y‘ .. .' 7— M . .. . .... brr" ' ' --
- . I 'J'A

:43:

.‘Q

I m REEATION narwmq. SENTENCE mom AND
3 ' 1m COMPREHENSION or wmmu ENGUSH

Thesis for than Dogma of M. A.
MICHIGAN STATE UNWERSITY
Damild Keith Darnel‘l
1960

 

 

 

11!: 3314110! 3m 3mm ORDER
AND

m OWENS!” or WRITTEN ENGLISH

By
Donald laid: Damn

Al ”mm
Submitted to tho Callas. of Communication Arts
Michigan seata Univaraity in partial
fulfill-ant of tha "quit-nun
for tho dear” of
mm 0! ms
Martina: of Ganaral Communication Arta

1960

m...“ fiWJB/M

Donald Kaith Darnell

no man of this ttudy Ina to lunatiaata tho "lationahip
bat-authortarotaaataaaaaiaaauaagaaudthacmprahaoaioa
otthataaaaazatythaintaadadraooivarf

150 Want rariabla in thia atody raa autanca ordar.
rha taa W variablaa van (1) touprahanaion of tha muga
aa aaaaurod by cloaa proaadurt and (1) ralatira m, of tha
aaaaaga aa aatiaatad fro- tha raapoaaaa of aotjaata to Manta in
a mttlattd fora of tha mama;

na aaaaral hypothala of tha ataiy raaa Mauro atqa
01 ml tron tit W“ aria: of 0.th will ruin tho
aaaaraoy of Wta' pradutioaa about tla aiaains parta o!
tha aaaaagu

A man at.“ Iaaaaga rat mitt. and arrangad ia a
W" «dart Iaataaoaa rara thaa toaaapoaaa la tha auaaaa
traatiu tin alttrnta for... ra- It: altaraata traatumtt raaaa
tro- aavaa to forty-tho. twat“ Iron tho otuilal cordon
III aorta traatautt war. '2“ M elm procadora and puadtad.
along with a control mama. to 1:50 5‘15”?” (20 par twat-at).

Ila nan oloaa proportion and tha warm ralativa antropy
vara toqutad for aact traat-ant. A: analyaia of varianca waa uaad
to taat for dittaraataa anon; can oloaa and entropy acoraa for tha
an- traatnanta. [ant ordar corralation coalfitlanta van alao
count“ to mania the ralationahip batman tha nubar of

tranapoaitiona tro- tha "tacauaa" order. tha nun cloaa aooraa.

Donald [01th him“.

and tha avaraga ralativa radundancy for tha aavan traatnanta. In
addition. judgmta of difficulty. interaat. and utility Ivor.
obtainad iron aaoh aubjact, and a Paaraou 3; computad to datarnina
tha ralation batvaaa.aa individual‘a cloaa acora and aach of hia
avaluatioua of tha aaaaaga.

tha concluaiona of tha atudy wart:

~l. tiara‘uaa a aiguifioant ralation batvaan tloaa acoraa
and tha transpoaitiona traa‘a "tacauaa" ordar.

2. Thou rara aignificant diftarancaa among ralativa radmdancy
aooraa for tha aavaa traataanta. but tha rant ordar corralation ba—
tvaaa ralativa radmdaaay and trauapoaitioaa from a "baoauaa" ordar
war not oiguifioant.

3. lo avidaaca‘vaa obtainad of a ralatiou batraan cloaa acoraa
and judgnaata of difficulty. iataraat, or utility.

.Aa altarnata aathod vaa auggaatad for obtaining approximationa
to a givaa ordar of a aaaaaao. and a corraction‘raa aada in tha
prooadura for altiaatiaa tha proportion of ralativa antropy from

tha raapoaaaa of a gray of aobjacta.

TH! RELATION 3mm SENTENCE ORDER
AND

THE W810i 0? WRIT!!! ”61.181!

By
Donald Kaith Darnoll

A man
Submittod to tho Collaso of Omication Arts
Hittigan Itata Uniroraity in partial
folfillnant of tho roquirananta
for tho dozroo of
MASTER Of ARTS
Dopartnoot of Oanorol Communication Arta

1960

rauwcmrms
ACWS..55.........-........iii
turACI.....l...............l...iv
usrormus.....................vu
LIMOIWTIGI.................-.ix
W........aa.a......s.~... l
m1..........s..a.saauua.a.4
mn...............~.....-..-..23
m111......~..~.......o..s....32
man...”...............-. My
lDLlOGRAUH..~..-..u....-.-..-.~..-.-... 56
amnl....................... 59

Amn'oOooaoootooooooo'aao‘ao'ao60

ti

ACME

1 out to upraaa oy approoiatioo to my advioor,
or. David K. mu. .5 aaoiotad aa mmugaa progn-
o! rooooroh) to It. lool W. Dr. Dov“ lottor. .
Dr. lunar! Insla. and Dr. John ltroot, tho holpod in tho initial
plan-in mitt; and'to Dr. $1601. Hookah Jr. a“ Dr. '
Irvi- Iottinshaoa rho criticiaod tho final roport.
I out aloo that tho ml protaooora who ooatributod
tlaaa um aod m... m thia nae-m; ‘
I“ “shoot hour it rout-rod for tho flatly, frioodo,
and oollaaxuao who liatoood patintly.

iii.

MAC!

tho oomunioation procooa oftan involvoa two pooplo, and
alwayo involvoa two rolaa, tho 323:9; and tho doatination.

A difforont olaaaifioation can alao ba oado of tho communication
attora, tho gaitiasor and tho W. Tho initiotor and

tho roapoodont aay bo diacrininatod by tho foot that a willinznooa
to oxpand tho noooaaary attort to accompliah a apaoifio parpoao
nay ba arm-nod on tho part of tho initiator but not for tho
intondod roapondont. that it, who ono of tho actoro initiatoa

o commotion it nay bo ant-ad that tho ratio at oxpottod
ravard to tho upaotad ottort ia aqnal to. or oaoaada, aaity.

Tho initiator _o_a_x aaauo that tho probability of tho daoirod
roaponaa ia approaioataly aqual to tho parooirad raword/offort
ratio for tho roopondaot; (for aaanplo. if ho ia cartaio that tho
iotndoo roapondnt howa that roaponao ia daoirod). no nay,
thu, wioh to introaao tho probability of tho doairod roaponoo by
dooroaainx tho affort that rill ba roqoirod of tho raapondant in
giving that roaponaa. it ho can do on without rodncinx critically
hia own mrd/offort ratio.

