framesper second still allowed a much finer anal-ysis than has been previously reported.

Figure 3, the volleyball time sequence,indicated that the base of support (right footforward) was established almost simultane-ously with the beginning of the left arm ac-tion at .300 seconds. Again, there were

5

'FABLE 1. SEGMENT AND PROJECTILE VELOCITIES FOR SUBJECT Dr.

Activity Elbow Extension Wrist Flexion Projectile Velocity

V13 Serve 30151sec. 45.5 ft/see.BAD Smash 16197scc. 3748^ /sec. 183.6 ft /sec.

TEN Serve 1371 ''./see. 2743 ysee. 128.9 ft/sec.

positioning movements of the right, or strik-ing elbow, but the hip rotation forward wasthe main detectable unwinding, movementand occurred at .217 seconds, very closeto the same action in badminton. A differ-ence in sequence was observed in compari-son with the badminton skill. Here the kneeextension at .150 seconds preceded theupper spinal rotation at .115 seconds.Therefore. the time span of the upper spinalrotation was less than that in badminton.The final forward movement was a com-bination of elbow and shoulder extensionand humeral inward rotation beginning atabout .065. The latter movements \ve:..confounded, but the last movement was el-bow extension. No wrist flexion was de-tected for this subject on the volleyball serve.

Figure 4. the tennis time sequence, in-dicates that there was no forward step in-volved in performance of this skill. Theskill is otherwise similar to that of badmin-ton in terms of order of segmental move-ment: hip rotation at .190 seconds, upperspinal rotation at .170 seconds, and kneeextension at .123 seconds. However, allof these movements occurred later in thetotal sequence, as if the tennis required afaster explosion. Hip flexion was detected at.095 seconds and accompanied the con-tinued hip rotation. Again, as in badminton,the final striking motion occurred with theelbow extension at .080 seconds and wristflexion at .043 seconds.

Angular velocities of the segments forfive frames (average of .006 seconds) priorto contact and projectile velocities were cal-culated for subject D.C. and are given inTable 1.

6

Although rotation occurred during variousportions of the skills, there appeared to beno forearm rotation during the final .006second, as determined by length of segment.Apparently the body, including the shoulder,was stabilized, and the movements occurringwere elbow and/or wrist actions. Thereforethe movements were basically in the a n tero-posterior plane and sufficiently parallel tothe camera that measurements were accurateenough to report. It is assumed that anyinaccuracies of reporting would be an under-est:iii.ite rather than an overestimate of thevelricities, since some translation did occur.

11 is evident that the badminton smashand tennis serve represent high velocity pro-jectiles, while the volleyball serve velocitywas above average but not much above theminimum of 40 ft 'sec. reported by Temple(5). However, since the elbow extensionvelocity was over 3000- 'sec., the lower pro-jectile velocity might be attributed to thecoefficient of restitution of the volleyball,the short lever arm, or the performer's eon-cern with accuracy and spin.

Since the findings reported here an., onlya portion of a larger study. the films for theother two subjects have not been analyzed inas great detail. However, some interestingcomparisons arise in the positioning of thesubjects at contact and at .034 seconds priorto contact. The tracings in Figures 5 and 6are the equivalent of the last two trningsin Figure 1, i.e., drawings 113 and 1E. Figure5A, at .034 seconds before contact, indicatesthat subject C.I3. had a narrow stance whichwas similar for all three skills. The amountof knee flexion was minimal, but hip flexionwas pronounced all three skills. Again,


A

A

A

BAD.

VB.

TEN.

B

B

A

A

BAD.

VB.

TEN.

A

B

B

B

Figure 5. A filmtracing sequence of the badmin Figure 6. A filmtracing sequence of the bakminton smash and the volleyball and tennis serves tion smash and the volleyball and tennis servesfor subject C.B. A, .034 sec before contact; for subject C.H. A, .034 SEC before contact;B, contact. B, contact.


the racket and forearm analogy suggested forD.C. applies, but the elbow flexion was morepronounced in the volleyball serve than forD.C. at this point in time. The kind of tennisserve (slice) probably accounted for thc dif-ference in the upper spinal rotation andhead position in the tennis serve as com-pared with the other two skills. The rightand left arm co-action was similar on allthree skills.

Figure 5B indicates the simikrity in thepoint of contact for all three skills with ref-erence to the vertical line of the body. Thesepoints of contact differed, in terms of dis-tance from the gravital line, from those ofD.C. A somewhat exaggerated hip flexionat contact on all three skills was utilized byC.B., indicating that the trunk movementwas an important part of her moment arm.C.B. still had a flexed elbow position, yield-ing a lower point of contact and a more up-ward flight of the projectile than for D.C.The trunk position for C.B. at contact inthe tennis serve differed from that of theother two subjects and was necessitated bythe kind of serve.

In Figure 6A, at .034 seconds before con-tact, C.H. exhibited less intra-individual dif-ferentiation that did either of the other twosubjects. That finding might carry the impli-cation that, as one skill reaches an extremelyhigh level, similar movement patterns mightactually replicate the one best skill. In thetracings, the stance, knee and hip flexion,amount of Np and spinal rotation, left andriOt arm action, and head position were,,cry similar for all skills. It will be notedthat the badminton racket was angled moresharply downward than the tennis racket.The major difference among the skills oc-curred in the amount of elbow and wristflexion, with the volleyball elbow still flexedand the badminton wrist still cocked. Thelatter contributed to the velocity of the strik-ing implement recorded later in the paper.

Figure 6B, point of contact, indicates dif-ferences in width of base of support, but theoverall body positions were similar. Thevolleyball and badminton points of contactwere closer to the gravital line of the bodythan that for badminton. This subject, like

8

C.B., used hip flexion at contact on all threeskills.

The velocity calculations for D.C. indi-cated a direct relationship between lever andprojectile velocity. Therefore only the veloc-ity of the final contributing lever was cal-culated for the other two subjects in thisportion of. the study. Bceause other factors,e.g., length of lever or coefficient of restitu-tion of the striking surface in volleyball,might have affected the projectile velocity,calculations and subsequent comparisons ofthese velocities were not made. Wrist flexionwas calculated for C.B. at 3432°/ sec. inbadminton and 39777sec. in tennis. Thevolleyball elbow extension was 3165 ° /see.The greatest velocity was achieved in hermost proficient sport despite the fact that,theoretically, wrist flexion velocities with abadminton racket are faster than with a ten-nis racket. C.H. had slightly greater elbowextension velocity in volleyball, 3200 ° /sec.The slow wrist action of 2500°/sec. in ten-nis was indicative of C.H.'s novice status.However, the wrist flexion of this highlyskilled badminton player was 6200°/see.,a figure which exceeds any published data.Cooper and Glassow reported wrist velocityfor a highly skilled woman thrower at52507see. on an overhand throw, a skillwithout a striking implement.

Discussion and conclusionsThe data presented indicate that t1re

are some intra-individual and interindividualdifferences in the patterning of the threeoverhand skills discussed: the badmintonsmash, volleyball serve, and tennis serve.The differences in bodily movements mayimply differences in neuromuscular pattern-ing. An interesting question to pursue fromthe mechanics- learning aspect would bewhether the neuromuscular system recordspattern parts (loops or subroutines) whichmay be assembled either in a given order orwith certain substitutions allowed in thechain, leading to different neuromuscularpatterns.

Two overhand patterns are noted in theliterature, and variations thereof were noted


in these data. One pattern involves a stepforward followed by pelvic rotation and theclassic unwinding sequences for force pro-duction. Another pattern introduces flexionin the anteroposterior plane, as a hip-trunkaction and/or a knee-trunk action. Anotherquestion arising from these data would in-volve an investigation of the mechanical ef-fectiveness of the two patterns. That is, isone pattern more effective for an individual,or is it more effective for a specific over-hand pattern? And, more generally, will theanteroposterior pattern find more acceptancein the teaching of certain overhand skills?

The subject with a high degree of skill inall sports (D.C.) showed greater intra-individual differences in the contact positionthan did the other two subjects. For thesubject who was very highly skilled in onesport in particular (CIL), the position atcontact in the other skills greatly resembledthe contact position in the best sport. Theslice serve in tennis for C.B. was responsiblefor some of the interindividual differencesnoted in that movement pattern.

On the basis of the data presented, itseems possible that the differences in per-formance of similar overt-imd patterns mightbe as important as the similarities, especiallyat the highly skilled level. In general, then,there are implications for transfer of learn-ing as well as for mechanical improvementof performance. The results of the study in-dicate the need to pursue some of the ques-tions evolving from the findings.

References1. Broer. Marion R. Efficiency of human inure-

ment. (2d ed.) Philadelphia: W. B. SaundersCo., 1966.

, and Houtz, Sara Jane. Pattern %,;mus-cular activity in selected sport skills. Spring-field. III.: Charles C. Thomas, 1967.

3. Cooper, John M.. and Glassow, Ruth B. Kine-siology. (2d ed.) Saint Louis: C. V. Mosby Co.,1968.

4. Enberg. Mary Lou. A comparison of two over-hand patterns. Unpublished report, PurdueUniversity, 1966.

5. Temple, Ina. Developing skill in the overarmserve. In Betty Jane Wills (Ed.). Volleyballguide. Washington: American Association forHealth, Physical Education, and Recreation,1965, pp. 23-31.

6. Wells. Katharine F. Kinesiology. (4th ed.)Philadelphia; W, ft. Saunders, 1966,


J. E. LINDSAY CARTERBARBARA H. HEATH

Somatotype Methodology and Kinesiology Research

IT IS NOW THREE DECADES since Sheldon,Stevens, and Tucker (40) introduced theconcept of somatotyping. This basic conceptwas unique and has proved useful in manyfields of study during the past 30 years. Inresponse to the -need to overcome some in-adequacies in the originally proposed system,a number of modifications to somatotypinghave been developed. However, an examina-tion of articles using the technique of soma-twyping, and books in various disciplinesdescribing the technique, reveals two prob-lems: a) the authors in general seem to besingularly unaware of the developments andmodifications in somatotype methodology;and b) inappropriate somatotype methodshave been used for examining relationshipsbetween somatotype and other variables suchas growth and motor performance.

The purpose of this paper is to describethe developments in somatotype method-ology and to point out their relevance tokinesiological studies.

METHODS OF SOMATCIYPING

Sheldon's somatotype methodsPrior to 1940 the physique of an individ-

ual was usually characterized by individualor combined anthropometric measures or by

J. E. Lindsay Carter is professor of physi-cal education at San Diego State College,San Diego, California. Barbara H. Heathis a consulting physical anthropologist inMonterey, California.

10

grouping people into broad discrete cate-gories or types on the basis of measures orvisual impressions. There was a need for aclassification of total body form on continu-ous scales which could be expressed in asimple value. A technique directed towardsthis end was described by Sheldon and others(38, 40), who called it "somatotyping," aterm they defined as follows: "... a quanti-fication of the three primary componentsdetermining the morphological structure ofan individual expressed as a series of threenumerals, the first referring to endomorphy,the second to mesomorphy, and the third toectornorphy." Sheldon wrote that endomor-

,y, or the first component of the morpho-i , ical level of the personality, is character-ize, by relative predominance in the bodilyeconomy of structures associated with diges-tion and assimilation; mesomorphy, or thesecond component at the morphologicallevel, by relative predominance of the meso-dermally derived tissues, which are chieflybone, muscle, and connective tissue; andectomorphy, or the third component at themorphological level, by relative predomi-nance of ectodermally derived tissues, whicharc chiefly skin and its appendages, includ-ing the nervous system. The somatotype,Sheldon holds, is a trajectory or pathwayalong which the individual is destined totravel under average conditions of nutritionand in the absence of major illness (38,p. 337).

Sheldon also used the terms morphophe-notype, or the present body type, and mor-phogenotyp2, or the genetically determinedbody type. Nevertheless, Sheldon maintainedthat a given individual's somatotype is un-


changeable, even though it can be assessedthrough inspection of a sequence of pheno-types through time. Sheldon's system intro-duced a seven-point rating scale for scoringthe expression of each of the three compo-nents of a somatotypc. The sum of theratings on the three components is between9 and 12. The system also contained tablesof age corrections (at intervals of five years)for somatotype distributions ranging from18 to 63 ycars, based on the height-weightratio.*

In its earliest form the somatotypc wasdetermined from 17 measurements taken ona negative or photograph, but this laboriousprocess was seldom used after its replace-ment with the photoscopic procedure. Toassign a somatotype rating, Sheldon usedsomatotypc photographs taken from thefront, hack, and side; a record of weights atvarious ages, where available; a table of thesomatotype distribution according to theheight-weight ratio and age of the subject;and standardized photographs and descrip-tions of the somatotypes against which tomatch each subject's photograph. Althoughthe method appears simple enough, thereare many subtleties, and satisfactory ratingscannot be achieved without considerablepractice and training.

This new method of evaluating bodyform had its critics and the fact that it sur-vived the somewhat acerbic but masterlycritique of Howard V. Meredith (24) prob-ably indicated the viability of the conceptand at least laid the groundwork for futuredevelopments.

Sheldon (37, 39) recently summarized themain objections to his system: a) the soma-totype changes; b) soniatotyping is not ob-jective; c) there are only two, not three,primary components; and d) somatotypingomits the factor of size. He offers comments

* The height-weight ratio referred to is theheight/cube root of weight (HWR). Strictlyspeaking, there are many height-weight ratio,,;one of these is the ponderal index which is thecube root of weight/height, and it is the re-ciprocal of this form, called the reciprocalponderal index (RPI), which is used in soma-totyping.


on these criticisms and presents sonic in-formation on further objectification of hismethod. Sheldon's "new" method, apparentlyin use since 1961 (36), is still inadequatelydescribed, and the associated tables havejust been published (39). The essential pro-cedures in the new method arc as follows:a) standardized photograph and weightrecord arc taken as before; I)) maximalstature and minimal height-weight ratios areestablished from height and weight histories;c) a trunk index is derived from planimctriemeasurements of the thoracic and abdominaltrunks as marked on the photographs; d)the somatotype is obtained from a table ofheight-weight ratios and trunk indices, thena second table of somatotypes plotted againstmaximal stature, and finally from the "basictables" for soniatotyping which combine thethree parameters. These tables arc age-corrected, and they arc read differently formen and women, and the seven-point ratingscale is retained. The sum of the three com-ponents is not limited to a range of from 9to 12 as in the previous methods, but nowranges between 7 and 15. Whether or notthese changes meet the objections to Shel-don's earlier system is quite debatable, as nosupporting evidence is given by Sheldon.However, it is important to notc that thetrunk index system bears little relationshipto his previous systems (38, 40) so must beregarded as another somatotype method.Furthermore, because maximal stature andminimal height-weight ratios are requiredfor the trunk index system, its practical ap-plication in growth studies is questionable.To obtain this information would requiretesting annually until maximal stature isreached.

Hooton's somatotype methodHooton's method (18) is essentially a

phenotypic representation of the somatotype.He and his colleagues used this method ex-tensively throughout the 1940's on UnitedStates Army men, and he based his ratingson inspection of the photograph and theheight-weight ratio. When compared to Shel-don's 1940 method, the first component

11

ratings in the lower grades were more "lib-eral," while the second component was ratedmore strictly, resulting in scores approxi-mately one unit lower. The third componentwas derived directly from scaled height-weight ratios. The height-weight ratios ofapproximately 40,000 army men were di-vided into seven equal categories and thethird component rating units were assigneddirectly. Hooton's system did not limit thesum of the three components to 9 through12. Another departure from the Sheldoniansystem was the use of the terms fat, mus-cularity. and attenuation for the three com-ponent names (12, 18, 28).

Cureton's somatotype method (8. 9)Cureton's system combines inspectional

ratings of the photograph, palpation of themusculature, skinfold measurements, height-weight ratios, and assessments of strengthand vital capacity. His studies were mainlyconcerned with young college men and ath-letes.. Curcton stated his system ". . . willplace any given case at the location in theSheldon-Stevens-Tucker triangle at leastclosely enough for all practical purposes byobjective measurements." (9, p. 15). How-ever, his ratings differ from Sheldonian ra-tings in the third component. Cureton's ra-t:ngs average close to one unit higher thanwould be possible by using Sheldon's height-weight ratio tables. In some of Curcton'ssamples of athletes the difference in ratingsis even more marked; for example, the meanheight-weight ratio for Danish gymnasts isgiven as 12 43, while the mean somatotyperating is 21/2-51/2-4 (9, p. 26). Accordingto Sheldon's distribution of somatotypes byheight-weight ratio (38, p. 267), the 3-5-4somatotype has a height-weight ratio of13,10. A height-weight ratio of 12.40 is op-tiinum for a 3-6-1 somatotype. In regard toCurcton's inclusion of performance scoresfor assessing somatotype, such criteria arearbitrary and may lead to spurious relation-ships, because the component ratings arenot independent of performance. Anotherdifference in Cureton's use of somatotypingis that he constructs his somatotype triangle

12

with ectomorphy on the left side ;:eld en-domorphy on the right side of the trianglea procedure which is opposite to that of allother authors.

Parnell's MA deviation chart methodParnell (29, 31) suggested and developed

the use of the following anthropometricmeasurements in conjunction with somato-type photographs to obtain component ra-tings: bone diameters, muscle girths, andskinfolds. These data wcrc entered on theM.4 deviation chart, which included all nec-essary tables, The anthropometric somato-type thus obtained corresponded with Shel-don's estimate of somatotype, but Parnellsubstituted the terms fat, muscularity, andlinearity (along with their respective abbre-viations F, M, L) for the component names.The endomorphic estimate is derived fromthe skinfold measurements, the mesomor-phic estimate from height, bone diameters,and limb girths, and the ectomorphic esti-mate from the height-weight ratio. Each ofthe component scales is corrected for differ-ent age groups. Although the chart is basedon male data, it is also used for femaleratings. Parnell developed similar M.4 chartsbased on anthropomctric measures for chil-dren aged 7 and 11 (30, 31, 32).