Co-ooioation raaaarch ia diroctod toward thoaa variablaa
which atfact tho rooord/offort ratio, and ono of ita goala ia to
oiniaiao tho 592; M raqnirod of tho aotora in a givan

canonization chain. Iran in tho ainploat form of tho comioation

iv

chain. howovor. wharo tho initiator porforna tho aouroo, ontodor,
and trananittar functiona, and tho roapondant portorna tho racoivar,
datodor. and daatination functiona, it ia not any to aolact tho
approach that will roquiro tho ninimn total affort.

a our-bar of diffarant approachaa night ba tahon. For ananplo.
tho manta: could aoloct a vary rodundant'codocoona a... uaaa a
largo mm of aynbola‘ in 51...... to tha a... of information,
hnowlodgo, to bo trananittod-ointading to lininiao tho load on
tho “odor and dooodor functiona. moopproach, howaoar. tnda
to incraaao tho button on tho tronanittor' and. rotoivor fuctiona
and nay raault in no aignificant ohongo in tho total ottort oprdod.
'lha oppooito approach could aloo bo adoptod.‘ to nininiao tho ratio
of aynbola to contant, but again tho roault ia a ahitting of affort
tron ono function to anothar rathar than a radootion in tho total
ottort of tho ayatan. A third approach (tho ona noat oftau adoptod)
u to aininioo tho affort roouirad of an initiator _o_l; tho roapoodaat
with tho conaoquanca‘ that an ancoaaivo burdan ia plaood on tho
othor. is haa boon ouggoatod. tho oolution to tho problan in not '
an aaoy mo. ‘ A O

1b alporinontal approach adoptod in thia atudy ia (l) to
hold that nuhor of aynbola tonatant. (2) to vary tho ordor of tho
aynbola'. producing aovon troatnanta of tho naaaago. and (3) to
oboarvo tho bohaviora of raopondonta to aaa if tharo aro difforontial
ottocto .oug tho troatnanto. If aigniticant difforonoaa ara
ohaorvod, and if it to aaarnnod that tho aamo amount of offort ia
axpondod by tho initiator and roapondont on aach of tho aovon
trootonto, than. that troatmont which ahowa tho groatoat offact

V

(o.g.. yialdo tho highaat noan couwrohonaion acoro) can ba aoid to
roquiro tho laaat total affort in tho trananiaaion of a conatant

amount of information.

vi

rahlaa
l.

3.

10.

« LIST OF TABLES

Page

Samplo calculationa of Entropy,
My. andxiadiraction. . . . . . . . 13

no Noon Ccoraa for tho Saran
Crown of lubjocto on tho Control -

‘anolrporinantalllooaagoa......... . 33

Bartlott Toot for Monaity of
Varianca Anong Saran M. of Cloaa
SooroaonthoControlMaaaaga. . . . . . . .3’0

Analyaia of Varianco of tho Control
mm.oooooooooooo0oooao035

Bortlott root for Hmogonoity of
Varianco Among Sovon Gmupa of Cloao
Scoroa on tho lxporimontal-‘rimod

m.ooaooo00ooooo'aoooo036

Analyaia of Vorianco -- ’Elttporimantalo
Tindscotoa................36

Analyaia of Varianca of Exporinontal-
rinod Scoroa Dividod on tho halo of
Scoraa undo on tho Control Mooaago . . . . . 37

inalyoia of Varianco w Exporinantal-
BotinadScoroa.........‘...... 38

Analyoio of Varianco of lxporinontal-

lhtimod Scorao Dividad on tho Baaia

of Scoroa nado on tho Untinod Control
baaogo...................38

; Toota Batman Adjacont mono --
hparinontal-Timod Cloao Scoroa . . . . . . . 39

vii

Tabla logo
11. 5 Taco hotwaon Adjacant Mona ~-
lxporinantal-Dntinad Clooo Iooroa . . . . . . 39

ll. Analyaia o! Varianco of Iotropy ~
“on. O O O D O O O . O Q Q Q C Q I O 0‘ 0 O ‘0

l3. Corrolation hotwoan Cloza Scoraa
and tcaloa oi toga. Intaroat. and ~
mtx1&,aooa.o¥1ooooo6ooaooooo"

1A. Analyaia o! Varitnco Among 'l‘raati-
nantator‘lhroaJongIntScalool é o a . . .‘3

viii

Figuro

l.

3.

b.

5.

LIST OF ILLUSTRATIONS

' Pogo

‘l'ho oovon ordora of tho mart-anal
moooogo...................25

Comoriaon of noon rolativo redundancy,
oloao proportion. and niadiroction for
thooovontroatmonto......_...... 41

hylor'a calculation of outropy for tvo
idaal cocoa and g'a of 10, 20, and 100. . . . 50

Corractad calculation of antropy for
two idaal caaoa and 33's of 10, 20, and
10° 0 O o I o O O O O O O O O 0 O a o o a o a 51

Compariaon of clooo proportion and

rolotivo rodundancy aa oouputod by

'l‘oylor'o nothod and tho corroctod

at“ O I l O O O O O o I o O 0 O O O O O O 53

WODMIOI

‘lho purpoaa of thia atuay io to invaatigata tho ralotionahip
botanonthaordoro! aantncoainonaaaagaandthaooqrahuaioo
of that aaaaaga by tho intandod raoaivor. rho aaaartion haa ha.
nado rapaatadly by authora oi tanthooha and taaohora o1 Ingliah
coqoaitioo tht it in important to ottud to ordor in crootiog a
writton or apohaa naaaaga. that in. it an individual trananita a
naaaaga in whatovor croor tho uita occur to hia. tho daairaa offact
nay not bo obtainad, whilo anothor ordor of tho aaoo unito night
achiovo tho doairod afloat. rho proaaat raaoaroh to an attanpt to
obtain aona qiriaal aupport tor a qaciiio oaaartioa which ia

daduoihlo iron tho noro guaral ono atatad abova. thia apacifio

”mum 1- cmmmugmmmam
W 2!. m2 3.22m-

tho unnum- of tho atody lioa not in proving to. ooporta
right or wrong but in tho poaaibility of inaraoaing tho otticiucy
of «Imitation. Thorn aro aovoral roaoono for arming that
ordor nahoa a ditforonco in tho noaning olicitad by a not of
aynhola (co-para g! with 55, 3355 m with gas 33;, or _l_1 with
m. but thara ia no raaaon to aroma that con arrangan-lt ia
oocoaaarily noro difficult to produco than any othor.

no noth of thia atudy ia to couparo noon conprohuaion

acoroa for groupa of auhjocta on traatnonta or a naoaago which

diflar only in tho ordar that tho oantootoo occur. rho out.“ ia
ohooao on tho unitto ho nanipulatod haoauaaa (l) tho afloat o!
ordar on ooaprohaooion can to cloarly dononotratad at tho lowar
lanolo, (I) it haa ottan boon aoaartad that thoro lo a "boot"
ordor- tor tho. largor .ito. and (3) at laaat coo invoatisatioa at
a highar lovol (paragraph) iailad to yiold a oigniticaot roault.
It to aoomad that an affoct of aantanco ordor can only to
prodiotad it dapaadonoy can to duonatratod botwaan tho oantucoo.
and it tho daotination io-oogniaont at tho dapandanoiao. with
that aaouqtion. tho following dotinitioaa aro critical to thia
otudyt
l) Civan a oat of uita, (a), (b). and (a); givan that
nit (a) coma. if can can prodict tho occurruoo
at (b) or (c) with graator than chanca ancoaao oooa
dograa oi dopadaooy obtaino botwa. tho nito (a).
(b). and (C). .
2) Givan a naoaaga, to tho outnt that tho ordor of tho
anita ia datorninad by a dqandut rolationahip bo-
twaan unito thoro lo M; in tho noaoago.
3) div. a atrnatnrad naaaago, to tho antant that tho

dootinotion haa accurato oxpoctotiona about tho
ordaring of tho unito within that naoaago. tho

dootinatioo lo M.
tho doiinitiono .1 othor ooncopto portinnt to thia otody

unat bo withhold nonantarily until thay can bo praautad in
thoorotical poropoctivo; howovor. tho pupoao of tho promt
raooaroh can now bo rootatod. It it to invootigato tho ralation
batwaan otroctural variation and organiaation.

thia otndy ia to ho pxaaaataa in four parto. chaptor x in-
olodao a thoorotical background for tho raooarch, a doacription of

tho naaouring inotruant. tho ganorol hypothaoia, and a rationalo.