Damon's anthropometric methodDamon, and others (10) predicted soma-

totypc from anthropometric measurementson white and Negro soldiers, using a multi-ple regression technique. Forty-nine anthro-pometric measurements, including weight,lengths, breadths, depths, circumferences,skinfolds, grip strengths, and pulmonaryfunction were used. Eighty percent of thepredictions came within half a rating uniton a seven-point scale of the photographicratings made by Damon (an experiencedSheldonian rater). Multiple correlation co-efficients (R) for endomorphy, mesomor-phy, and ectomorphy were .78, .66, .90 forwhites, .83, .84, and .88 for Negroes. Upto 10 different measures were used to predictsome of the components and the other com-ponent predictions were also used in some


of the equations. The grip strength scoresand pulmonary function were not used inany of the equations.

Petersen's somatotype method for childrenOut of a group of approximately 12,000

children, every child showing an illustrativeor striking build was selected to be somato-typed. The children were mostly Dutch, withsome Belgian children, undergoing schoolmedical examinations. Thcy ranged in agefrom approximately 5 to 14 years, with themajority prepubescent. The body build ofthese children was evaluated in accordancewith what was actually seen in the photo-graphs (a phenotype), without taking intoconsideration possible expectations with re-gard to future development. The somato-scopic criteria were those for Sheldon'syoung adult males, and the ratings weremade by Petersen and Van Galen. The datapresented is all cross-sectional and is thebest published series of photographs on chil-dren's somatotypes. However, the authorbases his general theme and orientation onSheldon's concepts and believes in the con-stancy of physique, although he really avoidsexamining the issue in detail (33).

Medford somatotype equationsEquations for predicting somatotype com-

ponents for boys 9 through 17 years of agehave been examined as part of the MedfordGrowth Study conducted by the Universityof Oregon under the direction of H. HarrisonClarke. Sinclair (41, 42) and Munroe (26,27) derived regression equations from an-thropometric and performance measures forpredicting somatotype components as ratedby Heath (see below). A variety of equa-tions are presented with the multiple regres-sion correlations for the first and third com-ponent usually quite high, but with lowerrelationships for the second component.The prediction for the second componentwas improved whenever values for the othercomponents were used in the regressionequation. It was found that the regressionequations were specific to the age at whichthe measurements were taken.


The Heath-Carter somatotype method

Heath (15) criticized certain limitationsof Sheldon's method and described modifica-tions designed to overcome these limitations.To begin with, the author stated that theseven-point scale was arbitrary and thenproposed a rating scale, with equal appear-ing intervals, beginning theoretically withzero (but in practice at one-half) and hav-ing no arbitrary upper end point. Secondly,she saw no good reason for limiting the sumof the components ranging from 9 to 12, sothis restriction was eliminated. Thirdly, sheobserved that in Sheldon's table of somato-types and height-weight ratios there was alogical linear relationship between someratios and somatotypes, but at other pointsthe distribution seemed quite arbitrary.Heath then reconstructed the table to pre-serve a linear relationship throughout.Fourthly, Heath seriously, questioned the"permanence" of the somatotype; therefore,she eliminated extrapolations for age andused the same height-weight ratio table forboth sexes and all ages.

Heath's somatotype ratings were presentsomatotypes, or morphophenotypes, andwere neither predictions of future somato-types nor estimates of the somatotype atage 18, as were the Sheldonian somatotypes.The elimination of extrapolations for agestrengthened somatotype methodology inseveral ways. The concept of a series ofsomatotypes for each individual replacedthat of one somatotype for a lifetime. Ingrowth studies it was possible to study theevolution of adult somatotypes from a suc-cession of pre-adult somatotypcs. Further-more, subjects in a reference population, butwith different ages, could be compared di-rectly, as they were measured on the samemeasuring scale. In the past 12 years Heathhas applied these modifications to somato-type data involving approximately 15,000ratings used in over 30 published studies.

Heath and Carter (17) further objecti-fied Heath's system by incorporating Par-nell's M.4 technique. Heath and Carter de-fine the somatotype as "present morpho-logical conformation." This somatotype isexpressed by a three-numeral rating of the

13

671

661

761

571

S.

181

281 182

MES011ORPHY381

431 172

371 2 2

361

461

651

751 541

641

741 642

631

731 632

621

72

8210 7722

372261

362

352

262

313252

162

263

451 253

CO

173

163

153

164

154

S 452 353 254 155

552 e 4111 3t.?144'

441..**,,,, '''

442 00is,542 w 443

543432

532

533

33

B2/ 534

523

623

612 51.3

613

713 614

524

514

343

4 .1.3

4 4

4 4

4 4

515

244

344

31g

334

335

325

425

315

415

416

245

135

235

236

3356a 226

326

216

316

317

146

136

126

215""'

137

,v**118o

218

44.

Figure 1. Somatochart showing distribution of the somatotypes of 35 San Diego State College footballlettermen. Somatotype ratings were made using Parnell's M.4 Deviation Chart method. The meansomatotype (triangle) is 4,7-5,5-2.1.

three primary components of physique whichidentify individual features of body morphol-ogy and body composition. The first, or en-domorphic component, refers to relativemusculo-skeletal development for the in-dividual's height. (It may be thought of asrelative lean body mass). The third, or ecto-morpitic component, represents relative lin-earity of individual physiques. Its ratings,derived largely but not entirely from height-weight ratios, evaluate body form and longi-tudinal distribution, or "stretched-outness"

14

of the first and second components. Ex-tremes in each component are found at bothends of the scales. That is, low first com-ponent ratings signify physiques with littlenonessential fat, while high ratings signifyhigh degrees of nonessential fat. Low sec-ond component ratings signify light skeletalframes and little muscle relief, while highratings signify marked musculo-skeletal de-velopment. Low third component ratingssignify great mass for a given height and lowheight-weight ratios, while high ratings sig-


781

...881 671 dr

561

661

481

181

281 182

MESOMORPHY

381

2 172

581 371

681 .m.411/ 471 261

11114

le34 44.

431. 352

312

671

272

262

313252

+

861 651453 .342

4 254 155761 551 452

4 4 I

%. ....

244751 541 442 343

........344

o,'851 641 542 443 245

N. ....444,.........543

741 642 432 3 3..",,,,841 631 532 j33

173

162

163

263

IDG 153

253

164

154

.....552 144

831

731 632

621

72

I,821

I722

%. 711O

%.*'..%:)111 71

.4.

X33

533

523

534

623 524

612 51,3

613

713 614

514

145

116

135

235 136

4 4 s..%.......%335 236 137."%.,.

33iEt 126-.......

325 226

`,N.

4 4

425

315

326

216

227

1

4.128

)71

49

4f

415 316 2. 7 c:,. 118,4.+

416 317 2165 5

Figure 2. Somatochart showing distribution of the somatotypes of 35 San Diego State College footballlettermen. Somatotype ratings were made using the HeathCarter Anthropometric somatotypemethod. The mean somatotype (triangle) is 4.2-6.3.-1.4.

nify linearity of body segments and of thebody as a whole (little mass for a givenheight), together with high height-weightratios.

For the assessment of somatotypes bymeans of the Heath-Carter method, the fol-lowing data arc needed: height, weight, fourskinfolds (triceps, subscapular, suprailiac,calf), two bone diameters (humerus, fe-mur), two muscle girths (flexed arm, call),information on age, and a revised height-weight ratio table. By entering data on the


Heath-Carter somatotype rating form theanthropometric somatotype is obtained di-rectly. Details of this procedure are fullyexplained elsewhere (17). The Heath-Carteranthropometric somatotype rating relatesvery highly with the criterion rating byHeath. However, a final somatotype ratingbased on both the photograph and anthro-pometric somatotype is determined by thesetwo sources of data and the distribution ofsomatotypes for the given height-weight ra-tios. For those trained in the method, a

15

rating may be made from the photographand the height-weight ratio table.

In addition to the above methods, Hunt(19) warns us of idiosyncratic rating meth-ods these methods develop when differentraters (simply because they are differentpeople) place different interpretations on thecriteria for the components, in making sub-jective photographic ratings.

It can be seen from the descriptions ofthe various methods above that they differwith respect to the basic premise of whetheror not the somatotype rating is an attemptat assessing the constitutional and unchang-ing pattern of somatotype, or whether it isa phenotypic (i.e., present) estimate of thesomatotypc. The estimates of the compo-nents are based on one or more combina-tions of photoscopic estimates, planimetry,anthropometry, and functional performance.When using any of the above methods thevalidity of the method, the reliability and ob-jectivity of the ratings, and the measurementused in the ratings are integral parts of a

study.Obviously, somatotype ratings calculated

by various methods arc likely to be different.For information as to the magnitudes of thedifferences between methods and relation-ships between them, the reader is referredto the following studies (10, 11; 12, 14, 15,16, 17, 19, 21, 23, 25, 26, 27, 28, 29, 31,34, 44).

As illustrations of the differences betweenmethods, two examples arc given. In Figures1 and 2 the same group of football players(4, 5) are somatotyped by Parnell's M.4deviation chart method (Figure 1), and bythe Heath-Carter anthropometric method(Figure 2). Comparison of both the meansomatotypes and the distributions on thesomatocharts shows considerable differencesbetween the two groups. One particularlynoticeable effect is the compressing of allsomatotypes with a mesomorphy rating of7, 71/2 , or 8 by the Heath-Carter methodonto a 7 rating, which is maximum for theM.4 method. This lock'of discrimination at

Figure 3. An individual somatotyped by four different methods. Age = 54.3 yrs.; height = 67.6 in.;weight = 145.5 lbs.; height-weight ratio = 12.84. HeathCarter Anthropometric somatotype, 21/2-63/2-21/2; HeathCarter Photoscopic plus Anthropometric somatotype (authors' indepencient ratingsagreed), 2-6-21/2; Sheldon's Trunk Index somatotype, 21/2-4-31/2; and Parnell's M.4 Deviation Chartsomatotype, 21/2-61/2 -4.


the upper end of the scale has forced a lowermean somatotypc rating by the M.4 methodfor mcsomorphy compared to the Heath-Carter method.

A second example is given in Figure 3 inwhich a middle-aged male is rated by theHeath-Carter anthropometric method, theHeath-Carter photoscopic method, the Shel-don's trunk index method, and by the M.4method. Only the two Heath-Carter ratingsare similar on all three components. Theother two ratings are dissimilar from eachother and from the Heath-Carter ratings,especially for the second and third compo-nents.

The essential point to be gained thus far isthat there arc different methods of somato-typing, and that the relationships betweenthese methods are generally known. Thus,the word somatotyping is a generic term em-bracing a number of different methods.

THE RELEVANCE OF SOMATOTYPE

METHODOLOGY TO KINESIOLOGY STUDIES

The choice of the method of somatotyp-ing in kinesiological studies is determinedlargely by the purpose of the study. In gen-eral, we are looking for structure-functionrelationships; therefore if we measure per-formances at a given time (and perhapsagain at a later time), then It would seemlogical to relate these performances to thesomatotypc ratings at the same time. Withrespect to the ratings, two simple criteriaare applicable: a) the rating method shouldbe independent of the functional measure-ments; and b) the ratings should be allowedto vary for the individual. The appropriatemethodology, therefore, is one in which theratings are based on present somatotypes(i.e., they are not age-corrected or presumedto be constant), and the performance scoresare not part of the rating method. In growthstudies, just as it is important to record ab-solute values such as height and weight onthe same scales at different points in time,it is equally important to rate somatotypecomponents in the same manner so that dis-tance and velocity curves may be plotted.


If, on the other hand, we merely wish toidentify the subjects as belonging to a parti-cular somatotype group, then the use of afixed or permanent somatotype rating mayhe useful. In this case, the rating should beregarded as one would regard blood groups,or ethnic classification. This procedurewould appear to have little to recommend itbecause of its limited usefulness in struc-ture-function studies. Furthermore, the gene-tic basis of somatotype has never beenproven or even the possible magnitude of itestablished. In spite of Sheldon's statementsto the contrary (37, 39), for which he pres-ents no supporting data, the majority of theevidence in growth and development studiesis overwhelmingly in the direction of plas-ticity and inconstancy and somatotype values(1, 2, 3, 6, 7, 13, 20, 21, 22, 25, 26, 27, 28,35, 41, 42, 43). Of course, the results ofsuch studies themselves depend on the spe-cific methodology used. If one wishes toprove that somatotype is constant from ageto age, one merely has to shift the scales foreach age group, as is done in the Sheldoniansystem and in Parnell's M.4 system. In thesesystems, different "looking" amounts of thecomponents at different ages are given thesame rating.

In summary, then, we have seen that thereare several distinct somatotyping methodswhich have different bases and meaning. Be-cause of these differences, several questionsneed to be answered: a) What method ofsomatotyping is to be used? b) Is the methodappropriate to the problem? c) Who is to dothe ratings and what is his reliability? Inkincsiological studies the use of a methodwhich is not age-corrected, and is based ona scale which reflects changes in somatotype,is recommended.

That certain important and substantalrelationships have been shown between so-matotype components and structural andfunctional variables is encouraging, and al-though different systems with different mean-ings have been used this does not obscurethe fundamental fact that such associationsdo exist. An understanding of the differentmethodologies may help us to clarify someof the relationships and to discover others.

17

References1. Barton, W. H., and Hunt, E. E., Jr. Somato-

type and adolescence in boys: A longitudinalstudy. Human Biology 34:254-70, 1962.

2. Bodel, J. K., Jr. Distribution and permanenceof body build in adolescent boys. Doctoral dis-sertation, Harvard University, 1950.

3. Carter, J. E. L. An analysis of somatotypes ofboys aged twelve to seventeen years. Master'sthesis, State University of Iowa, 1958.

4. . Somatotypes of college football play-ers. Research Quarterly 39:476-81, 1968.

5. Somatotypc characteristics of cham-pion athletes. Proceedings of the Anthropo-logical Congress, Praha-Humpolec. Czecho-slovakian Academy of Science, September1969.

6. , and Phillips, W. H. Structural changesin exercising middle-aged males during a 2-year period. Journal of Applied Physiology27:787-94, 1969.

7. Clarke, H. H. (Ed.). The Medford. Oregon,Boy's growth study. Curriculum Bulletin No.238, University of Oregon, Eugene, Oregon,Now :fiber 1963.

8. Cureton, T. K., Jr. Physical fitness, appraisaland guidance. London: Henry Kimpton, 1947,Ch. 11.

9. . Physical fitness of champion athletes.Urbana, Ill.: University of Illinois Press, 1951.

10. Damon, A., and others. Predicting somatotypcfrom body measurements. American Journal ofPhysical Anthropology 20:461-74, 1962.

11. DeWoskin, S. F. Soniatotypes of worrier, in afitness program. Master's thesis, San DiegoState College, 1967. (University of OregonMicrocard PE 1007)

12. Dupertuis, C. W., and Emanual, I. A statis-tical comparison of the body typing methodsof Hooton and Sheldon. WADC Tech. Re-port 56-366, ASTIA Document AD97205.Wright Air Development Center, U.S. AirForce, Wright Patterson Air Force Base, Ohio,1956.

13. Garn, S. M. The somatotyping of young chil-dren. I. Standards. II. Constancy of the ratings.(Proceedings of the Nineteenth Annual Meet-ing of the American Association of Physical

Anthropology) American Journal of PhysicalAnthropology 8:261, 1950.

14. Haronian, F., and Sugarman. A. A. A com-parison of Sheldon's and Parnell's methods forquantifying morphological differences. Amer-ican Journal of Physical Anthropology 23 :135-42, 1965.

15. Heath, B. H. Need for modification of soma-totyping methodology. American Journal ofPhysical Anthropology 21:227-33, 1963.

16. , and Carter, J. E. L. A comparison ofsomatotype methods. American Journal ofPhysical Anthropology 24:87-99, 1966.

17. A modified somatotypc method.American Journal of Physical Anthropology.27:57-74, 1967.

18. Ilooton, E. A. Handbook of body types in theUnited States Army. Cambridge, Mass.: De-partment of Anthropology, Harvard Univer-sity, 1951.

18

19. Howells, W. W. Variation of external bodyform in the individual. Cambridge, Mass.:Peabody Museum, Harvard University, 1957.

20. Hunt, E. E., Jr. A note on growth, somato-type, and temperament. American Journal ofPhysical A n thropology 7:79 -89, 1949.

21. , and Barton, W. N. The inconstancyof physique in adolescent boys and other limi-tations of somatotyping. American Journal ofPhysical Anthropology 17:27-36. 1959.

22. Lasker, G. W. The effects of partial starvationon somatotypc: An analysis of material fromthe Minnesota starvation experiment. Amer-ican Journal of Physical Anthropology 5:323-41, 1947.

23. Livson, N., and McNeill, D. Physique andmaturation rate in male adolescents. ChildDevelopment 33:145-52, 1962.

24. Meredith, H. V. Comments on "Varieties ofHuman Physique." Child Development 11:301-309, 1940.

25. Morton, A. R. Comparison of Sheldon's trunk-index and anthroposcopic methods of somato-typing alui Mar relationships to the maturity,structure, and motor ability of the same boysnine through sixteen years of age. Doctoraldissertation, University of Oregon, 1967. (Mi-crocard PE 1032)

26. Munroe, R. A. Relationship between somato-type components and maturity, structural,strength, muscular endurance, and motor abil-ity measures of twelve-year-old boys. Doctoraldissertation, University of Oregon, 1964. (Mi-crocard Psy206)

27. ; Clarke, H. H.; and Heath, B. H. Asomatotype method for young boys. AmericanJournal of Physical Anthropology 30:195 -201,1969.

28. Newman, R. W. Age changes in body build.American Journal of Physical Anthropology10:75-90, 1952.

29. Parnel! R. W. Somatotyping by physical an-thropometry. American Journal of PhysicalAnthropology 12:209-239, 1954.

30. . Physique and mental breakdown inyoung adults. British Medical Journal 1:1458,1957.

31. . Behaviour and physique. London: Ed-ward Arnold, 1958.

32. Parnell, R. W., and others. Physique and thegrammar school entrance examination. Brit-ish Medical Journal 1:1506.10, 1955.

33. Petersen, G. Atlas for somatotyping children.Springfield, Ill.: Charles C. Thomas, andAssen, the Netherlands: Royal Vangorcum,1967.