Chaptoo I! includaa tho doaign of tho alparinont and a daocriptioo
of tho tochiano of ohioiotration. Chaptor lll roporto roanlta o!
a otatiotioal onalyoia of tho data. Chaptor IV includaa a ohort
os-ary of tho anporinut, conclnaiono. a diacnaaion of tho rooulta.

aoaa ohoorvationa about intonation thaory. and finally. auggootiono
for turthor roaaarch.

CHAR!!!

'thio chaptor inclndaa a thaoroticol background tor tho
rooaarch, a daacription of tho naaouring inotruant.‘ tho ganoral
hypothoaio. and o [rationala for tho otudy.
momma. «mm-run. m ammun-

a najor part at that rooaoaing bohind tho praoont otody io
baaad on intonation “thaory. It io tharotora iqortant to in-
troduca at thia aarly point aooa of tho concopt'o at that thoory
and to datino than lingniotically and oath-atioally. ma at
tho control concapta ot intonation thaory lo M or
m. thaoa torno aro co-only oaad with divorgant naaninga.
intonation rotorring to ”hnowladga.” and antropy. in tho thaory
of thornodynanico. hoing aooociotad with tho probability a: o
givan diotribntion ot nonantua among nolaonloa. in information
thoory, howavor. tho two taroo aro aquivolant. rotorring to tho
unoortainty or loch ot hnowladgo in a commotion oyataa.
Hart-an flaavar orprooooo tho ganoral naaning oynboliaod hy tho
torno in thia way.

lotropy aaaociotad with o aitnation ia a naaouro

at tho dograa ot rando-Iaaa, or ”ohnttladnooa"

it you will, in tho oituotion.... M for a

to-unication aourca ono can any. . . ”thia

oituation in highly organiaod. it to not

oharactariaod by a largo dogroo ot randonooo

or of choico «- that ia to any. tho infatuation
(or ootropy) ia low."1

 

lClanda I. Ghana: and "arr. Woovor. n; W

_ m 2.1, W (Urbano: tho Univoraity ot lllinoio Prooa,
194’). p. 103. 4

Iron thia coplanotion. it would oaola that a nogativa ralation
obtaino botwo. ontropy and organioation. 'sinco organization in
a hay oonaapt in thia atudy. a noro oxact datinition ot antropy ia
roooirod to ootahlioh thia ‘rolotionohip. '

Giv. an infatuation oonrco which io producing a naooogo by
anacooaivoly aalacting diooroto aynbola tron a oat of aynbola, it
tho oynbola aro indopondant and aonally probabla that oourco haa

m m. tho valua ot nonin- ontropy tor a oat of 3 aynbolo
nay bo anproaoad
Hmcx '-' “(4i 1°32. 1!?)-

thno. itgoooala tour.§naaat (-k legal) at (.5) o a. that
ia to oay, a aitnotion in which tho oonrco nuat dacida botwoan tour
opally probabla oltarnativoo containo two "hito" ct intonation.
no nati— odor of "bito ot intonation" or “unito ot ontropy"
in a ayatan io aqual to tho nunbor ot tinoo tho amor ot altornotivoa
nnat bo dividad by two to obtain a ouotiant ot ono.

01v. an intorution aourco which ia producing a noaoaga by
oooooooivoly oolacting diooroto aywolo tron a aot ot aynbolo. it
tho probability ct choico at tho varioua aynbola. at any atogo in
tho prooooo, ia dopadant on tho proviooo choicoa tho aourco haa
m m a!” m. In tho aituation wharo tho oynbola aro
dqandant or finally probabla, tho oxpraooion tor tho abooluto
(oboorvod) ntropy io

H:"'[PI 1091 PI" P1 ‘09). Pa. t ' ' ‘ 4' Ph '09:. Po].

fhuo, it 3 aquolo tour ao in tho proviono onanplo, and p, pr. pa.
aodp. opal .30. .25. .15. and .10 roapoctivoly. tha abaoluto antropy

ot thia ayat- aqualo 1.71021 ”bito." no glotivo m o! thia
ayat- io tho ratio ct tho actual ontropy to tho noni— atropy
or .8113 (871.). and ano ninua tho rolativa aatropy it tho filativo

m at tho oyotu. ' a toroula tor aapraoaiu ralativa
radudancy tollolwoa2 ‘

.1_ [2191109191]
“‘03.- ‘ogzi‘i-J .

It. in tact. atropy ia nogativoly rolatod to organiaation, than a

 

poaitivo rolation ohould obtain botwa. orgaaiaatiou and root—duty.
At variowa pointo throughout tho mind“ of thia thaoio it will
bo couraniant to rotor to aithor ontropy or rad-Idancy. but in aoch
caao. tho obvaroa io iwliod. ‘ ,

I Itho ooncopt ot rolativa ontropy aoownaa a dotorainar or dator-
ainaro. without prodatorainatioa tho aouroo would alwaya havo
cooplota troodoo ot ohaioo -- noxiou- antropy would obtain. For
tho h-an oourco tho datorninaro aro tho ruloo which tho oourco
adopta tor tho conatruotion ot a particular naaaago. thaoo ruloo
nay ha of throo typaox (l) thooa which aro counonly accaptod in a
languago calamity (a.g.. troquoncy of proviona uaago). (2) thooo
which aro baaod on opacitio agroonant batwaan tho aourco and
doatinatioo, and (3) thooo which tho aouroo oolacta arbitrarily.

 

11:14.. pp. 103-105. room aro availablo for (-p lamp)

in I. b. howl-an. "Conputational Mothoda floatul in Analyoing a

Iorioo ot binary Data." tour 1259...}. 9; mm,

tho aigniticanca at tho diatinction ia that tor oach at tho throo
typoa than it a dittoraat probability that tho oourco and tho
daotination will agroa on tho oat. Irho oniotaco ot tho third
poooibility arguao that -tropy ia a choractariotic at tho oourco,
and that a cooaidoratioo ot rolotiva .tropy oo o charactoriatic
ot a nooaago or a languago involvoa tho aaouption ot ona at tho
tirat two typoo oi ruloo.

loppooo a givaa oourco cowpoooo a naooago and tronanita m
g; g to o raooivor who io aapoatad to oonploto tho noaaago (i.o..
play tho oourco rola). rho rolativo ontropy at tho two cont-coo in
wing tho withhold part of tho oooaaga, and conaoouontly tho
two voroiaoa of it. would ba orpactad to dittar to tho ontot that.
tho rulaa thay hava adoptad dittor. Moro. to tho aatant that
tho two varaiono ot tho withhold portion ot tho nooaago aro oinilar
itaanboaaa-ad that thorulooadoptad by thatwooourcaoaro
aioilar. And. it tho two voraiono of tho naaaaga aro idantiaal. or
vary ainilar. it would own to tollow that tho aituatioa io highly
coo-out for both aourooa, or that tho withhold part ct tho naoaago
oootaina littlo "intonation.”

bait iapoaoiblotocontraatandocqarathathroabooic
oanaoptootthiaotudy. div—o oatot ay-boloanongwhichdap-dont
ralatiooahipo obtain, and givan that a naaoago ia conotrnctad tron
tho out: it tho ordor of tho oylabola in tho noaaago ia dotoroinad
by tho dqoodnt rolationohipa thorn lo M in tho naoaagog
it tho ordar io datarninad by tho ruloo which tho oourco haa adoptad

thora lo M; in tho oourco; and it a rocaivor io abla to

prodict tho oyabola and tho ordor of tho oynbolo with grootar than
chantooucoooothoroioWinthadaotioation.

If. for a givon noaoago. tho rulaawhich'tho oouraahao adoptad
aro baoad on a knowlodgo at dopondut rolotioaohipa in tha opiol
oat; tho proportion of rolativo radudanay in tho oourco ia canal
tothodograoototruatnrainthonoaoago. it. torogiavaonaaaaga.
thoaouraaandtharoaoivuadoptthooanarulaa.rolotivoradundancy
andthadagraaotaruaiaationaroaowal. lt.toragio-naaaaga.
both at thaao condition aronot. rolatm rodndanay. atrooturo.
andorgaoioationhvothoo-avalu.

mm:

. Cloaapraaadoraiothanaaouriaginatruutchaoantoruoain
thia otudy. to lo dotinad by ita originatoraa "a popohologioal
tooltorgmiagthadqraaottotalcorroqondanoobatwa-(l)tha
ancodiag habito oi tron-ittoro and a) tho daooding habita ct
r'acoiovora.”3 thoprocaduraiaqnitaaiqla. motohaaawrittn
ooaaaga and dolotoa a portion of thaworda. tho dolatioa ia par-
tornadinouchawayaatobaindapondntotthawordathnaolvoo;
i.o.. tronaroadanotartovaryh*“wordiadolatad. thiaiodona
booauoa”itaa-aouaaoarytolotthaooourrucaootollaortoot

worda ba roproaantad according to tho proportion of thoir occurronto. "‘

 

vi

’Ililaon Luio 'Daylar. ”Application at ‘Clooa' and tntropy
Hoaauroa to tho ltudy of Contartual Conotroint in Ianploa ot
Continuouo Drona " (unpwliahad tint. diooortation, 'l'ho
hivoraity ot lllinoia, 195‘), p. 3.

‘33.. p. o.

Ilo'oko of nation oioo roploco tho dolotod torch, and tho outilotol
nooooao io prooutoo to o group of ouhjooto who ottupt to roplooo
tho mining outdo. \Iho m of dolotod outdo which o ijoct
mu: rnplocoo io hio "oiooo ocoro." which no, ho ”Mod
oo o propottioo for ooopototivo mono. . ~ - I t

' tho "olooo ooooo" to: o 31" oohjoot ood o givon oooooso
io tohon to ho o nooooro of tho dost-no of oomopood-oo
hot-o. tho uuoozo hohito hood by tho oourco ohilo
"attains" tho noooogo (fitting ooquonooo of Mo
oydzolo to tho nooning‘) ond tho immo hohito oood
by tho toooivot ohilo "heading" it (fitting noonio.
to tho mtilotod nooooso) nod, on tho hooio at noon“.
porooivod, ottqtiu to onoodo thooo oi-ooto thot
will noko tho oooooso'o ion filo min.’ '

to o Iota pobliootioo nyioo ontioiyotoo oono o1 tho fiction
thot night orioo no givoo thooo mm:

rum. q to noo‘iooiooto thot tho oooioot ooyo of
outlying olooo pmoduro noy ho boot for out oooo.
thou-o oouo to ho littio odvonmo in prooloooityiu
oordo no lioitins dolotioo to than. ood no odvootoso
to putting ooooolf to tho ttouhlo of. Judging 'ond
nearing oyoonyno. Aloe. it oppooro thot on ova-r
fifth-wort! oolotion oyotun opoooo blonho oo for oport
oo thoy nood to ho. turthnt. o oorioo oi ohout 50
hionko to roughly outfioiont to olloo tho ohontoo of
nothonioolly oolootin; oooy or had wot-do to oonool

out out! yiolo o otohlo oooro of tho ditiioolty of o
pooooso. or tho porter-on“ of on iodividooi. doopito 6
nht oponiiioootoothoooutiu-ont ptooooonoyooloto.

lono intention ohont Ihot olooo nooonroo oo oou oo in.
“motion ohoot ito roliohility io proviood by tho folio-ins

oomlotiooo. In o hotooo-ottoo looming outpost-at. olooo oooroo

4. . ._‘__. .

3333.. pp. 1-3.

q"noon L. toylor. "moot novoloonooto in tho in of Clooo
Mo." WW. mm (Ilium. 1956). 45.

10

woro ohtninod olong with. conprehuoion scores from way

wolidotnd ultiplo ohoioo tooto and AH?! intouigonco ocmo.
clooo corrolotod with tho "bolero“ omnhonoion toot .70 and with
tho "oftot" compo-hollow toot .80. Elm corrolotod with imlf
hotoro ood often: .88. contohoooioo oooroo oomlotod with um
'.65 ad .70. onto olooo and AN: corrolotod .73 on! .76. A11 .
thooo g'o oto oigniiioont ot tho .01 1on1 ood on m torso and
pooitiwo.’

Clooo procoduoo no originolly iotonood oi o nooooro of
“ooooohility.” noviooo rooooroh in thot oooo hot tho-II thot
ooooo'oo could nooooiiy ho ronhol ooooodin. to diuionlty by ,
oonoidoniu o won for muhloo. tho hon-Chou foo—1o. for
ooh-no. io hood on mono oootoooo loath and poroootogo of on-
toniliot was.” no Iloooh gain-.1. ooooihwo mun-bu of m. x
poo oontoooo ooi tho haul of oyllohloo por hondrod worth, It
who «rim. howovor. thot thooo tor-loo did not toho into oooout
on tho votiohloo thot‘oiioot toodohility. ”mo—y thiok of olooo
’rooodoro.“ ooyo “ruloo. “oo throwing on potatiol roodohiiity in-
floooooo in o pot. iottio. tho- too-out. tho-hoopla; tho mulb'do

A_i_‘_ .4 .M . ._

’ma.. 3.

“Mao! holo 'ond Joouo I. (boll. ”A m1. loo puma.
hoodohility: mutation." Banana 1 Reooorch allot“. um
(Noun. 1948). 37.

’kodolph floooh. ”A I. hoodohility Yordotioh," m 9;
Anti“ m. mu (Juno. 1948). 223.

to ’ '
Hilooo 1.. Taylor. ”'Clooo Pratodoro' I A low tool for

moo-tin Inability." W W. xxx (roll. 1953). 417.