34. Roberts, D. F., and Bainbridge, D. R. Niloticphysique. American Journal of Physical An-thropology 21:341-70, 1963.

35. Seltzer, C. C., and Gallagher, J. R. Somato-types of an adolescent group. American Jour-nal of Physical Anthropology 4:15:!;-68, 1946.

36. Sheldon, W. H. New developments in soma-totyping technique. Lecture delivered t;t Chil-dren's Hospital, Boston, Massachusetts, March13, 1961.

37. A brief communication on somato-typing, psychiatyping and other Shrldoniatzdelinquencies. Paper read at the Royal So-ciety of Medicine, Lo.:.don, May 13, 1965.


38. Dupertuis, C. W.; and McDermott.E. Atlas of men. New York: Harper andBros., 1954.

39. Sheldon, W. H.; Lewis, N. D. C.; and Tenney,A. M. Psychotic patterns and physical con-shim?. Invitation lecture, Eastern Psychiat-ric Research Association Symposium, NewYork, November 16, 1968.) In Schizophrenia,current concepts and research. Hicksville,N. Y.: D. V. Siva Sankar, PM Publications.(In press)

40. Sheldon, W. H.; Stevens, S. S.; and Tucker,W. B. The varieties of human physique. NewYork: Harper and Bros., 1940.

41. Sinclair, G. D. Stability of physique types ofboys nine through twelve years of age. Mas-ter's thesis, University of Oregon, 1966. (Mi-crocard PE 833)

42. .Stability of somatotype components ofboys ages twelve through seventeen years andtheir relationships to selected physical andmotor factors. Doctoral dissertation, Universityof Oregon, 1969. (Microcard PE 1042)

43. Tanner, 1. M. The effect of weight training onphysique. American Journal of Physical An-thropology 10:427-62, 1952.

44. . Reliability of anthroposcopic somato-typing. American Journal of Physical Anthro-pology 12:257-65, 1954.


ERNEST W. DEGUTIS

A Prober Solving Approach to the

Study ot Muscle Action

IT IS COMMON PRACTICE for physical edu-cation majors, in their study of the humanbody, to put considerable emphasis on thestudy of the skeletal, or voluntary, muscularsystem. The emphasis is justifiable since theskeletal muscular system: a) makes up over40 percent of the total body weight; b) byvirtue of its contractile properties, producesmovement for utilitarian, recreative, and af-fective (expressive) purposes; c) containskinesthetic-proprioceptive sense organswhich, from one point of view, render mus-cle as man's most important sensory mech-anism; and d) is critical to the develop-ment of the various organic systems becauseof the supportive role of these organic sys-tems in muscular activity.

One aspect of skeletal muscular systemstudy includes its gross structure. Invari-ably, the study of gross structure involvesa detailed description of many of the body'smuscles in terms of each muscle's attach-ment and the movements the muscle is pur-ported to produce when it contracts. Learn-ing points of attachment and movements oractions attributed to contracting musclescalls for considerable unattractive memori-zation. The problem of dealing with a newvocabulary, the extensiveness of the infor-mation with its similarities and differences,and the infinite detail can make the study of

Ernest W. Degutis is professor of physicaleducation at Western State College of Col-orado, Gunnison. The art work for this arti-cle was provided by Don Oden.

20

muscles and muscle actions a tedious, bur-densome, and difficult chore. Consequently,many students are likely to lose interest. Thisis unfortunate, as the subject is importantin itself and is fundamental to the student'sprofessional preparation, since such knowl-edge is related to the ability to deal satis-factorily with content in other courses (e.g.,kinesiology, physiology of exercise, athletictraining).

Most teachers are aware of the shortcom-ings of rote learning. When the attempt tolearn is obstructed by unattractive, boringmemory work, searching questions shouldbe asked regarding methods. This is espe-cially true if waning interest, low level moti-vation, and underachievement are in evi-dence. Thoughtful questions directed to theevaluation of methodology might include thefollowing: Is there concern for the masteryof principles, or only for the mastery of lim-ited content which seems to lack significanceand is readily forgotten? Is the student chal-lenged in a stimulating way? Is the student'scuriosity aroused, his appreciation deepened,and his motivation heightened? Does the ex-perience contribute to the student's abilityto "learn to learn"? Is the overall experiencesatisfactory? Is it productive? Affirmativeresponses to these questions are consistentwith a method that has much to commend it.

Various methods may be employed in thestudy of movements or actions purportedlyproduced by contracting muscles. A goodway to initiate such study is through a prob-lem solving approach, herein termed "mus-


LorIG 11-111:::q a4

irrA - Fitoo

t.4

mom"'

1

1

41 611111111

1

1

7 "I./ (tv..C*.asN]wxslasF--

P-.2' Fkonl-t-tNv)

NOV-0.M/ PoisTERA oP--I

I(.6.64T-rkt._ TRAtisAlat-sm-)

tiorAts

INF.MN

Figure 1. Fundamental body and movement planes (sagittal, frontal, and transverse), and funda-mental motion axes (transverse, anterior-posterior, and longitudinal). When an axis is named by theplanes in which it lies, the following applies: longitudinal or frontalsagittal axis; transverse ortransverse-frontal axis; and anteriorposterior or transverse-sagittal axis.


de action analysis" (MAA).* This methodhas the potential for eliminating many of theundesirable concomitants of the rote learn-ing approach. The study of muscle points ofattachment necessarily receives emphasis inMAA but it becomes subordinate to the me-chanical principles underlying the "how"and "why" of muscle action. In other words,MAA enables the student to develop a ra-tionale for explaining the movements pro-duced by the contraction of a muscle or amuscle group. It is not satisfactory to simplystate that a muscle may act in flexion or inany other movement. The reason must bestated in terms of the muscle's line of forceor pull, and the positional relationships ofthe muscle's line of force to the motion axisupon which it is acting. Thus muscular ac-tion, in the mechanical sense, must be rea-soned out. The invariably favorable responseof students who have used the MAA prob-lem solving approach to muscle study per-mits the enthusiastic endorsement of thismethod. The intent of this paper is to de-scribe the MAA method and the essence ofthe preparatory procedures by which thismethod may be satisfactorily implemented.

Preparatory learning experience, for muscleaction analysis (MAA)

It has been found that the MAA methodcan he efficiently and successfully imple-mented if preparatory learning experiencesof the type described herewith are providedfor the student. Attempts to implement thismethod without such experience have notbeen as satisfactory.

1. A study of pertinent infornmtion rela-tive to sonic basic anatomical considera-tions: a) identifying and defining funda-mental body and movement planes, andfundamental motion axes (Figure 1); b)anatomical terms of direction or position;zinc] c) anatomical movements. (Note: Inthis text, it is understood that all movementsare initiated from the anatomical position.)

* MAA is a modified form of a method longused in the Department of Anatomy at theUniversity of Wisconsin.

22

2. A study of the skeletal system, whichresults in a thorough familiarity with thesurface markings on hones which serve aspoints of attachment for muscles.

3. A study of the arthrodial system whichstresses the functional and structural classi-fication of joints. The structural classifica-tion of joints is based primarily on the ab-sence or presence of a joint cavity (the con-tinuity or discontinuity of skeletal parts).The functional classification of joints isbased on the number of motion axes aboutwhich the joint admits movement and, there-fore, the number of planes in which move-ment can occur. Considerations common tospecific joint actions would further includethe identification of the motion axis or axesinvolved, and the miming of anatomicalmovements and the planes in which move-ments occur.

4. A study of selected mechanics relatedto leverage and force, since MAA basicallyinvolves the apnlication of leverage and forceprinciples. It is especially important for stu-dents to comprehend that a muscle's line offorce or pull, when applied at an acute an-gle (and therefore coursing obliquely) to askeletal part and at a distance from its mo-tion axis, has two components acting atright angles to each other, alai each actingon the motion axis to produce angular :no-tion (Figure 2). Otherwise, a line of forcewhich is applied at a right angle to a skele-tal part and acts to produce angular motionhas a direct, single pull or line of action(figure 3).

5. The development of the "movementproducing combination" concept, whereinangular movement, in a mechanical sense, isrecognized as a function or a combinationof two factors: a) the direction of the lineof force, and h) the line of force's positionalrelationship to the motion axis.

Students have little difficulty in under-standing the movement producing combina-tion (M PC) concept if the approach is madeby use of the wheel and axle principle. Animprovised wheel and axle of sorts can beused to demonstrate what the students al-ready know, i.e., if a wheel is to turn on itsaxis (angular motion), a force must Le ap-plied to the wheel at a distance from the


Figure 2. Schematic illustration of the two components of force (fl, f2) of a muscle's obliquelycourse ig line of pull (F), both acting on a skeletalpart (b) to produce angular motion about amotion axis (a) located at the center of the jointcavity.

Figure 3. Schematic illustration of the line offorce (F) of a muscle acting at a right angle toa skeletal part (b), which produces angularmotion about an axis (a) by its single, direct lineof pull.

KINF_SIOLOGY REVIEW 1971

axis and in a plane that is perpendicular toit. This should be demonstrated with thewheel positioned in each of the three funda-mental movement planes. In order to iden-t;fy movement producing combinations thewheel, as it is positioned in each of the threeplanes, must be spatially oriented in termsof anatomical directions to shcw in whatdirection the turning force is acting, andwhat is the relationship of the line of force,positionally speaking, to the motion axis.

Generally, students haw; little or no diffi-culty in their understanding up to this point.The problem comes in their attempt to trans-fer the wheel and axle and the MPC con-cepts to the body and visualize them opera-tionally. For example, if these concepts areperceived in the body context there shouldbe recognition that when the wheel is posi-tioned in the sagittal plane and its motion,therefore, occurs in that plane, flexion (Fig-ure 4a-b) and extension (Figure 4e -d)movements are being considered. Figure 4eillustrates the importance of understandingthe positional relationship of the line offorce to the motion axis. In Figure 4e, thelines of force, F and F1, act in the same di-rection, superiorly and posteriorly. However,force F has anterior and superior relation-ships to the motion axis as indicated by itscomponents f1 and f., respectively. ForceF1 has inferior and posterior relationships tothe motion axis as indicated by its compo-nents, f3 and f4. Forces F and F1 will bothproduce motion in the sagittal plane but inopposite directions. 7igure 4f expresses thesame concept but with the lines of directionof forces F and F1 reNersed. The MPC con-cept related to flexion extension movementsoccurring about the transverse axis and illus-trated in Figure sequence 4af, is similarlyillustrated in Figure sequence 5af for ab-duction-adduction movements about an an-terior-posterior axis, and in Figure sequence6af for medial and lateral rotation move-ments which occur about the longitudinalaxis.

Some students grasp the concept underconsideration in its body context veryreadily. For these students, muscle studyfrom the standpoint of MAA can be a cap-tivating challenge. Others have more diffi-

23

,0 i 4

SUPERIOR

MK for counter clockwise movementor for flexion

4a.NIPC'f': sup. pull. ant. rel.MPC pos. pull. sup rel

4c.NIPC V: ant. pull, sup. rel.

inf. poll. ant. Tel

MPC for clockwise mos ementor for extension

Tk

4b.NIPC inf, pull. pos. rd.NIPC f': ant. pull. inf. rd.

Effects of lines of force with identicaldirections but %sigh different positional

relationships to the motion axis

4e.Sup. and pos. lines of force(F1. i producing counterclockwi e and clockwisemotion, respectively, dueto their different positionalrelationships to the motion

4d.ls1PC I': pos. pull. inf. rel.NIPC sup. pull. pos. rel.

INFERIOR

4f.Inf. and ant. lines of force

F,) producing clock-wise and counter clockwisemotion, respectively. dueto their different positionalrelationships to the motionaxis

Figure sequence 4af. Identifying MPC in the sagittal plane. The spatial orientation as shown iscommon to all diagrams and allows: a) for the determination of the directions of the line of force,and b) for the positional relationships of the lines of force to the transverse motion axis, representedas a point within the wheel. In the body context, sagittal plane movements are flexion and ,Ixtension.


a2

SUPERIOR

MPC for counter clockwise movementor for abduction

5a.MPC f', sup pull. lat rcl.MPC f': med. pull. sup. rel. IPC for clockwise movement

or for adduction

5c.MPC f' lat. pull, sup. ridMPC f:: inf. pull. lat. rel.

5h.MPC f": inf. pull, med. rel.MPC f': lat. pull. inf. rd.

Effects of lines of force with identicaldirections but with different positional

relationships to the motion axis

5d.MPC P: med. pull. inf. rd.M PC P: sup. pull. med. rel.

.Sc. 5f.Sup. and med. line of force Inf. and lat. lines of force( F', F:1 producing counter (F'. F 1 producing clock-clocks, ise and clockwise wise and counter clockwisemotion, respectively, due motion, respectively, dueto their different positional to their different positionalrelationships to the motion relationships to the motionaxis axis

INFERIOR

-4m

Figure sequence 5af. Identifying MPC in the frontal plane. The spatial orientation as shown iscommon to all diagrams and allows: a) for the determination of the directions of the lines of force,and b) for the determination of the positional relationships of the lines of force to the anterior-posterior motion axis, represented asa point within the wheel. In the body context, frontal planemovements are abduction and adduction.


O

2

ANTERIOR

1111PC for counter clockwise movementor for medial rotation

fia.NIP(' I': ant. pull. lat. rel.NI PC t:: tined pull. ant. rel.

fie.

NIP(' t' lat pull. ant re''sfP(' pus pull. lat rel

NNW for clockwise mm ennuior lateral rotation

fib.pos. pull. med. rel.

NIP(' lat. pull. pos. rel.

Effects of lines ta force with identicaldirection!: but with different positional

relationships to the motion asis

fieAnt. and med. lines 01 Imo:( 1' producing coonterclockwise and clock% isemotion, respectisely, dueto their different positionalrelationships 10 the motion:Isis

al.

NIPC med pus. rel.NIP(' ant pull. med. rel.

POSTERIOR

Pos. and lat. lines of force(F'. F2I producing clock-wise and counter clockwisemotion. respectively, dueto their different positionalrelationships to the motionaxis

tm

Figure sequence 6af. Identifying MPC in the transverse plane, The spatial orientation as shown iscommon to all diagrams and allows for the determination of: a) the directions of the lines of force,and b) the positional relationships of the lines of force to the longitudinal motion axis, representedas a point within the wheel. In the body context, transverse plane movements are medial andlateral rotation.


culty but they manage to some degree; theytoo seem to enjoy the confrontation withproblem solving. Unfortunately, there are afew students who have exceptional difficultywith this approach. It appears that they aredestined to approach the study of grossmovement as produced by contracting mus-cles from the uninteresting, ineffective, andconsiderably less meaningful avenue of purememory work.

In order to facilitate the development ofthe MPC concept it may be well to devisework experiences for students which chal-lenge their understanding. A number of suchhomework experiences have been devisedand are available upon request,

Suggested procedure for MAA

In the foregoing prelv;atory learning ex-periences for MAA, a mechanically orientedproblem solving approach to the study ofmovements produced by contracting muscleswas identified and briefly elaborated on. Inthe following, the procedure for MAA itselfis described in detail. The procedure hasfour emphases, which consist of determin-ing: a) the directions of pull (lines of force)of the muscle; b) the effective pulls of themuscle; c) the positional relationships ofthe muscle's pull to the axis of motion; andd) the movement producing combinations(MPCs). If the suggested procedure is ad-hered to, a minimum of difficulty should beencountered.

I. DETERMINING THE DIRECTIONS OF PULL(LINES OF FORCE) OF THE MUSCLE

A. Locate the points of attachment of themuscle.

B. Decide what point of attachment isfixed and which is free. The free point isattached to the body segment undergoingmovement; the fixed point is on the stabil-ized body segment. Ordinarily, the fixed at-tachment is proximal and the free attach-ment is distal in the extremities. When suchis the case, the terms fixed and proximal,and free and distal. are synonomous, or thereverse may be true.


(Note: The reversal of muscle action isdiscussed in section E.)

C. Determine the positional relationshipsof the fixed attachment to the free attach-ment b" observing these points of attach-ment and the course of the muscle. For ex-ample, the fixed point of attachment may besuperior or inferior, medial or lateral, andposterior or anterior to the free point of at-tachment.

D. The pulls of the muscle are determinedon the basis of the positional relationshipsof the attachment points to each other. Sincea contracting muscle may simultaneouslypull in several directions toward its fixedpoint of attachment, a muscle whose fixedpoint of attachment is superior, medial, andposterior to its free point of attachmentexerts its pulls or lines of force in those di-rections.

E. Discussion. With reference to the freeand fixed points of attachment and, there-fore, the lines of pull, the muscles them-selves have no preference. In a shorteningcontraction, the muscle's disposition is topull from each of its ends toward its center.It is the nervous system, fulfilling its integra-tive functions, that dictates to the musclesand establishes the free and fixed body seg-ments and, thereby, the free and fixed pointsof attachment and the muscle's lines of pull.Under varying conditions the points of at-tachment and the lines of pull can be re-versed ... Whereas in one instance the mus-cle's fixed and free points of attachment mayhe proximal and distal, respectively (e.g.,elbow flexors moving the forearm), underdifferent conditions the fixed points of attach-ment may be distal and the free point maybe proximal (e.g., elbow flexors moving thehumerus) . . . In those instances where amuscle or its tendon courses so as to changeits direction (pulley effect), one of the pointsof attachment is determined at that pointwhere the pulley effect occurs . . , Once thedirections of pull of the muscle are deter-mined for a given condition (e.g., the ana-tomical position), they will apply at all timesfor that condition. Therefore, all of themuscle's pulls should be determined at theoutset and noted.

27

II. DETERMINING THE EFFECTIVE PULLS OFTILE MUSCLE

A. The effective pulls of a muscle (lineof force) are those which, by their action,produce angular motion. In order to produceangular motion the pulls of a muscle mustact at a distance from the motion axis andin the plane in which motion occurs. Con-versely, an ineffective pull would be onewhose action does not produce angular mo-tion about a given motion axis. Rather, itstendency would be to produce linear motionby acting in a direction parallel to the mo-tion axis. In a three-dimensional system amuscle may have pulls which are superioror inferior, medial or lateral, and anterioror posterior. Two of the three pulls will beeffective; one will not be. The chart belowidentifies the effective and ineffective pullsfor each of the three fundamental motionaxes.