 

o , o b I. . 4 _
' h V I ~
, .
| ‘
. w
. o
n . (r
A .. bl t .h .
. 1 . . . . w
I. A r o
. i o ‘ y I
o, . , o o .
I o I II
0 A I
t A I fit . .
Con .
o.
h .
'- . n‘
I . .
U i l
.
O
o 7! O
o . I .5 ¢
I . u
.
’ . o . ..
.u _
.l .. . A .
. ‘ o a 'Q. J I O
o
. I . .
o a , .
. ‘ . . r .

.1
. I . . Ha
I g ,
O U . c x o
In
I to .
. a
.
1 n \o i I.
J ..n . 1 . o .
A
l
. q ‘ . 1 ‘
-
cl 0 I I L
o .
a
I
- a
I n o n l ‘ I
. .
. . .
. u .
’ I 1‘
1 1 .
o
. | ‘
. .
. . o . b N
o
3 O . J.
or o ‘
w \ 4
.o .
1 .¢ . . o .
h»
.- l . .
o - A h
J ...
1 ..
.. r
.3. h . . I
.
1 . I...’
I ’
o t.
l n t
5 A t I
o :r l y I. .1 1 .
O, I! I
I l
y 4 u. . o
.4. . . A
I h \I:
. . ¢ . . J
J . I O .
n .
o
o ‘ y '1 o I
o- t .n I .
O '
o . A:
. I. s
a
A . I
. V
O l
I l .
I
y o

11

C10“ touts corroloto highly with findingo of the Polo-Chou and
flush fomloo oo otondord mtoriolo,‘ hot when tho nochoniool
tutu-loo oro oppliod to non-standard materialo cloao oooroo ooon

to ho bottor noosuroo of tool difficulty.11

- Clozo oourco, thou,
woy' ho on indoor of honing. oomprohoaoion. iotolligonco. or woooogo
difficulty. om! it io o nottor of control thot dotorniooo that tho,
Moo in o portioulot oitootioo. . ‘

no prootiool ointficonto of olooo proooiuro for thio otody
it thot olooo prooodoro it o oothod of. obtoinin; prodictiono Iron
on individul. It in on”! that no mun-‘1'. proiiotiooo m
on indoor of hio onoototiooo. tho ptoportioo of hio prodiotiooo
thot oro "oottoot" io. thou, on ootinoto of hio ouoninotion. nod
tho oooo olooo oooro {or o 31‘9" of oohjocto io oo ootinoto of tho
ovorogo onooiootion thooo owhjoeto hovo in rolotion to o givon
noooogo. no oooortion thot olooo oooroo on on inch: of orgoniootioo
in oddition to hoist: on iodox of learning. oooprnhfioioo. in- .
tonigonco. or oooooxox difficulty. ohoold not ho oorprioios, oinoo
on of tho lotto: ooooopto oro cortoinly rolotod to tho proton of
(mil. omoot oxpoctotiooo (orpninotioo).

By 05min tho prodictiono of graph of owhjocto. scan in-
iornotioo ton ho ohtoinod ohoot tho ohoiooo thot o m hoo udor
opooifio conditiono ood thoir tolotiwo prohohilitioo. tron thooo
on notinoto oi tho tolotivo rodudonoy or .tto" of on "mop"

oourco con ho ulculotod. To obtain thio ootinoto ono trooto oooh

 

12

hlonk of tho mutilated message as a diooroto system and the rosponoo
of oach subject oo o possihlo outcome of that system. By computing
tho prohohility of oach word'o occurrence, it io poooihlo to approxi-
oloto miricony tho degroo to which the blank io dctcmiucd by in
contact.“ mm ontropy would thou ho indicotod u: all ouhjccto
.iwo difforont rooponooo to o blank, ond toro oatropy if o1]. owhjooto
giro tho tono rooponoo. Givoo tho roopooooo of o group of oohjocto
it it poooiblo to compute noxim ontropy, rolotivo ontropy. ond tho
rolotiro rodundaucy of o gorticulor hl__§_n_k (ooo pogoo 5 tad 6). tho
oworogo ontropy or rodondoncy for tho moggo io oioply tho oritbotio
noon of thooo ocoroo for on hunko.

Taylor offoro otill onother mom tbt any ho ohtoinod in tho
oono ooopototiooo. thot of "milnfdroctiou.‘l Tho clozo proportion
(proportion of right oomro) io ronovod tron tho oot of oltomotivoo
and on outropy woloo cooputod from tho twining roopooooo. Th1!
ohooloto voloo is then divided by tho maxim ontropy voloo for tho
cholo oyoton to ohtoin ronoining rclativo atropy. Wining
rolotivo ontropy pluo tho clooo proportion ouhtroctod from ooo givoo
tho niodirctioo fixer- (Lo..CP+RRH+M= 1 3-” 'l‘ohlo 1 than
o ooqlo colooiotioo of ontropy, rodmoonoy, clooo proportion, ond
oiolirootioo. lhio tohlo con-u toooty roopoodooto for o oioglo
block. of when too givo tho right onovor. fivo giro o oinglo wrong

ooowor, and tho ronoioin; iivo giro difforoot wrong onovoro.

 

u‘l‘oyl.oi'. "inflation of 'CIooo' cud Entropy Hooonroo. . .."
y. 15.

l3
2“. . 59.61o

13

TABLE 1.--Samplo Calculations of Entropy, Redundancy, and

 

 

Hildiroction
geoponaeo Frequency gggortion - p log a E
o 1 .05 0.2161
b 1 .05 0.1161
c 1 .05 0.2161
4 1 .05 0.2161
o 1 .05 0.2161
f 5 .25 0.5000
g (c. P) 10 .50 0.5000
'56" 3736” "523-365"
Maximal ontropy o 20 (0.2161) 8 4.3220
lolotivo ontropy o 2.0805 I 4.3220 o .46
lol. Macy o 1.00 - .46 o .54
Clean proportion :- .50
In. rol. out. 8 1.5805 I 4.3220 o .36

'1“th 3 1.00 . (o50+o36) I .1“

 

 

14

A TREATMENT OF STRUCTURE

Bocnuoo clooo procodurc requires a aoooago approximately 250
vordo in length, and because the Fletch formula indicotoo that a
oixtoon word oontence is £115}! difficult. it vao decided to croato
o fifteen oeotenco meooogo. Since it would he impoosihlo to vary
otructuro through on poooihlo orders of fiftooo ooutonceo (thou
oro 15! poooihlo ordoro) it woo ooceooory to ooloct o opociiic kind
of ocean otrnotoro..

'l‘ho "hocouoo" ordor ooeood to provido tho noooooory condition
for o otmturod noooogo oineo. ao Millo ooyo. "l‘horo io ooo oooontiol
rolotionohip coon: idooo: ooch oubpoint out oooiot in proving tho

"1‘

point to which it io inodiotoly onbordinotod. . . In tho "hocoooo"

ordor tho toot of ouhordinotion it tho inoortion of "for" or

15 Coordiooto

"hocoooo" hotwooo o topic point and in oohpointo.
ouhyointo nndor tho oomo heoding con ho connoctcd by ”and" and
or. logicolly intorchongoblc.

Admittodly, o "hocoooo" ordor outlioo io not 32: way to
otructore a moss-13o, not it it opplicahlo to oil hiodo of ootoriolo,

hot it io ooo way on which many outhoritioo ogroo. and is thoroforo
oolootcd for tootio; tho hypothooio.

‘— __.__ .L-

i w—vv - , v-v— v- v

4
‘ Gloo I. Hillo. c oi tho m (Now York: Prontico
Boll. 1no.. 1951). p. 7.

ӣ814 ond Donold c. Bryant and Earl E. Wallace. Fondonutolo

of Public £39.21 (Now York: Appleton-Cotton Coupony. Inc”
1347). p. 229.

15

To slazpiify tho process of creating an moose o fifteen unit
outline in skeleton for: woo set up and oootoncoo which not all tho
requirmto for logical subordination and coordination were writt‘
to fit. Too independent jedgoo verified tho writer'o judgment. to
outline in which eoch noln point hoe the oeno oubor of oobpointo
woo mood to equalize contextual Gifford-coo end to oi-pllfy the or-
po‘rioontol mipolotiooo’ of otructoro. '1'ho novelties noooozo. in
outline form, can he food in Appendix A. the dependacy between
ontonoos io clear: given my one of the nojor hoodo. only two of
the fourteen other ontoncoo noy logioolly follow 31v. one of
thooo, there are two coordinate pointo which ooy follow.

W818

the gonorol hypothooio of thio otudy in
MM“ 2.022.! 2!. “m1 m .9. "twee-m" 3.4.3.! 23.11 m L!
occorocz g; rooggndmto' prodicgiaoo M £13: Biotin mg 2; 32
music.
narrow:

The general hypothooio ototed above it e cooclooioo heood on e
mice of proniooo. ‘l’hooo ore:

1} Structure in o oyotoo incrooooo the intornol predictohiliey
oi the ole-onto io thot oyoton.

1) The "hocoooo" ordor io o kind of otrnotoro.

3) The "hocoooo" ordor io o Emilior pottorn to noot our.
of English.

4) League” oooro ore capable of roopoodin; to voriotiooo in
otrntturol predictohility.

16

The firot of thooo pronieee ie eeeuned ee e metter of definition.

Given thet the "bounce" order impoeoe dependenciee mung unite with-

in the noeongo. it ie otroctorod. he "become" order io taught in
nony echoole end it recmondod by textbooho on o good way to "orgenieo"
neoeogoe, eo it Ioeue roooonehle to ocean that noot univoroity
etudente hove hod experience with the "bounce" order. The fourth

‘ io oupported by ompiricel date.

Stndiee which hove dealt with nooeoge order. etrocture, or
orgooioetioe will be diecoeood in three groupe: (1) thooo which treat
but do not emport the preeent hypothoeie. (2) thooo which follow
einiler proceduroe hut ere irrelevent to thie orgu-ent, end (3) thooo
which eupport a. gonorol hypothoeio of m. etody.

he only etody which hoe cone to thie writer'o etteotion thot
telle in the firet cetogory ie that of K. c. Boighley. Beighley had
eowerol opooch etodonto outline the notoriel tron two epoechoo end
orriwod et tho "hoot" order by domocrotie noene. the peregrephe
were then tendonieod with the following roetrictione: peregreph
o‘er one not not ellowed to he in the firet iiwe. no two peregrepho
were per-itted to he in the originol order. end no three poregrophe
of the tone nein point were ollowod to follow ooch other. emporieon
of the Inn emrehonoion ecoree tor the "orgenieed" end "dio-
orgeoiood" wereiooe toiled to chow o oignitieent difference.16

Mh Boighloy'o rooolt once not eupport the proeont hypothooio,

 

16
I. c. heighloy. "An prerinontol Study of tho Bfioete o!

rowr Speech Veriohloe on Lieteeor Conrohonoion," .225; m.
xxx (love-her. 1952), 2119-258.

17

it con not he ooneidorod no negative evidence boceweo: (l) foilnro
to reject the null hypethooio dooe not noooooerily dioprowo the
theoretieol hypothoeie. (2) it it poeeiblo thet o peeoogo of
perogreph length providoo onfficiont content to rooch no option
level of orgonieetion beyond which euhjeete do not ettend to etrwttwrol
dependencioe, end (3) loighley etotoe.

Where it woe dooireble to elter the wording of e treneition

et the beginning of o peregreph in order to echo the

neteriel reed noro onoothly, eoch e ehongo woo node.
slight rewarding wee nocoeoery et tho beginninge of only

twoporogrephoioeoohefthotwoepeecheo.l
thie could be interpreted ee toying thot tho rendon wereione were" y
ooorly eo well patterned on the otrwctnrod woroiooo. end the onoll
difforoneoo which did occur were proqtly corrected. the nosetiwo
reowlt night hove been predicted, depending won whet Beighloy went
by ”reed nore monthly.”

the following etndioe ere included only heceneo they hove deelt
with the effect of noooego order on receiver'eonprohonoion. annoy
end Berle. defining e coll ee "o noeoege enit. coneioting of on
eeeortion end the ewidaco eupporting thot eoeortion.“ weriod the
order oi eollo on well no the order of eeoortiooo end ewidaeo with-
in the cell. prodeeing oin oxporinontcl nooeogoo. they fownd no
oignifioent difference enoug the ole trootoonte on the retention of

18

the propooitione, eeeortione or evidence. Sponherg found totol

1
71514. g ’e ”lo

18llclhert I. Galley end lewid l. lerlo. "lffoct of Intereellwler
end lntrooollnler Speech Strwctnre on Attitude Cheogo end morning."
£29.22! W, XXIII (Novonbor, 1956). 287-297.

._
. ‘
. ,dl
. .
. v
.
v
.
.
.
\ I u..
.
.
..
i.
.
.
.
.
. ,
1.
a
1 ‘
. ,
.
.
.2.
D J
l
I I
. .
.

18

retention of three eooortione were oignificootly euperior whu the
etrongoot eeeertion wee proeontod firet; that in. in onti-vclinox

order.” however. replioetione oi thet otody by Gilkinooo 35 31.

here not produced the tone route.”

nough otndioo of thie kind
hove invoetigetod offette of order of neeeege unite on receiver
cooprehoneioo. they hove gonorelly towered two or nore "etructwrod"
ordere.

A etody which doolt with once of the tone concepte lee thie otndy,
m which need e different opprooch. ie .‘m of Moreno. Petereon
tooh two peooegee fron e oociel otndioo hooh end nodifiod then for
”better orgeninetion." reteroon conoidorod fourteen itene in her
nodificetion of tho peeoegeo. end though eigniiicont roenlte were
ohteinod. it in inpooeihlo to dieoovor which or how my of thooo
weriehloo were reeponeihle for the incrooee in coqrohenoion toot
occivroe.n

The mining etndioe to he cited tell in the third clone; i.o..
tend to eepport tho proeut hnothooie. noniltoe eondnctod e oerieo

of minute in rooding end outed:

 

1’Iorold ”when. "no Roletiwe lifectiwoneoe of Cline:

end Anti-Cline: ordore in en erg-Interim Speech." w
W. XII We 1. t’“). 35"“o ’ ‘

”now-re ailhineon. s. 1. mm. end 1:. r. 9mm. ”lifeote
of Order end Authority in on Argumentetiwe Speech.” 92129.12

m; 25 m. :1. (Ann. 1954). 25-26.