IneffectiveMotion axis Effective pulls pulls

transverse a. superior or medial orinferior lateral

b. anterior orposterior

anterior- a. superior or anterior orposterior inferior posterior

b. medial orlateral

longitudinal a. anterior or superior orposterior inferior

b. medial orlateral

B. For a single joint, determine the axesof motion and movement planes permittedby the joint's structure. Use bony landmarksto indicate the approximate location of eachaxis on a specimen and on the living sub-ject. Generally, the transverse and anterior-posterior axes of motion may be consideredto pass through the center of the joint, whilethe longitudinal axis may be considered topass lengthwise through the center of thebone. There are a number of exceptions tothis statement but this basic approach gen-erally will be satisfactory.

C. When determining the effective pulls,consider only one motion axis at a time.

28

III. DETERMINING THE POSITIONALRELATIONSHIPS OF THE MUSCLE'SPULLS TO THE MOTION AXIS

A. Whenever possible, examine the mus-cle and the joint structure from a view whichshows the axis under consideration in crosssection, i.e., as a point.

B. The muscle's pull acts on the motionaxis at the points where its line of directioncrosses the motion axis. The positional rela-tionships of the muscle's line of pull to themotion axis should be determined at thesepoints, not where its attachments are. Thisis particularly necessary in those cases wherethe line of direction of the muscle's pullneeds to be extended in order to cross themotion axis because the muscle does notactually cross the motion axis (e.g., the longhead of the triceps around the transverse oranterior-posterior motion axes of the shoul-der joint). Accordingly, the positional rela-tionships of a muscle's line of pull to a givenmotion axis may be superior or inferior,medial or lateral, and anterior or posterior.

C. The positional relationships of themuscle to the motion axis may be deter-mined schematically as follows: a) Examinethe muscle-axis relationship as indicated in A.b) Intersect the point representing the mo-tion axis at right angles with .wo brokenlines. Extend the broken lines until they in-tersect with the line representing the mus-cle's line of pull (Figure 7a). It is at thepoints where the b oken lines intersect withthe muscle's line of pull that the positionalrelationships of the line of pull to the motionaxis'are determined (RI, R.). Obviously, thenaming of the positional relationships de-pends on the motion axis under considera-tion and the appropriate spatial orientation.To be consistent with the wheel and axleprinciple and to maintain sensitivity to an-gular motion, it may be well to proceed asabove but to visualize the axis as a pointwithin a wheel (Figure 7b).

D. If a muscle is so placed that an axispasses through it, the muscle needs to beconsidered as two or more units and shouldbe analyzed accordingly (e.g., the deltoideusaround all axes of the shoulder joint).


910454EasE-ivto-rtoti Aios

-tut?

Figure 7ab. Schematic determination of posi-tional relationships (131, R2) of a muscle's line ofpull to a transverse motion axis.

IV. DETERMINING MOVEMENT PRODUCINGCOMBINATIONS (MPCs)

A. For a given axis of motion, identify thetwo anatomical movements possible (e.g.,flexion and extension about a transverseaxis).

B. MPCs are a function of: a) the direc-tion of the muscle's line of pull, and b) thepositional relationships of the muscle's lineof pull to the motion axis under considera-tion. Generally, the muscle's line of pull willhave two components and two positionalrelationships to the motion axis. When suchis the case, the proper matching of pulls andrelationships will result in two MPCs foreach movement (refer to Figure sequences4-6).KINESIOLOGY REVIEW 1971

Figure 8. The right posterior deltoid muscle andits line of pull from the anatomical position.

MAA illustratedFigure sequences 4-6 have used the wheel

and axle principle to illustrate the MPC con-cept. In this context the MPC concept isreadily understood by most students. Theproblem for the student arises when the at-tempt is made to transfer these understand-ings to the body context. While the instruc-tor may help, it is the student himself whohas to understand. To illustrate the MAAprocedure in the body context, the posteriordeltoid muscle has been selected for anal-ysis (Figure 8). This muscle acts on the idaxial shoulder joint to produce movementin each of the three fundamental bodyplanes. The deductive analysis is shown onan MAA worksheet (Figure 9) used bystudents to facilitate their efforts and foreasy checking by the instructor. Repeatedexperiences of this type will lead the waytoward mastery of MAA.

It should be noted that the MPCs asshown emanate from the anatomical posi-tion. As a segment moves through its rangeof motion, the line of pull of the involvedmuscle and its positional relationships to

29

OM

USC

LE

AC

TIO

N A

NA

LY

SIS

(MA

A)

WO

RK

SHE

ET

Func

tiona

lM

. Pro

xim

alM

. Dis

tal

M. D

irec

tions

Join

tC

lass

ific

atio

nM

uscl

e (M

)A

ttach

men

tA

ttach

men

tof

For

ce (

3)Sh

ould

erT

riax

ial

Post

erio

rL

ower

bor

der

Del

toid

(gle

no h

umer

al)

delto

idof

sca

pula

rtu

bero

sity

spin

e

Ana

tom

ical

Mot

ion

Mov

emen

tA

xis

Supe

rior

(su

p.)

Med

ial (

med

.)Po

ster

ior

(pos

.)

MO

VE

ME

NT

PR

OD

UC

ING

CO

MB

INA

TIO

NS

(MPC

's)

Eff

ectiv

e L

ines

M. P

ositi

onal

Rel

atio

nshi

psPr

oof

of f

orce

(2)

and

to M

otio

n A

xis

(2)

Dia

gram

s1.

Ext

ensi

onT

rans

vers

eM

PC-1

:su

p.an

dpo

s.(T

rans

vers

e-M

PC-2

:po

s.an

din

f.Fr

onta

l)

2. A

dduc

tion

Ant

erio

r-M

PC-1

:su

p.an

dm

ed.

Post

erio

rM

PC-2

:m

ed.

and

in:.

(Sag

ittal

-T

rans

vers

e)

SO,

It4F

3. L

ater

alL

ongi

tudi

nal

MPC

-l:

med

.an

dpo

s.R

otat

ion

(Fro

ntal

-M

PC-2

:po

s.an

dla

t.Sa

gitta

l) Figu

re 9

. MA

A w

orks

heet

illu

stra

ting

anal

ysis

of

the

righ

t pos

teri

or d

elto

id m

uscl

e.

MtO

AN

T

Museswpm OFPut v

MTlegicte. %smut*.MOTION Pacts

LINT-

Figure 10. Proof diagram for the right posterior deltoid muscle assisting in abduction when thehumerus approximates 180° abduction.

the motion axis change; therefore the move-ment produced by the muscle, mechanicallyspeaking, is a function of different MPCs. InFigure 10 the line of pull of the posteriordeltoid is shown schematically in a proofdiagram when the humerus approximates180° abduction. In this position it is likelythat this muscle, ordinarily a strong abduc-tor, assists other muscles in the abductionmovement because of the shift in its line ofpull and positional relationships to the an-terior-posterior motion axis.

Concluding remarks

This paper has been written to explain amethod of analyzing the action of contract-ing muscles from a mechanical standpoint.

Experience supports the view that MAA isreplete with the elements of problem solv-ing. It can be, therefore, a potentially chal-lenging, stimulating, productive, and satis-fying way of studying muscle action.

It is recognized that muscle action is amultidimensional phenomenon and thatthere arc limitations in viewing movementsproduced by contracting muscles in a purelymechanical sense. Results from electromyo-graphic studies, for example, tell us thatmuscle action is not simple. It must hepointed out, however, that if a muscle is toproduce a movement, mechanically speaking,it must first be in position to do so; whetherit does or not depends upon environmentalconditions and the integrative action of thenervous system.


T. L. DOOLITTLE

Errors in Linear Measurement withCinematographical Analysis

IT HAS BEEN OVER 30 YEARS since Cureton(1) set forth elementary principles for theuse of cinematographical methods in anal-yzing human performance. These were inessence reiterated by Glassow (2) andHubbard (3), and numerous studies inphysical education have used them; how-ever, there has been a paucity of literaturein the discipline regarding any improvementin or critical analysis of these methods. Noss(4), a noted exception, published his workregarding the inherent errors in angular meas-urements. Techniques for the reduction oferrors in linear measurements have not beenelucidated. Since the utilization of cinema-tographical methods for analysis of complexor high speed human movements is becom-ing increasingly popular, it appeared thatthere was a great need for establishing moreaccurate techniques for reducing these linearerrors.

The assessment of linear distance fromfilm frequently is a primary parameter eitherfor distance per se or for the ultimate deter-mination of velocity. This method is subjectto errors that the naive investigator may failto take into account, even though he is fol-lowing what is generally thought to be ac-

T. L. Doolittle is an associate professor atCalifornia State College, Los Angeles. Hewishes to extend his sincere appreciationto Alton Boynton and Barbara McClure,former graduate students at CaliforniaState College, for their able assistance inthis investigation.

32

cepted practice. Perspective errors resultwhen three-dimcnsional objects are reducedto two dimensions on the film, with thosenearer appearing larger than those that aremore distant. Increasing the camera-to-sub-ject distance, while partially reducing thissource of error, does not completely elimi-nate it, nor is this recommendation alwayspractical. Utilization of a grid for a back-ground creates perspective errors for whichcorrections should be made. A second sourceof error scaling errors may result whenthe projected image is other than actual(life) size. An object of known size isusually included in the field of view, meas-ured on the projected image, and a scalefactor derived to convert measured values(in inches, millimeters, etc.) to the desiredvalues (in feet, meters, etc.). The accuracyof the scale factor obviously influences theaccuracy of the ultimate determination. Thesignificance of this potential source of error,often overlooked, will become quite apparentlater in this report.

The placement of the object of knowndimension in the plane of action will reduceperspective errors, but unless it is of largesize scaling errors will be enhanced. Re-gardless of its size, the object may impedethe desired action, requiring its placementparallel to the plane of action. This is thesituation that exists when a grid is used fora background. Figure 1 illustrates the phys-ical arrangement of a camera, a distance ofhypothetical action (r), and a grid. On thetwo-dimensional image of a photograph r


will appear as n.* This may be observed inFigure 2 where the hurdles appear to besuperimposed on the grids. The effect ofaltering d and/or s (Figure 1) also may beobserved in these photographs. The purposeof this study was to evaluate the relativeeffects of perspective and scaling errors, andto investigate techniques for eliminatingthem.

ProcedureFrame by frame analysis, such as might be

employed in determining the distance be-tween two points for subsequent computa-tion of velocity, was simulated in each ofeight still photographs. In actual analysisthese two points would undoubtedly occurin different frames; however, for purposesof this study having them in the same framepermitted standardization without detractingfrom the basic technique. Hurdles, in a planeperpendicular to the axis of the camera,were utilized to represent the points. Thedistance (r) between the right vertical barson the right ttnd middle hurdles (Figure 2)was held constant at 12.5 feet throughoutthe investigation. The axis of the camera wasmaintained perpendicular to the plane of thegrids at the center of the middle grid. Thevalues for d and s (Figure 1) were variedas indicated in Table 1. Frames 1, 2 and 3,and 4 and 5 were originally photographedfor purposes of testing different exposures,but ultimately were utilized to determine.intra-operator error. Frames 1, 5, 7, and 8are exemplified by Figure 2, A, B, C, and D,respectively. Three operators independentlyanalyzed the film using a Kodak RecordakModel MPE-1.

PERSPECTIVE ERRORS

The first subproblem in this investigationwas to determine the magnitude of perspec-

* In keeping with the terminology and symbolsused subsequently in this report, it shouldbe noted that r and n are life size; thus, inreality, on the projected image one is oberv-ing ni which when multiplied by the scalefactor (k) becomes n (n = km).


camera

Figure 1. Diagram of camera, grid, and hurdleplacement, showing various variables.d distance from camera to plane of hurdles(i.e., simulated plane of dction); line also repre-sents axi_ of camera lens.s distauce hurdles were in front of referencegrid.r actual distance between points being measu refl.n uncorrected value of r measured from theprojected image from the film as m (in mm) andconverted to feet by the scaling factor k (seetext).r' distance one of the points was offset fromthe center of the image.n' apparent value of measured from the film(as with n)a distance between two hypothetical points pand q that is not parallel to the grid. (Note also,that a projects the same m, thus the same n, asr unless precautions and/or corrections aretaken into consideration).N angle opposite side n; also opposite side r;equals angle R minus angle N'.N' angle opposite side n'; also opposite side r';tangent N' equals n'/(s+d) or r' /d.R ank:e opposite side n-I-n% also oppositer+r'; tan3c nt R equals ;n-l-n')/(d+s), or(r+ r') /d.

33

A

Figure 2. Four views of grid and hurdle placement. Distance (r) between right vertical bars of rightand middle hurdles was 12 9 ft. in all exposures. Camera to hurdle plane (d) and hurdle plane to gridplane (s) distances were varied as follows: (A) d 80 ft., s = 4 ft., (B) d = 56 ft , s = 4 ft.;(C) d = 46 ft., s = 9 ft.; and (D) d = 96 ft., s= 9 ft. Grids were made of squares 1 ft. on a sideand were placed: 13'2" on center for views A and B; and, 11'7" on center in views (.; and D.

TABLE 1. VARIABLE VALUES FORTHE EIGHT PHOTOGRAPHS

Frame d (ft.) s (ft.)d

d s

1,2,3 80 4 .952384,5 56 4 .933336 4f 4 .920007 46 9 .836368 96 9 .91429

d & s are as indicated in Figure 1.

Live errors and explore techniques for theirreduction. A scale factor (k) was determinedfor each frame by noting the number ofmillimeters on the graph paper that wereequivvii-mt to one foot on the projected imageof the grid. The distance between the two

points the right vertical bars of the twohurdles) was measured in millimeters (m).The scale factor was used to convert m inmillimeters to n in feet (n = km). Formula(1)

% Error n r X. 100 = n12.512.5

X 100 (1)was utilized to determine the percentage oferror in n (with respect to the actual distanceof 12.5 feet) prior to correction. Formula(2)

1 =d

X n (2)+ swas employed to correct n for perspectiveerror and resulted in the calculated value, 1.If no other errors were involved / wouldequal r, or 12.5 feet. The percent error inthe calculated 1 was determined by substi-tuting / for it in Formula (1). Table 2 de-


TABLE 2.ACTUAL

OBTAINED (n) AND CORRECTED (1) VALUES FOR THEDISTANCE (r) WITH PERCENTAGE ERROR FROM 12.5

FEET IMMEDIATELY BELOW EACH VALUE

FRAMEOPERATOR

nA

1

OPERATORn

13.17

5.36

13.175,36

13.175.36

13.538.24

13.568.48

13.41

7.28

14.6417.12

13.387.04

B1

12.540.32

12.540.32

12.540.32

12.631.04

12.661.28

12.34-1.2812.25

-2.0012.23

-2.16

OPERATOR Cn 1

13.73 13.089.84 4.64

13.73 13.089.84 4.64

13.73 13.089.84 4.64

12.82 11.972.56 -4.24"

12.86 12.002.88 -4.00"

12.76 11.742.08 -6.08"

14.22 11.8913.76 -4.8813.43 12.297.44 -1.68

1

3

4

5

6

7

8

ft.

ft.

ft.

ft.

ft.

ft.io

ft.

ft.(70

13.45

7.60

13.35

6.80

13,306.40

13.21

5.68

13.205.60

13.366.88

14.7217.76

13.95

11.60

12.81

2.48

12.71

1.68

12.671.36

12.33-1.3612.32

-1.4412.29

-1.6812.31

-1.5212.75

2.00

n Denotes increase in magnitude of thedenotes corrected value was less thanof the error.

picts the measurements in after they hadbeen converted to n and corrected to 1, andthe respective errors, obtained by the threeoperators for each frame. In Frame 1, forexample, Operator A determinNi that thedistance was 13.45 feet, based on measuringthe p.ojected image (m) and converting tofeet (n) with his scale factor (k); this re-sulted in an error of 7.60 percent; employ-ment of Formula 2 resulted in an I of 12.81and a remaining error or 2.48 percent. Itshould be noted from the values for Frames7 and 8 that increasing the distance(s) be-tween the grid and the plane of the actiongreatly increased the error in the uncorrectedvalue n. Increase of the camera-to-actiondistance (d), from 46 feet in Frame 7 to 96feet in Frame 8 compensated for the error inpart, but not to the degree that Formula (2)did, as may be noted from the respectivevalues for 1.


error with correction applied. Minus sign12.5 ft. and solely indicates the direction

SCALING ERRORS

The second subproblem, investigation ofscaling errors, developed when it was ob-served that considerable residual error re-mained in many cases after n had been cor-rected to 1. The negative percentages forerrors in I implied overcompensation, or thatthe value of in was too small. The positivepercentage implied undercorrection, or thatin was too large. This variance of results,combined with the fact that for given valuesof d and s the multiplier of n, d/(d s), inFormula 2 remains constant, dictated thatthe source of the error was in n and that itcould not have been in the degree of correc-tion. There were two probable sources forerror in n: (1) measurement of the image in(in millimeters), and/or (2) inaccuracy inestablishing scale factor. Analysis of two ofthe operators' raw data revealed that they

35

had identical measured values for Frames6 and 7, and varied only .5 mm on theimage measurement for Frames 1, 2, 3, and8; the scale factors, however, varied bygreater percentages (Table 3). A compari-son of the differences between operators,from the data in Table 2 with the percentvariations in Table 3, made it rather ap-parent that the primary source of interopera-tor error was in the scaling factor.