1“Honor 1!. retoroon. m g; mdohilig in the Sociel V-
tnl ' (11. York: hereon of rublieetiono. 'fooohoro 6511....
to Uniworeity. 195‘), pp. 25-26..

19
It h. e e e M 2m tut f0! ‘11 ”jut. r.um
iron the ege of nine or ten up to netnrity the con-
etitnent elenente of content. 3.3.. the phreee. the
tentence, the peregreph. etc. , here eech e definite
neeenreeble welne ee en eid in the perception of
worde. the influenced theee factore weriee with
different individuele. with netnrity, with erection.

end with difterent eelectione of tending netter. the
letter werying directly with the perfection of eontent.’2

enerutly whet limit“. neene hy "perfection of content" ie the
degree at dependency between neeeese cite. for hie nethed wee to
rendoniee neeeeae unite of differnt eieee end to one“ the
”ewereae ninin. tending. tine per word." thee in. for e given
me. one mm: m. wee eeneted hy rue-nun. the
lemma. enother by undo-nun men-nu. m hy mum;
phreeee, end the tinel one hy tenet-hie. werde. end the ewereae
u.- ugme tereoosnieeeeehword inerenendwitheeehredction
in the eiee oi the etrnetnred unite}3

Snort for the welidity of Innilten'e tine index to toad in
e etndy hy lowee. Iowee chose the duetien threehold (teehietoeeopic
preeentetion) end recognition tine (eentinnene emanation) ere
inwereely rented to word prohehiiity. thie reenlt nee ohtelned by
eorreietiu ch taro indaee with note frequenciee whieh were ohteined

by eetnei word count.“

A _1. h A—Ak _._. A 4L

”tr-ecu union Benin-e. m We; Legtore .1! need ~ /
("Arehiwee e! teycholony.” Vol. I. no. 9; In Yorh: The 831;?
be“. 1907), y. 3:.

”2i!- . 14.

“Bowie 1!. Rowen. “the, Definition end ”want of Ice-d
Irohehility" (nynhliehed Hun. dinnertetion, Dept, of reyehology.
nerverd thiwnreity, 1950).

 

20

It it inferred iron theee two etndiee that the nine of the
«Intention! nit (Le... the length of the unbroken eeriee of deo
m M) efleete the conditionel prohebilitiee oi thoee
ei-ente. end thet the effect in enbeeeaently reflected in the
tine MW to mantle eeoh elw in eeqnence. Sane "port
ie provided (or thie inference by enother etudy. Miller end
tried-en. neina e very ehort peeeele with one letter omitted end
touting the percent right on the edjette' iiret m. conclude
that (l) the nore content the earn money. (2) eecereey ie
ereeteetwheneontent ieeyenetricelerou-dthedeletedchereeter.
end (3) left content ie eeeier to nee then right cutout.”

the niller end tried-en etudy elee ehowe the nhtimhie be-
twe. length of the depudent equate end enhjette’ ebility to
nehe prediction ebont nieein; unite. Another etedy which deele
with predictebiuty end eontextnel eonetreiet ie thet of teltridye.
ho peremhe iron e ehildren'e etory were reed to e group of
enhjeete; th— eeeh eehject wee eehed to [teen end write the next
we“. “jute nere then told the correct word end eehed to gene
the next. end no on. the note predictehility (yore-t of right
We) net .3322. “the low wee .05 for the firet word in en
independent clenee end the high .68 for the thirtenth word in en
indepndent elenee. (to few oleneee in the peeeeae conteined note
thee thirteen worde thet reenlte beyond thie point were urelieble.)

”George A. Killer end Iliubeth A. tried-n. "the teemtrnction

of lutileted halieh rate." W 533 M. I (September.
1957). 38-55.

21

I! e differ-at experiaeet mud in the eene theeie. with
leaning reeell ee the dqendut weriehle. the firet word in eech
elenee won the noet difficult to lent-n. end difficulty deereeeed
with the mount of content "needing the teeted word. A poeitive
reletion nee found heme. retell end th degree of contextul

deter-tendon.“

me finel etndy reinvent to the preent reeeereh deelt with
eeneenn-neteriel. 'thie etedy fond thee efijeete retelled nere
nowhenee eylle‘hlee free e not with e lower «crepe rote ug-
m pa ”1mm. [etc of inforeetion wee eentrolled hy
hewine the enhjecte leern pe'tterne in ednnee. the petterne
repreentin different denreee of oontentwel deteninetion. Up
to e tertein point e toeeteet mount of inferential wee learned.
hnt beyond thet point eylleble leernin. did not inereeee in
proportion to the redeetion in intention." 'fhie etedy mort-
the lelfridge teenlt on the reletien hetween predictehility end
leerninc rete. n hee the edditionel edwenme of «mutating.
the effect of etrnctnrein menu uteriel where there to little
poeeihility thet the teenlt eonld hewe he. predneed hy enother

weriehle.

 

“Jenifer eelfridge. "Inweetioetiene into the etrnctnre of

Verhel Content" (nnpnhliehed hoeore theeie. Dept. of reyehology,
hedeliffe cone... 1949).

17 '
may thorn end herbert tnbeneteie, "Intonation theory

end I-ediete Retell," 1 of W rezchglm.
xuv (septa-oer. 1952). 2 » a:

22

It ie contended thet theee tewernl etndiee, tehen ee e group,
leewe no doubt that lensnege ueere cen reepond to differencee in
home etrnctnre. If the "becenee" order it e etrnctwre of
which eon neere of mu» ere even. 1: eeene thet elteeetion
ef the dqendat reletionehipe enoe‘ eentencee to produce etrnctnree
which ere leee fenilier to the reepondent ehonl‘d rednce the econreoy
of prediction nede ebont the neeeeee. the hypotheti- thnt
eweeeeeiwe etepe of renowel fron e “hecenee” order will redwee
the eeenrecy of reepondente' prediction ehont the nieeins perte
of the neeeege ie therefore enbjeoted to en upiricel tent.

o.

cram 11

Thin chapter included the denign of the enperinent end e
deeeription of the technique of ehin‘ietr'etion.
333108 0! 1'!!! MIMI

the premt etedy wee deduced to more the effeete of
weryin; centence order on (I) ooqreheneion of the neeeese end
(2) the reletiwe redudency of the neeeege.

the independut reriehle in thie etndy in lent-re order.
the two dependent weriehlee ere (1) nonreheneion of the neeeene
ee neeenred by close ecoree end (2) the reletiwe' reddeney of
the neeeege ee eetineted free the nder end freoneney of different
rupee-u to than in e nntileted forn of the nee-ego.

In order to nehe prediction ehont differencee in expected
decree. tend order-treetnente of the neeee'e were ohteined by e
eyetenetie nethod. 'fhe firet order in e W order. Given e
"heoenee" etrett'ed neeeece. the next etep wee to control the
eerietion of etteetnte. A tendon order night hwe been weed. hut
there wee the poeeihility thet e tendon drewie. wonld prodnoe the
etreetured order or cone clone epproninetion of it. ‘l'hie wee
ohwiooely to be mided. the problen wee eon-het oonpliceted by
the foot thet there ere 128 orders thnt ere logieelly correct
(i.e., eny coordinete unite end their eubnnite one be interchenged).
Moreover. if thie order ney he called deductive. there in e like
“or of W ordere which ere eqully well etrnetnred end.