Assuming agreement between operatorsdemonstrated a reasonable degree of ac-curacy, Frames 6 and 7 were utilized todevelop a theoretical scale factor. This wasdetermined by calculating the actual n thatwould be subtended on the grid by r anddividing the measured image ni by it. Theresultant theoretical scale factor for the twoframes in which there was absolute agree-ment and the operators' scale factors are

TABLE 3. COMPARISON OF RAW DATA AND SCALEFACTORS BETWEEN TWO OPERATORS

Measurements from Scale Factor (mm/ft) Deter-Projected Image (mm) mined from Projected Image

Frame B C % Var. B C % Var.1 151.5 151 .33 11.5 11.0 4.35

3

4

5

6

7

8

151.5

151.5

216.5

217

261.5

263.5

140.5

151 .33

151 .33

218 .07

218.5 .07

261.5 0

263.5 0

1.1.1 .35

11.5

11.5

16.0

16.0

19.5

18.0

10.5

11.0

11.0

17.0

17.0

20.5

18.5

10.5 0

4.35

4.35

5.88

5.88

4.88

2.70

TABLE 4. COMPARISON OF THEORETICALAND EMPLOYED SCALE FACTORS

Actuala Image Theor. B's %b C's %bFrame n (ft) m (mm) Scale Scale Error Scale Error

6 13.59 261.5 19.25 19.5 1.30 20.5 6.49

7 14.95 263.5 17.63 18.0 2.10 18.5 4.93

as Actual distance on grid subtended by r (See Figure 1).b Denotes operator's error in percent with respect to the theoretical

(i.e., operator C's scale for Frame 7 was 4.93% in error).

TABLE 5. ACTUAL VERSUS THEORETICAL ERRORSa(all values in percent)

Operator B Operator C

Frame Actual Theor. Diff. Actual Theor. Diff.

6 1.28 1.30 .02 6.08 6.49 .41

7 2,00 2.10 .10 4.88 4.93 .07

a Actual errors are the percent I was in error with respect(Table 2); theoretical errors are the percent the operator'sfactor was in error from the theoretical scale factor.

36 KINESIOLOGY REV;EW 1971

scale

to rscale

contained in Table 4. A comparison of thetheoretical errors in scaling from Table 4with the actual errors for / in Table 2 re-sulted in residual error differences that werenegligible (Table 5). This demonstrated thaterrors in scaling could account for remain-ing error after correction for perspectiveerror had been made.

FindingsThe findings of this investigation were

summarized as follows:I. Formula (2) was found to correct ade-

quately for perspective errors regardless ofthe magnitude of d and s.

2. Accuracy in the establishment of thescale factor was highly important; crrors inscaling can overshadow any benefit derivedfrom correction for perspective crror.

3. Accuracy in measurement of the pro-jected image was of lesser importance thaneither of the above, and discrepancies hereinwere not significant.

4. Intra-operator error between duplicateframes was minimal.

DiscussionThe traditional recommendations for

cinematographical analysis have indicatedthat the linear reference for determining thescale factor should be in the plane of theaction. The potential hindrance of suchplacement has already been noted. In addi-tion, even in a straight run it is extremelydifficult to maintain the performer in a con-stant plane. Utilization of a grid solves theseproblems in part but does enhance perspec-tive errors. Formula (2) corrects for the per-spective errors and potentially allows for theinconsistency of the performer by permittingd and s to vary by known amounts. If a gridis utilized as a background, and if the scalefactor is initially determined from an actualn calculated from known d, s, and r values,then as long as d s remains constant thescale factor (k) will remain constant. Thiswill permit the above mentioned variation ofd or s, and make possible the correct de-termination of subsequent unknown valuesof I by appropriately modifying the multi-


plier, d /(d s). For example, the originalscale factor could be determined for themiddle of the long jump runway (proposedplane of action), and correct values for Idetermined if the juniper were off to oneside, as long as the deviation from the cen-ter were known so that d could be modified.The scale factor would not change as thatrelates to the grid, but the multiplier wouldbe modified to correct n for the offset.Knowledge of the offset could be obtainedby having a series of parallel lines, of knowndistance:: apart, perpendicular to the axis ofthe camera. Two precautions should benoted: a) the scale factor should be deter-mined over the greatest lineal dimensionpossible in order to minimize any error in itsdetermination, and b) in order for the scalefactor to remain constant, as indicatedabove, the sum of d and s must remain con-stant (i.e., if one increases the other mustd.:crease by the same amount, or putting itanother way, the distance between cameraand grid must remain constant). In additionto improving the accuracy of linear measure-ments in cinematographical analysis, thesetechniques will permit filming at closer rangewhere insufficient space prevents the use ofthe often recommended telephoto lens.

The units of kinetic energy as shown inbe perpendicular to the camera axis has beenvirtually inviolable. This practice was fol-lowed throughout the present investigation.The basic logic for such doctrine may beobserved in Figure 1, where it should benoted that a would produce the same n aswould r. From the practical standpoint, theconcept of perpendicularity undoubtedlyshould be followed; however, it is theoreti-cally possible to ascertain the correct valuefor a if the procedure described below isemployed. The distances that points p and qare from the camera must be known thismight be accomplished with the aforemen-tioned series of parallel lines or with con-centric arcs of known distances in the fieldof view. Forniala 1.3), an application of theLaw of cosines, where b and c are the dis-tances p and q are from the camera, permitssolving for a:

a2 b2 c2 2bc(cosN) (3)

37

The cosine of angle N may be obtained fromtrigonometry tables, since the following re-lationships hold true:

tan N = tan R tan N' n + n'd + s

n' nd+sd+sThe value for d + s "ill be the constant dis-tance between the camera and the grid. Thevalue for n will be km, where k (the scalefactor) has been determined as describedabove and nn is the projected image valuefor a. The significant difference between thisapproach and the one verified in this in-vestigation is that after obtaining a, insteadof applying the multiplier, d/(d + s), onedivides by (d + s) to obtain the tangent ofN, and thcn knowing b and c utilizes form-ula (3). The increase in complexity of this

procedure should not be minimized, butit could be effectively cmployed.

References1. Cureton, Thomas K., Jr. Elementary principles

and techniques of cinematographic analysis asaids in athletic research. Research Quarterly10:3-24, May 1939.

2. Glassow, Ruth. Photographical and cinema-tographical research methods. Research methodsapplied to health, physical education, andrecreation (rev. ed.). Washington. D.C.: Amer-ican Association for Health, Physical Educa-tion, and Recreation. 1952. Chapter 9, pp. 204-218.

3. Hubbard. Alfred W. Photography. Researchmethods in health, physical education, andrecreation (2nd ed.). Washington, D.C.: Amer-ican Association for Health. Physical Educa-tion, and Recreation, 1959, pp. 128-47.

4. Noss. Jerome. Control of photographic per-spective in motion analysis. Journal of Health,Physical Education, Recreation 38:81-84, Sep-tember 1967.


M. MARILYN FLINT

A Differential Study of the Hip Extensor Muscles

THE GLUTEUS MAXIMUS and the three ham-string muscles, the semimembranosus, semi-tendinosus, and biceps femoris, are generallyascribed the function of thigh extension atthe hip joint. How much effort each con-tributes to the total movement of hip ex-tension and the phase during which each isworking has not been determined. There issome disagreement in the literature as tothe extent of involvement of the gluteusmaximus and the hamstrings in other primemover actions such as adduction, abduction,lateral and medial rotation.

It is a generally accepted fact that musclesmay perform differently during activity thanwhen performing prime mover activity in aclinical situation. Basmajian (1), Close (2),Eberhart (3) and Inman (6) and otherkinesiologists have found that muscles donot act strictly as agonists nor antagonistsbut rather that they work in groups and con-tract in a coordinated, associated, orderlymanner. Wheatley and Jahnke (12) deter-mined the functions of hip and thigh muscleswith the thigh held in various positions andfound that postural differences do affect theextent of activity of muscles during move-ment.

The extent of effort made by the gluteusmaximus during the performance of a fewselected activities has been investigated elec-tromyographieally. One could expect that amuscle with such a large cross-sectional areawould have a major responsibility in theperformance of fundamental movements.

M. Marilyn Flint is a member of the De-partment of Physical Education at the Uni-versity of California, Berkeley. This studywas supported by University of CaliforniaGrant #140.


Yet the evidence reveals that the gluteusmaximus has a relatively minor role in dailyroutine movements. This may explain whyit is one of the first muscles to lose its firm-ness and tone. The activities which havebeen selected for measuring the muscle potential activity of the gluteus maximus andother lower extremity muscles include walk-ing, riding a bicycle, and balance movementssuch as maintaining and assuming an up-right posture. How much the gluteus maxi-mus contributes to the performance of otherexercises and activities is questionable.

The purpose of this study was a) to com-pare the functions of the gluteus maximus,the semitendinosus, and the biceps femorisduring a wide variety of prescribed move-ments and I)) to determine the activities inwhich the gluteus maximus was most active.

METHOD AND PROCEDURE

EquipmentA Meditron* electromyograph, Model

302, was used to record the data for thisstudy. This instrument was equipped witha dual channel cathode ray oscilloscopewhich made it possible to record simultane-ously the potential activity of two separatemuscles.

A Bo lex 16 mm reflex camera equippedwith an optical beam-splitting device wassituated in such a way as to simultaneouslyrecord the traces on the oscilloscope, the

* Appreciation is extended to the Meditron Co.,A Division of Crescent Engineering and Re-search Co., 5440 North Peck Road, ElMonte, California, for use of their machine.

39

data board mounted above the oscilloscope,and the subject. An exact reading of thesubject's movements and the correspondingmuscle responses was thus obtained on apermanent record. The physical arrangementof the instruments and apparatus are de-scribed in detail in a previous article by theauthor (4).

Monopolar needle electrodes used to pickup the action potentials were placed approxi-mately in the center of the muscle belly ofeach of the muscles under observation: thegluteus maximus, semitendinosus, and longhead of the biceps femoris. The semimem-branosus was not selected for study becauseof its inaccessibility. It is to a large extentcovered by the scmitendinosus and by theadductor magnus, making an accurate place-ment of the electrodes difficult.

The sensitivity of the eleetromyographwas set at 1,000 microvolts per centimeterfor both channels and the sweep speed wasset at 20 milliseconds per centimeter. Themachine was adjusted for each muscle ofevery subject, using a contraction againstevery heavy resistance to adjust the ampli-tude to 2 cm of vertical deflection.

In order to confirm the accurate place-ment of the electrodes, the subject performeda maximum contraction of each muscleagainst resistance, the gluteus maximus inhip extension and the hamstrings in kneeflexion. Inspection, palpation and electricalactivity determined whether the correct sitehad been obtained. The pattern of thesepotential readings also served as an indexfor a very strong contraction when the re-cordings were analyzed.

MovementSeven University of California students,

five women and two men, served as sub-jects. They were instructed to perform onlythe selected exercises listed below. No re-cordings were made until the exercise wasperformed correctly. All subjects performedthe same movement in an identical, or asnearly identical manner as possible. Exer-cises which were considered more difficultwere practiced prior to the recording ses-sion. The movements were not recorded40

until the subject had relaxed to the pointwhere clectromyographie silence was ob-tained on the oscilloscope.

In some exercises moderate resistancewas applied throughout the total range ofmovement in order to place a greater workload on the muscles being used. In theexercises requiring the support of both thebody and the left leg, two tables were em-ployed. One table supported the body andthe second table, parallel and adjoining thefirst at one corner, supported the left leg insuch a manner as to allow the right leg freerange of movements.

Movements were performed in a widevariety of positions and in every possibledirection about the axis of joint rotation inorder to learn, as nearly as possible, thetotal function of the muscles. Because only adual channel instrument was used, the sub-jects performed one complete series of exer-cises testing the gluteus maximus and semi-tcndinosus. The series was then repeated totest the biceps femoris and gluteus maximus.Muscle fatigue did not appear to be a fac-tor. The subjects were very well conditionedand the exercises did not require great ef-fort. However, rest between the exerciseseries and between the exercises was used.

THE BASIC EXERCISES

1. Extension of the thigh at the hip joint.The body and left leg were supported in theprone position on tables, the right thigh flexedto approximately 60° with the right foot touch-ing the floor, knee extended. The subject ex-tended the right thigh from 60° flexion to ap-proximately 45° hyperextension and thenslowly returned to the initial position.

This same exercise was executed while keep-ing the knee flexed, and performed with andwithout resistance which was applied manuallyat the knee throughout the total range of move-ment.

2. Extension of thigh against wall pulley.While standing facing a wall pulley, the thighwas extended. A leather cuff attached to thepulley rope encircled the lower part of thigh.

3. Abduction of the thigh. The subject wasin a left side-lying position, with the legs to-gether, knee extended and thigh in a neutralextended position. The leg was then abdu-ted


at the hip, from 0° to approximately 30° andreturned to the initial position. Resistance wasapplied manually at the knee throughout theentire movement.

The exercise was repeated with the thighflexed and again with thigh held in hyperexten-sion.

4. Adduction of the thigh. The subject wasin a left side-lying position with the right thighabducted with the knee extended. The rightthigh was then adducted from 30° to 0°, oruntil it touched the left thigh. Resistance wasapplied manually at the knee throughout theentire movement. This movement was repeatedwith the thigh held in a flexed position andagain when the thigh was in hyperextension.

5. Abduction and adduction of the thigh ina standing position. (An isometric contraction.)The subject stood with the weight of the bodyon the outside borders of the feet. He thenforcibly tried to adduct the legs by attemptingto pull his feet together against the coercion ofthe floor. After several attempts he then at-tempted to draw his feet outward in abduction.

6. Isometric contraction of gluteus maximusand hamstrings. Lying in a prone position, thesubject tightened as firmly as possible the glu-teus maximus and hamstring muscles and heldfor 10 seconds.

7. Lateral and medial rotation of the thigh.The body and the left leg were supported in aposition by tables, with the right leg, knee ex-tended, supported in the neutral position by anassistant. Resistance was applied manually atthe knee against lateral and medial rotation.

8. Treadmill. Walking was performed on thelevel, and on a 30° incline. The subject walkedon a portable treadmill* which was activatedby the walking of the subject. A waist-highhand grip was used for stability. Treadmillrunning was performed only on the level.

9. Vertical jump. The subject performed onemaximum vertical jump for height utilizingfull extension of the body.

10. Low bench stepping. Ten-inch and 17-

inch benches were used. The subject steppedup on a bcnch with the right foot, extended toa standing position, and then stepped off thebench with the left foot, while the right footsupported the weight of the body.

* Hamlin Health Hiker. Hamlin Products, 2741Wingate Avenue, Akron, Ohio 44314.


11. Bicycle riding. The subject rode a sta-tionary bicycle while in a sitting position. Thethigh moved from 60° flexion at the top of thepower phase, to 15° of hyperextension at theconclusion of the power phase.

12. Pelvic tilt. Rotating the pelvis up andback (posterior tilting) until the small of theback touched the table.

13. Toe touching. From the standing posi-tion with knees in easy extension, the subj. stleaned forward to touch his toes with hishands, and then returned to the normal standingposition.

14. Body sway. The subject maintained goodbody alignment and swayed forward from theankles until he nearly lost his balance.

Treatment of dataThe cinematographic recordings were

analyzed on a 16 mm Kodoscope Analystprojector modified for stop and intervalreadings, and a Recordak microfilm reader.Amplitude and frequency of the action po-tential response were the parameters usedfor measuring the magnitude of muscle activ-ity. The electrical activity was then ratedaccording to zero, trace, mild, moderate, andstrong for each phase or division of the exer-cise or activity performed by the subjects.Arbitrary number values were awarded eachof the categories: 0 for zero reading; 10 fortrace; 20 for mild; 30 for moderate; 40 forstrong. The exercise was analyzed in de-grees if the movement took place aroundan axis of rotation as in hip extension. Eachphase of extension, for example, consisted of15' of movement. If the movement wassustained as an isometric contraction or alinear movement such as a vertical jump, theaction was broken into logical segments orphases such as tighten-release, or up-down.

Mean readings were obtained for eachmuscle for each phase of the exercise andgraphed for analysis. Subsequently, a grandmean was made from those means to deter-mine an average potentional reading for eachmuscle in each exercise. The results of thisprocedure were also graphed. Although anaverage was established for the amount ofinvolvement each muscle had in the totalexercise, a true picture of muscle activitywas not shown on this composite.

41

DEGREES OF MOVEMENT

Figure 1, Mean readings in extension, knee extended. Biceps femorisGluteus maximus . ; SemitendinosusMean readings in extension, knee flexed. Biceps femoris ; Gluteus maximum = : =;Semitendinosus ==

Negative degrees on extreme right and left of graph indicate leg is in flexion;leg is in neutral position; positive degrees, leg is in hypertension.

MOD 30

MILD 20

TRACE 10

ZERO 0

0 degrees,

F0-

-7-

=19 30 tl5 90 25 20 15 10 5 0

DEGREES OF MOVEMENT

Figure 2. Mean readings in abduction, 0° - 45°. Bicep.. femoris , Gluteus maximus ;Semitendinosus --Mean readings in adduction, 0° - 45°. Return of leg from abducted position against resistance.Biceps femoris -------; Gluteus maximus = : =; Semitendinosus = = =-_


ANALYSIS OF THE DATA AND DISCUSSION

Extension exercises* (Figure 1)The potential recordings for the three

muscles investigated followed a similargraphical pattern through the total range ofextension from 60' flexion through the 0or neutral position to 45° hyperextension.The potential activity for each of the mus-cles gradually increased from a low inten-sity at 60° flexion to a strong intensity at450 hyperextension. A reverse pattern wasevident on the lowering or final phase ofthe exercises.

The hamstring muscles begin and controlbackward extension when the movement isinitiated with the thigh in a flexed position.Not until the hip was in 5° to 10° hyper-extension did the gluteus maximus show in-volvement, which gradually increased to amoderate reading at 30° and a very strongreading at 40°. Hyperextension beyond ap-proximately 10' is known to take place inthe lumbar articulations. The high potentialactivity is very possibly the result of strongaction of the extensors to overcome the re-sistance of antagonistic structures on the an-terior side of the hip. Levine and Kabat (9)observed a similar response from the an-terior deltoid in arm elevation. In agree-ment with other studies (1, 12), the gluteusmaximus was found to be most effective asa hip extensor when the muscle workedagainst heavy or moderate resistance. Aone-lift maximum and extension against aheavy weight by use of a wall pulley in-volved the gluteus maximus earlier in thecycle of extension.

As stated earlier, the pattern of muscleactivity in prime mover activity may be dif-ferent than when performing a like move-ment in coordinated activities. This be-comes quite apparent when comparing walk-ing with the controlled exercise of extendingthe leg through its full range of movement.In walking the gluteus maximus contractsat contact when the leg is in forward flexion.

* A brief analysis of extension has been shownin a previous article (4).


Whereas, as shown in this series of exten-sion exercises, it does not work to any ex-'tent until the leg is in hyperextension.

Whether the hamstrings work best at thehip if the knee is held in extension or flexionhas frequently been questioned. Wells ( 1 1 )

for example, believes that "the effectivenessof the biceps femoris as an extensor of thehip is in reverse proportion to the degree offlexion at the knee joint." Markee and as-sociates (10), in explaining the activity oftwo-joint muscles of the thigh, utilizemorphology in addition to the theory of dis-sipation of tension within the amplitude ofmuscle contraction as Wells has done. Theyhave classified the semitendinosus as a digas-tric muscle due to an intervening tendonousinscription. Consequently, the muscle canwork at one end without involving the otherend, or as a "powerful h:p extensor irrespec-tive of whether the knee is flexed or ex-tended." Basmajian (1, p. 68) in electro-myography studies, was unable to substan-tiate the thesis presented by Markee andhas shown that a two-joint muscle operatessimultaneously in both joints. In attempt-ing to determine the function of the ham-strings at the hip joint, measurements weretaken with the knee both in passive flexionand in extension.

A comparison of the electromyogramsbetween hip extension movements when theknee was flexed and when it was extendedrevealed only a slightly variable graphicalpicture. The semitendinosus did recordslightly stronger potential activity when theknee was held in extension. The bicepsfemoris recorded slightly stronger potentialswhen the knee was in flexion. However,since the difference was minimal, the read-ings of all extension exercises were com-bined and recorded in one graph.

Abduction :xercises (Figure 2)The gluteus maximus recorded potential

activity during abduction whether the thighwas is a position of flexion, extension orhyperextension. This is in disagreement withothers. Gray, in his Anatomy, does not in-clude it as an abductor of the hip. Otheranatomists claim that only the anterior or

43

upper fibers contribute abduction. Wheatleyand Jahnke found it to abduct against heavyresistance when flexed at approximately60°. Inman (6) determined that the posi-tion of the gluteus maximus in relation tothe hip joint limited its capacity to serve asa hip abductor. Neither of the hamstringmuscles contributed appreciably to the move-ment of abduction.

Adduction exercises (Figure 2)The gluteus maximus recorded no elec-

trical activity during adduction. Wheatleyand Jahnke found some activity in the glu-teus maximus when the thigh was held in anabducted position and worked against heavyresistance. During forceful movements ofadduction, the potential readings for boththe biceps femoris and the semitendinosusranged between trace and mild.

Lateral and medial rotation exercises(Figure 3)

Whether the thigh was held in extension,flexion, or hyperextension, the biceps femoriswas found to be a strong lateral rotator andthe semitendinosus a strong medial rotator.

The gluteus maximus readings on twoof the subjects were of mild intensity duringlateral rotation. Five subjects, however, didnot record electrical activity. The gluteusmaximus recorded no electrical activity dur-ing medial rotation of the thigh for any ofthe subjects.

Isometric contraction (Figure 4)Tightening and holding the gluteus maxi-

mus and hamstring muscle recorded mod-erate to very strong readings for the gluteusmaximus whether the subject was in theprone, supine, or sitting position. The ham-string recordings during this exercise showedconsiderable variation between subjects. Thiswas possibly due to the difficulty in isolatingthe hamstrings for an isometric contraction.

Pelvic tilt (Figure 3)The pelvic tilt can be performed without

activating the hip extensors although thmean recordings show trace plus activity.The amount of effort the subject put intoholding the position determined the po-

44

tential activity of the muscles. (Two sub-jects recorded moderate to strong. Two weretrace, plus, and two were zero.)

Treadmill walking and running (Figure 5)The graphic pattern of potential activity

for the three muscles under investigationduring running and walking are quite similar.The least amount of activity was found tohe just before and during push-off.

The gluteus maximus makes a negligiblecontribution to the whole walking pattern.Its trace-to-mild potential activity is recordedimmediately after the heel contacts theground for one-third of the supportingphase. During the run it seems to be in-volved over a longer part of the contactphase.

The semitendinosus and biceps femorisbecome active in the final third of the swing-ing phase of the walk and remain activethrough the first third of the stance phase.During the run, the semitendinosus recordspotential activity throughout most of theswinging phase. In general the data sup-port findings of other researchers who havestudied the walk. Eberhart, and associateshave recorded more extensive muscular in-volvement of the walk.

Vertical jump for maximum height(Figure 3)

Potential activity was recorded betweenmild and moderate for the gluteus maximusduring the preparatory and push-off phasesof the vertical jump for height. During ex-tension and contact only slight deflection ofthe potentials was evident. After contact,when the body began righting itself to thestanding position, trace to mild readingswere recorded. A somewhat opposite pat-tern was recorded for the hamstrings. Littleactivity was seen during the push-off phase.Just before contact mild, plus was recordedby the semitendinosus. The electrical activityto the regaining of the standing position waslimited but similar to that of the gluteusmaximus.

Bench stepping exercise (Figure 6)On the step up, as the body was raised

upward, the gluteus maximus recordings


MOO 30

MILD 20

TRACE

ZERO 0 I I8 9

GRAND MEANS

Figure 3. Total movement patterns for all subjects.Semitendinosus iS Biceps lemoris I Gluteus maximus

1. Abduction 4. Lateral rotation 7. Bicycle2. Adduction 5. Medial rotation 8. Treadmill3. Extension 6. Pelvic tilt 9. Vertical jump

MOD 30

MILD 10

TRACE 10

11 13 1GRAND MEANS

Figure 4. Total movement patterns for all subjects.Semitendinosus

10. Stepping up11. Toe touch12. Lean forward

15 IS 47

Biceps temoris Gluteus maximus

13. Prone isometric14. Supine isometric15. Adduction isurnetric

16. Abduction isometric17. Extension pulley resistance


MoD 30

MILD 20

Q

0zTRACE P

ZERO 0

SUPPORTING PHASE

Figure 5. Mean readings in treadmill walking. Biceps femorisSemitendinosus --Mean readings in treadmill running. Biceps femoris Gluteus maximusSemitendinosus = = =C Contact; BT, MT, FT - Beginning, middle, final third; PO - Push-off

SWINGING PHASE

; Gluteus maximus . ;

MILD

'MALI

ZERO 0

-./

2 3 CONTACTFOOT LIFT EXTENSION

MOVEMENTSTEP OFF

3

Figure 6. Mean readings in bench step. Biceps femoris ; Gluteus maximus . ;Semitendinosus --Movement in each phase, foot lift, extension, and step-off was broken into three parts to showgradual change in hip action (slow motion filming). During step-off phase, weight supportedon test leg.


were rated as mild. There was no potentialactivity before this during the actual prepara-tion phase nor at moment of contact with thestep. On the stepping down phase, no activ-ity was recorded until the foot contacted thefloor at which time a trace, plus potentialrecording was made. The hamstrings showa somewhat similar pattern with additionalrecordings made during times when the kneewas flexed. In this exercise as in some of theother exercises, a difference in ;...;tivity pat-tern was found between the two hamstringmuscles.

Bicycle riding exercise (Figure 7)Limited potential activity was observed

for the three muscles miring the bicyclingexercise. The gluteus maximus was mildlyactive during the power phase as the footbegan pushing down on the pedal. (Thighapproximately 60° from vertical extensionof 0°.) This activity appeared to last onlya few degrees in the downward cycle andthen ceased. The hamstrings showed slightinvolvement throughout most of the cycleof movement. Phases of inactivity for thetwo hamstring muscles were not synchro-nous.

MILD 20

TRACE r0

ZERO 00 -IS -90 -45

Houtz and Fisher (5) found the gluteusmaximus to be completely inactive through-out the movement and the hamstrings tomake a negligible contribution as the footwas underneath the body. Bicycling requiresprimarily hip flexors, knee extensors, andfoot and ankle muscles.

Balance activities (Figure 4)Toe touching and body-sway-forward

were exercises used to determine the effec-tiveness of the three hin extensors in re-gaining balance. Only the recordings for onesubject were available for these tests. Whenthe subject swayed forward from a standingposition no activity was recorded for thegluteus maximus. The semitelidinosus didhave a reading of trace, plus to mild whenthe body nearly lost its balance forward.Joseph (8) found a similar response fromthese muscles. He found no activity l'omthe gluteus maximus but considerable po-tential activity was recorded for the ham-strings. Jonsson (7) found no gluteus maxi-mus activity in the body sway.

During the standing and toe touchingphases of the toe touching exercise, the elec-trical activity was zero for the three muscles.

UPWARD CYCLE.-60 -75 $0 75 60

DEGREES OF MOVEMENT

45 30 15

DOWNWARD C1

Figure 7. Mean readings in bicycle riding, sitting. Biceps femoris --; Gluteus maximus . ;Semitendinosus

At 0° thigh is in extension; 90° thigh is in flexion.


During the actual movements of bending ning, bicycling, standing, and the mainte-forward and straightening up trace activity mince of balance.was recorded by all three muscles. 5. Variability in potential activity of a

muscle during various phases of an activityindicates the independence of muscle actionand emphasizes the individual responsibility

SUMMARY AND CONCLUSIONS of muscles in the execution of coordinatedmovements.

Eleciromyographic recordings of the glu-teus maximus, semitendinous, and bicepsfemoris were recorded simultaneously withmotion picture studies of subjects perform-ing a variety of movements involving the hipjoint. An evaluat:on of the differential activ- 3.ity of the hams:rings and gluteus maximuswas made as well as a graphical representa-tion of the muscle action during the variousphases of the movements.

The evidence presented supports the fol-lowing conclusions:

1. During exercise, the hamstrings initiateand control backward extension of the thighwhen forward of the neutral position, andadduct the thigh. The biceps femoris is astrong lateral rotator of the thigh and thesemitendinosus is a strong medial rotator.Neither muscle is an abductor.

2. The glutei's maximus works most ef-ficiently as a thigh extensor when workingagainst resistance and when the thigh movesbeyond neutral position into hyperextension.

3. The gluteus maximus abducts the thighwhen held in any position, may contributeslightly to lateral rotation, but is neither anadductor nor medial rotator of the thigh.

4. The activities selected for this studywhich involve the gluteus maximus to thegreatest extent are: stair stepping, verticaljump, isometric contraction, and hyper-extension (particularly against resistance).Those requiring minimal effort on the partof the gluteus maximus are walking, run-

2.

4.

5.

6.

7.

8.

9.

10.

12.

ReferencesBasmajian, J. V. Muscles alive. Baltimore:Williams and Wilkins Co., 1962.Close, J. R. Motor functioning in the lowerextremity. Springfield, Ill.: Charles C. Thomas,1964.Eherhart, H. D., and oTHIRS. Fundamentalstudies on human locomotion and other in-formation relating to design of artificial limbs.A report to the National Research Council,Committee on Artificial Limbs. University ofC: ,rnia, Berkeley, 1947.Huh,. M. Marilyn. Photo-recording electro-myography in clinical kinesiological studies.Biomechanics 1, First International Seminar,Zurich, 1967. Basel/New York: S. Karger,1968, pp. 132-37.Houtz. S. J., and Fischer, F. J. An analysis ofmuscle action and joint excursion during exer-cise on a stationary bicycle. Journal of Boneand Joint Surgery 41-.t: 123-31, 1959.'nman, Verne T. Functional aspects of theabductor muscles of the hip. Journal of Boneand Joint Surgery 29:607-619, 1947.Jonsson, Bengt, and Synnerstad, Bjorn. Elec-tromyographic studies of muscle function instanding. Acta Morphologica Neerlando-Scandinavica. 6:361-70, 1966.Joseph, J. Man's posture-clectromyographicstudies. Springfield, Ill.: Charles C. Thomas,1960.Levine, Milton G., and Kabot, Herman. Dy-namics of normal voluntary motion in man.Permanente Foundation Medical Bulletin 10:212-36, August 1952.Markee, Joseph E., and OTHERS. Two jointmuscles of the thigh. Journal of Bone andJoint Surgery 37-A:125-42, 1955.Wells, Katherine F. Kinesiology. Philadelphia:W. B. Saunders Co., 1966. p. 315.Wheatley. Max D.. and Jahnke. William D.Electromyography study of the superficialthigh and hip muscles in normal individuals.Archives of Physical Medicine 32:508-15,1951.


GLADYS E. GARRETTCAROL J. WIDULE

Kinetic Energy: A Measure of Movement Ind:vkluality

AS A CIIILD MATURES, his movement be-havior undergoes both quantitati,e andqualitative changes. While these changesmay be detected by the trained eye, quanti-fication is necessary for future reference andcomparison. Previous investigators of themechanical characteristics of human motionhave used measures of the body and bodysegments such as displacement, velocity, in-clination, and slope (2, 7, 8, 10). A meas-ure of human motion less frequently used iskinetic energy, which Eckert (5) claims isa next step in improving the accuracy ofanalyses in the study of human movement.One of the parameters necessary in kineticenergy ;:nalysis is the moment of inertia ofeach body paw Experimental methods fordetermining moments of inertia have beendescribed in the literature (1, 3, 4, 9).

To a scientist, energy expenditure issynonymous with motion. Since by definitiona moving body is said to posses energy ofmotion, or kinetic energy, kinetic energywould appear to be an appropriate measureof movement of the human body. This studywas undertaken to explore the possibility ofusing kinetic energy as a measure capableof distinguishing individual differences inperformance.

Gladys E. Garrett is an assistant professorand Carol J. Widule an associate professorat Purdue University, Lafayette, Indiana.The study is based on a Ph.D. dissertationof G. Garrett, under the supervision of C.Widule, major professor, January 1970.


Kinetic energy analysis

Any body that is rotating about an axisnot passing through its center of gravity mayhe considered to have a motion of translationand a motion of rotation about its center ofgravity combined. The body shown in Fig-ure 1, which may represent a human limbor segment, may be considered to be madeup of a translation with the velocity Vcg ofthe center of gravity and a rotation aboutthe center of gravity.

The kinetic energy of the body can berepresented by the following expression:

KE; = 1/2 Mi Vcgi2 + 1/2 Icg; (142where the subscript, i, represents the th

segment of a system, and where

M = mass of segment in units oflb-sec2

inVcg = velocity of the center of gravity of

insegment in units of sec

Icg = mass moment of inertia of the seg-ment about its center of gravity in

in-lb-sec2units of rad2

= angular velocity of the segment in

units of radsec

The units of kinetic energy as shown inthe following check of units are in-lbs.

KE = 1/2 M Vcg2 + 1/2 Icg (02lb-sec2 in2 in-lb-sec2

inX sec2 rad2

in-lbs

rad2X sec2

49

Figure 1. Notation of ith segment.

Since each segment of the subject's bodyis represented by its center of gravity (whichis assumed to he the center of the distributedmass), it was first necessary to compute thex and v coordinates of the segment center ofgravity. This was accomplished by multi-plying the difference between the respectivecoordinates by the center of gravity propor-tionality factor (car) as obtained fromDempster (4).

For example, the coordinates of the centerof gravity of segment number three, whichis bounded by joint numbers three and four,would be obtained by

CGX3 = CGP3 (X1 X3) + X3and CGY3 = CGP3 (Y., Y:s) Y3.

The sources of other information for thenecessary inputs to the kinetic energy equa-tion were as follows:

1. M (mass) = wg

acceleration of gravityThe weights of the segments were ob-

tained by multiplying the segmental weightproportions of Braun and Fischer (11) bythe subject's total body weight. The accelera-tion of gravity is 386.4 in/see.

2. V (velocity)The velocities of the centers of gravity

were obtained by numerically differentiatingthe center of gravity displacement informa-tion with the aid of a finite difference tech-nique. Specifically, the second central dif-ference expressions for velocity componentswere used. This expression tends to give

weight of segment

50

more accurate results for experimental data,since it takes a look at displacements up totwo stations to the right and two stations tothe left of each position being analyzed.Vcgx, = [--CGXJ 1 2 + 8 (CGX,--1)

8(CGX,_,) CGX, -2]/I2DTVcgy, = [CM' 2 -f- 8 (CGY,-1)

8 (CGY;-- ) CGY, --:1/12DTwhere DT is the time increment betweenpositions.

The total velocity for the center of gravityat that position is given by

Vcgj = I(Vcgx,)2 (Vegy,)2

3. io (angular velocity)The angular position, 0, of each segment

at each position is given by the trigonometricrelationship

Yi Y3tan 0, =X3

for the case of segment number three.The angular velocity of segment three is

then obtained by03(at station j) 03(at station j-1)

(113

DT4. I (moments of inertia of body seg-

ments)The units of the moments of incrtia of

body segments in Dempster were found tobe weight moments of inertia (gm crn2). Toget mass moments of inertia these units hadto he divided by g(980 cm/sec2). Therefore

gm crn2 sec2 gm cm sec2980 cm 980

which was found to be analogous to lb-insee in U. S. measure.

In order to get the units in lb-in see thefollowing formula was used with Dempster'svalues (DV):Given 1 gm = .0022 lbs

1 cm = .3937 inTherefore

DV X .0022 lbs X .3937 in sec2980

=_ U. S. MeasureDV X 2.2 X 10-3 X .3937

980 X .!03= U. S. Measure

DV .844 X 10-6= U. S. Measure


Since Dempster's values are in the formgm cm2 X 106, they were used directly(without the 106) and multiplied by .884.*

The preceding analytics provided the basisfor a computer program which subsequentlyresulted in the calculation of the segmentalkinetic energies as well as the total bodykinetic energy of the subject at the threeages.

Data collection proceduresA film record of a boy subject executing

a jump from a 12-inch bench was obtainedat the ages of two, three, and four years.From the image projected by a GriscombeMicrofilm Reader, the rectangular coordi-nates were obtained for each of the follow-

ing locations on the body: toes, ankles,knees, hips, shoulders, top of head. elbows,wrists, and fingers. The coordinates were ob-tained from frames selected at intervals ofapproximately .05 seconds, from approxi-mately .25 seconds prior to take-off to themoment of landingl

The segmental and total body's centersof gravity were calculated from the basicgeometric data obtained from the film rec-ords, and the discrete data points weremodeled using the finite difference technique.From the modeled data, corresponding mo-ments in time could be determined betweenthe intial frame and final frame, for the threeages. Kinetic energy values were derivedfrom the modeled data.

TABLE 1. VELOCITY AND KINETIC ENERGYOF THE LEFT I'OOT FOR EACH FRAME

Positon Segment End 1 End 2 Vel CG Omega K. E.3 2 3.981 -5.437 .3544 2 6.825 .795 .0485 2 9.817 -2.399 .1516 2 6.786 -.047 .0407 2 7.210 .289 .0478 2 7.568 -3.622 .2019 11.722 -2.220 .1770 I 3.572 3.478 .150I 27.176 2.723 .734I 47.492 -13.199 3.9823 I 33.612 -6.431 1.4674 I 24.155 -.829 .5205 I 47.073 .979 1.9566 I 58.937 1.609 3.0797 I 81.196 -4.323 6.0028 2 97.566 7.963 9.0859 2 80.848 9.331 6.738

20 2 64.016 4.879 3.87121 2 69.964 -1.804 4.33422 2 96.686 -.445 8.20923 2 77.298 1.633 5.27624 2 29.018 4.155 .93825 2 6.061 -5.965 .44126 2 6.985 2.762 .13127 2 16.758 -3.221 .36628 .250 -1.490 .02629 2 11.050 .293 .108

*See Garrett (6:68) for corrected moment ofinertia values.

1: For details related to the photographic situa-tion, the film speeds, and linear conversionfactors, see Garrett (6).


01

TA

BLE

2. E

NE

RG

Y D

IST

RIB

UT

ION

IN T

HE

SE

GM

EN

TS

ON

PE

RC

EN

TA

GE

BA

SIS

- A

GE

TW

O

LFoo

tLL

egLT

hiR

Thi

RLe

gR

Foo

tLH

and

L4A

rmLU

Arm

RU

Arm

RU

Arm

RH

and

Tru

nk T

otal

K E

.

32.6

48.

844.

785.

961.

52.2

9.8

84.

317.

9728

.93

4.92

5.17

23.8

013

.422

4.4

62.

253.

0429

.92

.08

.68

.43

3.17

3.75

6.86

7.59

6.55

35.0

110

.519

51.

362.

08.5

95.

59.8

9.2

7.5

43.

582.

8930

.68

21.4

514

.49

15.4

811

.103

6.4

82.

4012

.29

14.3

63.

494-

;'21.

351.

387.

7211

.93

17.6

025

.40

1.28

8.37

47

.08

.55

15.0

26.

59.6

3.0

2..0

6.5

1.6

32.

262.

014.

5867

.05

56.3

29

8.5

11.

5815

.52

2.94

.17

.03

.83

2.02

4.48

4.95

5.84

8.34

52.8

039

.355

9.7

41.

895.

172.

48.1

2.0

72.

514.

9010

.63

8.82

8.98

13.9

939

.90

23.9

00

10.6

219

.08

12.1

91.

62.6

2.1

35.

819.

636.

232.

376.

1011

.89

23.7

024

.338

111.

815.

3811

.56

4.15

1.04

.30

4.48

10.5

99.

393.

585.

156.

8435

.70

40.5

7412

6.28

22.8

48.

897.

70.7

1.1

53.

258.

345.

755.

573.

604.

3922

.53

53.3

3113

5.09

7.21

7.17

6.25

6.77

.37

6.13

13.7

36.

044.

074.

126.

9725

.97

28.8

2014

.87

2.57

9.57

6.91

1.25

.16

2.20

5.11

8.32

.25

1.12

2.77

58.9

369

.977

151.

854.

1614

.56

11.4

12.

51.7

61.

153.

501.

671.

331.

23.9

254

.86

105.

793

163.

369.

277.

788.

386.

23.5

7.0

6.9

23.

25.8

51.

16.5

757

.49

91.6

6317

5.87

7.85

3.06

17.8

86.

432.

38.6

01.

804.

174.

242.

582.

4340

.71

102.

323

185.

379.

025.

939.

233.

612.

29.6

82.

182.

243.

634.

094.

4947

.24

169.

071

193.

528.

019.

276.

456.

724.

891.

313.

934.

274.

765.

269.

0432

.57

191.

604

201.

522.

855.

216.

938.

765.

021.

685.

394.

6512

.85

7.26

9.44

28.4

425

4.26

821

1.09

1.95

4.76

6.19

5.89

2.52

1.03

2.83

3.15

37.2

34.

625.

7923

.15

398.

054

222.

495.

5310

.90

10.9

16.

182.

87.8

42.

783.

003.

574.

305.

9640

.67

330.

169

231.

164.

2112

.37

13.5

34.

691.

26.5

02.

183.

403.

343.

323.

4746

.57

456.

619

24.3

55.

649.

047.

914.

03.6

0.6

22.

794.

034.

495.

264.

5050

.76

270.

541

25.2

2.2

418

.59

23.0

71.

11.1

41.

734.

804.

535.

753.

943.

7632

.12

200.

742

26.0

6.3

31.

00.7

8.6

0.0

21.

283.

927.

865.

202.

853.

2872

.82

202.

500

27.8

63.

0110

.87

8.22

2.23

.72

2.05

7.49

1.92

1.54

4.36

4.99

51.7

542

.332

28.0

3.3

82.

541.

70.7

8.1

6.7

51.

30.5

22.

611.

281.

6886

.26

83.3

5329

.07

.46

5.98

2.51

.74

.06

.36

1.95

1.92

3.74

3.81

2.80

75.6

015

7.05

6

Results and discussion

Table 1 is an example of the results ob-tained from the computer program of thelinear velocity of the center of gravity of asegment. the angular velocity of the seg-ment. and the kinetic energy of the segmentfor each frame used in the analysis (with theexception of the first two and last two frameswhich are "lost" when using the second cen-tral difference expression to calculate veloc-ity). A similar page was obtained for eachsegment for the three aces. For the sake ofbrevity, only one page has been included.

The relative increase and decrease inkinetic energy of the segment can be notedin these records. Fluctuations in the kineticenergy values are probably due to errorstill present in the modeled data.

Table 2 is an example of the results of thetotal kinetic energy of the body and of thepercent of kinetic energy of the segments inrelation to the total for each frame for agetwo. For example, the 16th frame, which isthe take-off frame for the two-year-old, dis-closes that about three percent of the totalkinetic energy at this moment can be at-tributed to the left foot. about nine percentto the left leg, about seven percent to theleft thigh. etc. It roust he remembered thatthese are merely percentages of the totalkinetic energy and not absolute values inthemselves. Results were also obtained forages three and four but have not been in-cluded.

A study by Glassow, and others (7) in-dicated that greater angular velocities of thethighs and trunk at take-off result in betterjumping performance. The results of thisanalysis substantiate the finditigs of Glassowas far as the thighs are concerned, but donot concerning the trunk. In Table 3 are theangular velocities of the thighs and trunk

TABLE 3. COMPARISON OF ANGULARVELOCITIES (rad/sec)

Age L.eft Thigh Right Thigh Trunk

2 3.0 0 2.63 8.7 5.04 5.0 6.5 1.7

at take-off at ages two, three, and four. Thekinetic energy developed by the thighs attake-off is given in Table 4.

A plot of total body kinetic energy versustime for the three ages studied can he seenin Figure 2. The curves were so positionedthat take-off occurred at a common instantin lime in order to allow for some relatedobservations. It can definitely be said thatthe energy level increased with age. Sincethe same moment of inertia of the totalbody is used for the three ages, and sincethe body weight increased only slightlyfrom 32 to 36 pounds the increase inenergy is directly attributable to the seg-mental speed developed. It is interesting tonote that the kinetic energy has reached apeak at take-off at the ages of three and four,where the performance was better. Also, thesmoothness of the kinetic energy curve atage four would appear to be a result of asmoother, more totally integrated jump.It was also found that in the jump at this agethe upper limbs constituted a larger percent-age distribution than did the lower limbsbefore take-off. Upon take-off, however, adefinite reversal of this situation was noted.Does this mean that the upper limbs playan important role in developing the energynecessary for a skillful take-off? Answersto questions such as this will not be knownuntil more extensive investigations are per-formed on more subjects. Hopefully, theanalysis techniques presented here will pro-vide a vehicle for pursuing the answers toquestions of this type.

Because a change in the kinetic energywas reflected as the child matured in hisjumping pattern, kinetic energy shows prom-ise as a measure of movement individuality.The results of the study warrant additionalresearch of the energy characteristics of hu-man motion.

TABLE 4. COMPARISON OF THEKINETIC ENERGY OF THE THIGHS

Age Left Thigh Right Thigh

7 183 38 374 48 55


800

700

Total Body Kinetic Energy

600 vs

Time

500

1) 400

L.)

300

z

200

100

0

2

4 year old

3 year old 'S2 year old

Take-of f

4--

8 10 12 14 16 18 20 22 24

Time Increment (.015 sec. approximately)

Figure 2. Total body kinetic energy for each frame at ages two, three, and four.

References1. Bouisset, S., and Pertuzon, E. Experimental

determination of the moment of inertia oflimb segments. In J. Wartenweiler; E. Jokl;and M. H. Hebbelinck (Eds.), Biomechanics11. Basel, Switzerland: S. Karger, 1968, pp.106-109.

2. Cooper, John M., and Glassow, Ruth B.Kinesiology. St. Louis: C. V. Mosby, 1968.

3. Damon, Albert; Stoudt, Howard W.; and Mc-Farland, Ross A. The human body in equip-ment design. Cambridge, Mass.: Harvard Uni-versity Press, 1966.

4. Dempster, W. T. Space requirements of theseated operator. WADC Technical Report 55-159, Aero Medical Laboratory, Wright AirDevelopment Center, Wright-Patterson AirForce Base, Ohio, 1955.

5. Eckert, Helen. A. concept of force-energy inhuman movement. Journal of American Phys-ical Therapy Association 45:213-18, 1965.

6. Garrett, Gladys E. Methodology for assessingmovement individuality: A computerized ap-proach. Doctoral dissertation, Purdue Uni-versity, 1970.

bi 28 30

7, Glassow, Ruth B.; Halverson, Lolas E.; andRarick, G. Lawrence. Improvement of motordevelopment and physical fitness in elemen-tary school children. Cooperative Research,Project No. 696, University of Wisconsin,Madison.

8. Hellebrandt, F. A., and others. Physiologicalanalysis of basic motor skills. I. Growth anddevelopment of jumping. American Journalof Physical Medicine 40:14-25, 1961.

9. Holowenko, A. R. Dynamics of machinery.New York: John Wiley and Sons, 1955.

10. Kessen, William; Hendry, Louise S.; andLeutzendorff, Anne-Marie. Measurement ofmovement and the human newborn: A newtechnique. Child Development 32:95-105, 194.194.

11. Woods, Bryan Duncan. Validity of the seg-mental method of assessing the position of thecenter of gravity of the body in the supineposition. Master's thesis, University of Ore-gon, Eugene, 1967.


SHIRL JAMES HOFFMANKENNETH WORSHAM

ANGLCALC: A Computer Program for Descriptive

Analysis of Movements Used During RepeatedPerformances of Skills

THE OBSTACLES TO THE CONDUCT of com-prehensive cinematographic movement anal-yses have been sufficient to discourage re-searchers from filming subjects over a seriesof repeated performances of a skill. Ratherthan recording longitudinal data on subjects,kinesiologist-cinematographers have found itmore feasible to photograph each individualas he performs a limited number of trialsand to consider this small sampling of be-havior representative of the subject's move-ment characteristics. Generally, a large num-ber of filmed trials for each performer is sacri-ficed so that a greater number of subjects canbe studied.

In selecting only a few trials for analysis,the experimenter ignores the likely possibil-ity that individual performance styles varymarkedly from trial to trial. A representa-tive performance style may be attributedmore accurctely to an individual if the de-scription is based on observations madeover a mulitude of trials rather than one ortwo isolated attempts.

Shirl James Hoffman is an assistant pro-fessor in the Department of Physical Edu-cation at the University of Nebraska atOmaha. Kenneth Worsham works at theComputer Center at the same university.


A serious consequence of the kinesiol-ogist's hesitancy to conduct longitudinal in-vestigations has been the retardation ofmeaningful integration of kinesiological andmotor learning research. Rather than study-ing relationships between movements andlearning, kinesiologists have focused primaryattention on problems requiring the applica-tion of Newtonian laws to human move-ment in an attempt to understand more fullythe mechanical bases for efficient perform-ance in specialized athletic tasks. Conse-quently, knowledge is lacking concerning theeffects of practice on component movementsof gross motor tasks, the stability of individ-ual performance techniques over repeatedtrials, or the nature and order of changes inmovements that occur during learning ofmotor skills.

The primary deterrent to longitudinalcinematographic studies has been the volumeof physical work represented by the process.Even where a small number of measurementsare to be recorded for a few subjects, thedata extracted multiply in direct proportionto the number of trials included in the ex-periment. For example, when informationconcerning the position for five subjects inthree different frames on five consecutivetrials is desired, 75 separate frames must beanalyzed. If the subjects are to execute 10

55

trials the number of frames to be analyzeddoubles to 150. The investigator who naivelytackles such a problem soon realizes hehas created a project of monstrous pro-portions, while the sample size and numberof trial repetitions are still embarrassinglysmall.

Tn order to make feasible the conduct oflongitudinal studies, the computer program

ANGLCALC has been written. Thisprogram expedites collection of data fromfilm records while eliminating much of themeasurement error associated with manualanalyses. ANGLCALC provides a descrip-tive rather than mechanical account of themovements used by subjects during repeatedexecutions of a skill. Angles are calculatedfor the hip, knee, ankle, and metatarsal-phalangcal joints and for the inclination(with the horizontal) of the trunk, thigh,leg, and foot. The angles of inclination arecalculated to the right of the segment, asillustrated in Figure 1. The program is de-signed to take measurements on each of theeight variables from three separate frameson each trial. Any number of subjects (inmultiples of 10) may execute any numberof trials (in multiples of 10).

For convenience, the trials are dividedinto three sets of 10 cach and labelled: Ses-sion 1, Session II, etc. Subroutines calculatethe mean measurement on the variables foreach subject on the first 5, last 5, and entire10 trials in each session. In addition, thestandard deviation of the measurements foreach subject is calculated, as is the variancebetween the subjcct means for each set of10 trials. Blocking the trials into sessionsfacilitates statistical comparison of subjectperformances on carly and Iate trials.

The method of collecting the data for in-put is similar to the technique suggested byRushall and Pyke (2). A laminated coor-dinate grid 15" x 15" and calibrated alongX and Y axes at 40 parts to the inch ismounted on the viewing screen of a Re-cordak P-40. Locations of relevant anatomi-cal landmarks are plottcd directly on thegrid surface with a fine-tip ink indicator.Coordinate readings for the landmarks aretranscribed on data sheets for key punching.The ink marks then are erased from the56

Figure 1. Angles of inclination of body segmentswith the horizontal as calculated with ANGLCALC.a) Trunk, b) leg, c) foot, d) thigh.


CARD "I

CARD 2

CARD 3

/

4LL

/2

3

20

CO4

0-IX0I

/79

2/1

,290

a0

I

/97

284

279

aI

202

223

275

aI

129

193

189

X

z

149

189

257

W}

2

/09

/97

/34

w X

z4

/80

/79

306

W}

24

85

99

a

169

162

307

35

70

79

0

159

/50

300

33

69

r w ww w w

UwUU U

2 2w w2 2

w wCO X a r x rw w w wa a m

aw

0w w w

IZ

200

1%

/96

356 497 350

386 494 380

399'14192 387

Figure 2. Format for data input. Coordinates for each frame are punched on separate cards. Data inillustration are for the backward standing jump for distance. Frame 1 occurred at .15 seconds priorto take-off; Frame 2 occurred at take-off, and Frame 3 at landing.

laminated grid surface With a moist cloth.The process of transcribing the data can hespeeded up considerably if an assistant is

available to record the data as it is readaloud by the investigator. Using the teamapproach, the data can he collected and re-corded in reliable fashion at a rate of ap-proximately one minute per frame of film.

Data is input as subject, trial, frame, andsession followed by X and Y coordinatesfor the shoulder, hip, knee. ankle, meta-tarsal-phalangeal joint, and tip of toc. Co-ordinates for all anatomical landmarks andcoordinates for two constant backgroundreference points which lie on a horizontalplane in the visual field also must be inputfor each frame (Figure 2 illustrates the for-mat used for inputting the data). The Ter-ence points are used to calculate a correctionfactor for slight rotation of the projectedimage on the grid surface.

The output is arrayed in tables where thedata for all 10 subjects for an entire blockof 10 trials are displayed for each variableunder examination. Figure 3 illustrates adisplay of sample output from data taken on


the angle at the hip and angle of inclinationof the trunk for 10 subjects during one frameon the first 10 of 30 trials in the standingbroad jump. In an experiment where 10 sub-jects execute 30 jumps and measurementson all eight variables arc collected for threeframes on each trial, a total of 72 such tablesis printed. If the camera speed is known,average angular velocities can be computedby calculating the difference in joint posi-tions between relevant frames and multiply-ing by the appropriate reciprocal.

Currently, ANGLCALC is being used atthe University of Nebraska at Omaha to in-vestigate the effects of practice on alterationsin component movements of motor tasks.The subroutines for calculating the stand-ard deviation of individual measurementsand variance between subject means hasbeen particularly useful for investigatingproblems related to intra- and inter-individ-ual variability in movement. While the pro-gram has been used primarily for analyzingmovements used in jumping, it appearsreadily adaptable for analysis ol any skillswhich consist of uniplanar .motions.

57

C71 oo

AMIE AT NIP

SESSION NO 1

PPIL^IE 2

TRIAL

12

34

56

78

910

V-1-5

yv-6-0

mN-1-0

STD DEV

S. 0

147.82

131.58

150.50

131.0

122.54

148.86

131.61

116.05

130.P6

122.65

136.69

130.01

133.35

11.3702

s= 1

128.05

130.82

130.01

141.75

124.04

140.73

130.62

108.25

131.64

109.19

130.93

124.09

127.51

10.6798

S= 2

130.78

138.30

136.01

136.10

136.10

141.00

135.15

133.85

137.94

135.24

134,67

136.44

135.65

2.83Po

S= 3

135.05

108.68

94060

109.19

114.89

138.87

123.55

113.99

122.27

120.79

112,48

123.89

118.19

12.3305

S= 4

138.54

523.78

126.30

119.01

108.60

113.01

122.52

105.60

12p.08

97.64

123.25

112.17

117.71

11.1487

s= 5

120.83

123.05

126.71

110.901

129.87

119.34

112.23

120.23

132.10

116.77

122.29

120.13

121.12

6.6081

s= 6

149.64

87.45

-.L.99

144.99

148.33

137.42

153.63

154.65

145.97

141.15

137.48

146.56

142.02

19.0801

s= 7

136.46

135.17

144.93

138.03

116.46

134.81

14409

135,93

148.96

145.63

134.21

141.89

138.05

8.6728

S= 8

145.12

147.05

131.93

137.63

142 04

127.32

43.47

143.38

130.16

140.95

140.75

117.06

128.90

29.1678

s= 9

160.46

127.07

133.76

124.40

119.32

133.27

130.65

149.36

124.19

114.90

133,00

130.47

131.74

13.0886

INCLNTN ANGLE AT HIP

SESSION NO 1

FRAtZ 2

TRIAL=

12

34

56

78

910

v-1-5

NN-6-0

V21-1-0

STD DEV

S. 0

103.42

109.65

102.32

111.56

121.33

100.81

107.43

113.77

113.89

112.04

109.66

109.59

109.62

5.9681

s= 1

114.62

112.62

112.28

106.77

111.54

103.88

110.16

128.77

110.64

121.22

111.57

114.93

113.25

6.7571

s= 2

104.71

107.43

104.74

103.28

110.34

93.69

103.89

111.37

100.58

104.65

106.10

102.84

104.47

4.7339

S= 3

101.77

120.82

135.00

118.41

119.36

99.33

116.94

114.38

114.04

111.8o

119.07

111.30

115.19

9.4421

s= 4

98.04

103.24

104.74

112.04

113.11

111.32

114.86

119.40

117.45

126.03

106.23

117.81

112.02

7.8452

Ss 5

107.62

103.31

113.67

118,72

104.60

110.43

109.41

115.02

109.21

108.23

109.58

110.47

110.03

4.4567

S= 6

99.46

155.56

96.79

102.42

98.6o

101.96

91.93

85.03

90.60

98.78

110.54

93.58

102.06

18.5629

s= 7

111.37

109.12

104.74

109.80

124.29

109.52

108.43

109.29

100.71

97.37

111.86

105.07

108.46

6.7803

s= 8

107.35

103.36

115.24

106.63

104.38

111.56

100.24

101.71

109.52

106.57

107.39

105.92

106.66

4.3356

S. 9

94.29

118.97

106.99

108.92

111.25

112.46

109.60

104.42

116.57

121.12

108.08

112.83

110.46

7.3555

Fig

ure

3. A

rray

of d

ata

in o

utpu

t for

AN

GLC

ALC

. "IN

CLN

TN

AT

HIP

" re

fers

to a

ngle

of i

nclin

atio

n of

trun

k w

ith th

e ho

rizon

tal.

(Tab

ulat

ion

of th

e va

rianc

e be

twee

n th

e m

eans

has

bee

n om

itted

from

the

illus

trat

ion.

)

Calculation of the angle of slope of majorbody segments is an especially valuable func-tion. For some time kinesiologists have rec-ognized that joint angles alone do not depictthe relative positions of the body segmentsduring movement. In an early study of thestanding broad jump, Johnson first pointedout that motion at any joint alters the rela-tive position of all body segments above thejoint ( I ). If an analytical strategy is to fur-nish a complete account of the movementsof the body through space it should indicatethe position of body segments with respectto the horizontal as well as the anglesformed at joints.

With minor modifications, the use ofANGLCALC can be extended to under-graduate and graduate courses of instructionin kincsiology. It is becoming increasinglycommon for kinesiology instructors to re-quire their students to conduct elementarycinematographic analyses as course projects.The unglamorous task of tracing and mea-suring angles frequently makes such assign-

ments exercises in perseverance rather thanmeaningful studies in movement. WithANGLCALC the sophistication and scopeof course projects can be increased whilethe student will have available more time forinterpretation and application of his findings.

NOTE: ANGLCALC was written in Fortran IVlanguap specifically for the NCR 315 RMCcomputer. Disc accessing routines are from theUniversity of Nebraska at Omaha ComputerCenter program library. Application of ANGL-CALC to other systems will require minormodifications, Copies of the program may beobtained by writing to: Shirt James Hoffman,Department of Physical Education for Men,University of Nebraska at Omaha, P. 0. Box688, Omaha, Nebraska 68101.

References1, Johnson, Barbara. An analysis of the mechanics

of the stamiing broad jump. Master's thesis,University of Wisconsin, 1957.

2. Pyke, Frank, and Rushall, Brent. A computeranalysis of skills of projecting objects. Paperpresented at Kinesiology Section, AAHPERNational Co.ivention, Boston, April 1969.


WALTER S. REEDRICHARD E. GARRETT

A Three-Dimensional Human Form

and Motion Simulation

A BETTER UNDERSTANDING of human mo-tion characteristics has long been a goal ofresearchers in the area of kinesiology (2, 3).The application of high speed cinematog-raphy techniques for data collection and themore recent use of the digital computer fordata reduction have greatly increased ourknowledge of how man moves.

Recently researchers in the area of humanmotion analysis have been experimentingwith the use of the computer not only toanalyze cinematographic records but alsoto assist in obtaining information about cer-tain characteristics of human motion whichcannot be observed on film. Studies haveattempted to describe accurately the dis-placement, velocity, and acceleration char-acteristics of particular body segments dur-ing isolated motions. More recent studieshave involved comparisons of individual seg-ment energies and their contribution to themotion of the body as a whole (1) .

In reviewing this research, it has becomeclear that many of the problems facing re-searchers in human motion analysis are thesame problems to which engineers work-ing in the field of kinematics have addressedthemselves for sonic time. With this fact inmind, it would seem quite logical to viewmotion problems from a kinematician's pointof view.

Walter S. Reed is a graduate instructor,and Richard E. Garrett an associate pro.fessor, both at the School of MechanicalEngineering, Purdue University, West La-fayette, Indiana.

60

The purpose of this paper is threefold: a)to develop the assumptions necessary toclassify the human body as an engineeringsystem; b) to create a three-dimensionalcomputer model of the human body, employ-ing computer graphics techniques to dis-play the model; and c) to apply motiontechniques to the model and simulate a givenmotion.

The body an engineering viewAn understanding of the body's material

properties was one of the first steps in de-veloping a model for this system. Each hu-man body segment is composed of manymaterials with vastly differing physical prop-erties. The skeleton, however, is the frame-work upon which the entire body is con-structed, and although many other tissuesmay add rigidity, the strength of each bodysegment is directly proportional to thestrength of the bones forming that particularsegment (3, 4). Tensile and compressivetests on long hone show that over the normalrange of stresses to which bone is subject ineveryday use the resulting strain is small.Based on this fact, a first assumption is thatall body segments containing continuouslong hones will be considered completelyrigid for modeling purposes. For instance,the forearm will be considered a rigid seg-ment or link because of the radius and ulna.

Other body segments such as the hands,feet, and trunk are not constructed aroundcontinuous long bones and do not fit underthe above assumption. However, to avoid


unnecessary complexity in the first modelsthese additional segments will also be as-sumed rigid.

The human jointA joint or articulation will be defined as

the articulation between two bones. The em-phasis in motion analysis is of course placedon the synovial or freely moveable joints.Almost every synovial joint in the humanbody has a very complex three-dimensionalmovement over its range of motion. Forinstance, the knee, although appearing tohave a purely hinged or revolute type ofmotion, is quite complex. The primary mo-tion is indeed revolute in nature. The axisof rotation, however, is in constant motionitself. In short, the problem of correctlymodeling the synovial joints is a major under-taking itself and is beyond consideration atthis point in the modeling process. There-fore, each of the major synovial joints of thebody will be replaced with one of two en-gineering joints. The two joints are a purerevolute joint and a pure spherical joint. Theknee, for example, will be replaced bya revolute joint and the shoulder by a spheri-cal or ball-and-socket joint.

A threedimensional modelThe surface of the human body must be

classed as a very complex, highly irregulargeometric contour. To make matters worse,not only does the body shape of an in-dividual vary with age, but body shape fromindividual to individual is highly irregular.Compoun6ing the situation even further isthe fact that as an individual moves, hisbody changes shape slightly. This occurs be-cause of the movement of the boi.es andmuscles supporting the outer layer of thebody. All these facts combine to make agood model of the human body very diffi-cult to achieve.

The first attempt at representing thehuman body in three dimensions was donewith little regard for aesthetics. Much moreattention was paid to thc selection of thebody segments that would be consideredsolid. Considerable attention was also paid


to the individual origin and coordinate sys-tem by which each segment was defined.Each body segment was represented by someconvex polyhedra which roughly simulatedthat particular segment. As an example, theforearms were represented by rectangularpolyhedra. Rectangular polyhedra werechosen because they were thc simplest three-dimeasional solids which roughly simulatedthe forearms. In this manner, experience inthe definition, manipulation, and visualiza-tion of a three-dimensional model wasgained without the problems involved instoring and using thousands of data points.

The body was divided into 13 solid seg-ments and each segment was defined withrespect to its own origin and coordinate sys-tem. The body segments considered solidfor this first model were the head, neck,trunk. the arms, the forearms, the thighs, thelegs. and the feet. Each of these segmentshad an origin established at one end of thesegment representing the particular synovialjoint characteristic to the region of the bodyin question. Following this format, the legwas defined with respect to an origin rep-resenting the knee articulation. The properplacement of all such origins with respectto some fixed Cartesian coordinate systemresulted in a three-dimensional model of thehuman form. A three-dimensional modelcreated in this manner is shown in Figure1, which shows a right side view and frontview, respectively, of the model in its initialposition. Initial position refers to the factthat each body segment is stored in thisorientation permanently by the computer.

Motion simulationSince each body segment was defined and

stored completely separate from all othersegments, and each segment was defined withrespect to the joint about which it rotatesin the body, motion simulation was quitesimple. Each body segment was given acommand to rotate about some axis throughits origin. Following these rotations, eachsegment origin .vas given a command totranslate to some point in the fixed coor-dinate system. With the proper axis of rota-tion, angles of rotation, and translation of

61

origins, the human form model could bemoved from its initial position to some neworientation. The initial position was definedas that position shown in Figure 1.

The points describing the human bodyin a three-space were assigned values cor-responding to a right-hand Cartesian coor-dinate system, x,y,z. Letting ii be a unitvector in the direction of some axis of rota-tion which passes through the origin of thecoordinate system, a rotation matrix [R]may be defined as follows:

[(p.x2vers4) +cos4))

[It] = (tt,p.verso-kuzsinct.)Gtzttxvers4)p-sin4))

(r.,p.sversolizsino) (u,p,verso+toino)(p.,.2vers4) +cos4,) Cup-,versop.,sino)0,,,u,vers4)+tt,sin4)) (ft,'vers0H-cos0)where:

0 = angle of rotationvers4) = I cos4)

Tt. = (W-1- 14 + ki,1c)

Figure 1. Front and side view of three-dimen-sional human model.

62

Positive angles of rotation are defined ac-cording to dlr. r:ghthand screw rule.

A point may be rotated about an axiswhose direction is specified by p. and throughsome angle 4b in the following manner:

[V1 = [R] [V]where:

[V] = vector describing point to be

[x]rotated, yz

[x'i[V] = rotated vector, y'z'

[R] = rotation matrix

The translation of body segment originswas an operation requiring only the use ofselected visualization routines. The routineswhich were rued for graphic display providefor the translation of a point fr,m one posi-tion to another.

Combining the rotation and translationtechniques described above and applyingthem to the three-dirnensional model pro-duced tIle motion simulation shown in Fig-ure 2. This figure shows a sequential recordof the motion characteristics of a subjectduring a jump from an elevated platform.This view was chosen because it correspondsto the view of the motion captured by thecamera for data collection. However, sincethis is a three-dimensional simulation, anyother view of the motion could be displayedwith equal ease.

A detailed three-dimensional modelFollowing the successful application of

three-dimensional motion techniques to ahuman model, a more comprehensive de-scription of the human body was initiated.The surface of the human body, as men-tioned earlier, is highly irregular and com-plex. Because of this fact, any detailed rep-resentation of the body form using discretedata would involve a large number of points.


Figure 2. Motion simulation using the three-dimensional human model.

VisualizationThe scheme chosen for representing the

surface of the body in three dimensions wasto define a series of parallel, transversecross sections. Etch point describing thesurface of the body was then fixed in thethree-space by recording the vertical positionof each cross section. Figure 3 shows amodel created in this manner. Since thismodel is a three-dimensional representationof the human body, Figure 3 represents onlyone of an infinite number of possible viewsof the moclel.

A promising tool for researchers in thearea of kinesiologv is the ability to isolateand study any body segment. An example ofthis isolation technique can be seen in theviews of the head as shown in *rigure 4. [hehead zinc! shoulder areas are shown 'Tingrotated around an axis lying in the pictureplane.

This three-dimensional model is in theearly stages of developmeat with major prob-lems remaining in the areas of visualizationand motion.

Conclusions and recommendationsThe human form and motion models de-

veloped in this paper form a basis on whichto construct a more sophisticated and use-ful model. However, many problems havebeen encountered during these simulationsand remain to be solved prior to the devel-opment of any new model. One such prob-lem is in the area of graphic visualization.


The hidden-line problem makes it verydifficult to distinguish one surface fromanother cn a three-dimensional solid. With-out a bidden-line removal technique, theentire image of an object is projected ontothe picture plane as though the object weretransparent, Because this projection is mono-scopic in nature, and because the front andrear surfaces of the three-dimensional objectare displayed simultaneously, the visualiza-tio s may become unrecognizable. This be-conKs especially true if the observer is un-familar with the particular view lie is see-ing. For instance, if the detailed three-dimensional model of the human form shownin Figure 3 was rotated so that an observerwould be looking down across the bodyfrom an.approximately vertical position, thesurfaces of the body would become very dif-ficult to distinguish. The difficulty occursbecause the body is very rarely viewed fromthis position.

To solve this problem would require theuse of a rather sophisticated hidden-line re-moval technique. A technique to handle thisproblem is near completion at the presenttime.

Another area in which problems havebeen encountered is the area of measure-ment and data collection. Current measure-ment techniques employ the use of markersattached to the skin or tight ents to fixthe position of particular joint!, during mo-tion of the ')ocly. Since the skin and under-lying layers of the body shift somewhat dur-

63

J

..-,

f, ,-,..), ...,:... ,._.,t, r ---"'

l!' -.7 ---", C-

---, i-----",,

':',. ) ' -.------- -,-_.,---- '

c

.-- ----'

- -A -:- ......(

1:,:-:. 1.--

-

." -::-----: -?,

-- - -

..

c......_,3 : "' "-- 4 !; - : , ,

--I4:-----;4

:('-". 1.--."11

Figure 3. Detailed three dimensional human model,

Figure 4. Study of the head and shoulder regions of the detailed three'dimensional model.


ing motion of the body it is very difficult torecord the position of any joint precisely.Precise techniques for data reduction andmodeling are of little value if the raw datado not corrr-tly define the motion of thebody. Time and effort will be nccdcd in thisarca if the form and motion models are tobe realistic,

Further efforts at describing the surfaceand motion characteristics of the humanbody should be made by isolating each of thesolid body segments. Each of these segmentsand the joints that connect them should beapproached as a separate modeling problem.A better mechanical description of eachjoint, including inherent stops and some dy-namic characteristics, should be included inthis phase of the modeling process.

New methods of displaying existing formdata should be approached using the ca-thode-ray tube because of the instantaneousdisplay capabilities of this device. At thesame timc, an effort should be made to re-duce visualization problems caused by thehidden-line problem.

The application of three-dimensional mo-tion routines to the model should follow thedevelopment phase outlined above. A set ofmotion commands coupled to these motion

routines would enable a researcher to pro-duce any desired motion of which the modelwas capable.

A final phase of development should beaimed toward making the size and shape ofthe model adjustable. By adding scalingroutines the model could be tailored to rep-resent any individual. Pathological defectscould also be created in the model to serveas a research tool in the area of biomechan-ics.

The successful development of thesemodeling phases should yield a good workingmodel of the human body which would serveresearchers in the areas of kinesiology, bio-mechanics, and human factors engineering.

ReferencesL Garrett, Gladys E., and Widule, Carol J. Kine-

tic energy: A measure of movement individual-ity. Kinesiology Review 1970.

2. Rasch, Philip J.. and Burke. Roger K. Kine-siology aniVapplied anatomy. Philadelphia: Lea& Febiger, 1967.

3. Steindler, Arthur. Kinesiology of tlw humanbody. Springfield, Charles C. Thomas,1964.

4. Sweeney, A. W.; Kroon, R. P.: and Byers, R.K. Mechanical characteristics of bone andconstituents. Publication 65-WA /HUF-7, Amer-ican Society of Mechanical Engineers, NewYork, 1965.


DOCUMENT RESUME SP 006 737 Kinesiology Review 1971 ... · ED 084 219. TITLE INSTITUTION. PUB DATE....

Documents

Transcript of DOCUMENT RESUME SP 006 737 Kinesiology Review 1971 ... · ED 084 219. TITLE INSTITUTION. PUB DATE....