23

 

 

;\

26

poeeihly, eqnelly fenilier to noet neere of Englieh. Ewen
elinineting theee poeeibilitiee. there would be no ney of
eetleetiec the "dietente'? betwe. the rendon end etructered
ordcre and no beeie for nehins predictione ehont the reletiwe
effectimeee of nore than two ordere.

It wee obecrred thet novice the topic eeotenre to the niddle
of the neeeeae pieced the topic eat.“ on fer ee poeeihle tron
ite proper poeition in either e deductive or indentiwc pottern.
Having the two enhheedinge to the center of their eelle eleo
pleccd then on fer ee poeeihle fron their poeitione in e dednetire
or indentire ecqoence. It wee further oheerved tht tb fleet of
theee none required eeren binery' trenepoeitiene (“hence of
edjeccnt ectencee), end the eecond rec-ired eir nore. 'lhe
poeeibility of en ordinel eeele of etrnotore wee enpoeed. lone
triel end error chilling prodnced no order which wee e uni—
dietence (43 trenepoeitione) iron e dednctiwe or inductive order.
end eny fnrther trenepoeition novnd towerd one of theee ordere.
Three nore orderewere produced which ere et lenet 10. 27, end 35
trenepoeitione fron e deductive or indentiwe order. the new. en-
perinentel nenee'e ordere ere repreeented eynholicelly in lit. 1.

cm

Difference in difficulty enon. enpcrinentel treetnente reenltin;
fron neriehlee eoch ee word length end eentence length were controlled
by reins identicel eentencee in ell treetncnte.

to control for differencee in individual ebility to produce

cloee ecoree. e eecond neeecge wee prepered which netchod the cr-

 

 

 

 

""1515. l.“ The ecven ordere of the enperinentel neeeen

“ M # 01:13:33 A __ h w:
L:: 4 ,3 -g_ 3 4 2
me, i u r n m m w.
1 n n1w n1 ~.n..~. -_ 13- In,
n “I. .m. H? ' 141.24.14.32 M
W. . n2 1. n n. n ' :1
mi. .1: my ‘13 .- 131 131 an
n , . mo :31 m in 4m , 1m,
n1, n1 . m. m m, In ' n2
m 1.5. 1.3. r... r.s. . 41.3. Ls.
I! ,. n -, .tn. ,IIA. nu 3m 1m-
in m 1 .11“. all 13 IA . 13
an int .lm, In: “In: .. in: mm
In: mt : n. If h 1 4 I - 'I'
it} In V m. 4 m ' int. 1m 1131
ml ml a .mi , ml in, .» _ ~_ m u , 3
m: m: .1131, . gm: , m: . out. 1132'
'6 if“ ~, 13. 3677, i5. 35 43‘.

fl mg 5 03 RON can

 

(m ore repreeented by conwntionel outline eynhole to
ehew the lesion]. reletionehipe.) .

26

perinentel neeeoge in difficulty eccordin; to the fleech reedebility
formula. ‘ mtileted in the come way no the eacpcrincntel neeeege.

the control onstage wee eubnitted' to ell eobjecte elon; with one
order of tho exporincntel neesog-e. 7

Different“ in cloeeroon groope of eubjecte were controlled by
(l) limiting eobjecte to undergreduete etodente in the College of
Conciliation Arte end in the mm of eminence Shille
end (2) by giving etch cleeeroon group cppreninetely ncnel proportione
of the different enperinootel treatments.-

In eddition to the primry «richlee of thie etody the en-
perinenter wiehed to hnow if eohjecte could predict the decree thet
they would nehe on the teete; the in. if they were conecionely ewcre
of cny difficulty prednced by trenepeei‘ euteneee in e neeeese.

A five point ecele wee prepered on which enbjecte were eehed to renh
the teet beacon teey end Difficelt. Ae e poecible neeeere of
notirntion. two other ecclee. lntereetin. . . . mintereetinr end
denial . . . Worthleee. were edded. ‘lhe net of three ecclen wee
provided for both the control end the experinentel neeeegee.

- the pen of printed inetrectione nee eleo prepered. te-
etreotione. mu, end judgnent eeelee were reprodeced by
nineonreph. A typicel enperinentel pechet it pretented in
Appendix B.

Amara-runes or 113: mm

'l'eet pechete were node n of one inetrnction eheet. the control
neeecge, the encperinentel neeeeee. end e eheet of judsncnt cooled.
in thet'order. Inch teet peehet conteined only one trentnent of the

27

experientel neeeege. The teet peokete were errenged in eeriee, eo
thet eny eewen peokete need in eeriee would repreeent ell ordore of
the experimental neeeage. Subjecte were teeted during e regular
oleee period. the writer eerved ee experimenter in ell ceeee.

In the eerly pheeee of teeting, eewernl gredoete etudente were
inedwertently teeted. They were dropped from the eenple for the
eehe of hoeogeneity end their teet peckete were repleced. One
h-dred forty undergreduetee were teeted. end theee were dietributed
in eewen gronpe of twenty eeoh: therefore, it wee poeeible to do the
enelyeie with eqnel _e_’e without diecerding qualified eybjeote. '
senate were etedente in the department of Cmic’etion sun. end
the College of Co-wnieetion Arte, Billet School, Hiehigen Stete
Dniwereity. the euple wee emceed of 561 freehaen, 191. eophomoree,
121 juiore, end 137. eeniore. Seventy-two pereent of the enhjecte
were melee, end the eeen ege wee 21.8 yeere.

upon entering the oleeeroon, the enbjeete were infer-ed thet
they were to teke e teet which wee e part of e theter’e thoeie er-
periloent, end which wee, in feet, a mm of their reading ebility.
they were told, "Since thie experiment nearer wolnnteer enbjecte,
eeyeee who doee not wieh to take pert eey leewe without penelty."
lone left. In eighty perent of the oeeee eebjeete were told by
their inetruotore thet the teet eooree would be weed dnwetioelly
to pinpoint eny pertienler reeding difficulty, but in ell eeeee
euhjette were told thet their ecoree would not count direetly on
their coeree rude. 1

, The teet when were peered out by the experimenter, end the

enbjette were eeked to reed the firet page of printed inetructione.

e p e \l e .‘
.1

1 l
I. .
e
. e
u .
_
.e
O
.l:
o .
\ .fi
3
.
~
5
e
.
I,
. x
. .
l
. v i
. l
u I
. A
e I
t
u .a re
.
s
I.
_ .
v
I u
.

28

Pencile were furniehed by the experimenter. After everyone hed
eufficimt time to reed the inetructione, the experimenter went
over the inetructione rapidly culling ettention to the etetenent.
”one word end only one word in eech hlenk," end edding thet thie
could he eny word .. lerge or email. Stheete were told thet they
would have ten ninetee, end thet they would he given e five end e
one eiuute wernina. (lhe tine werninge were given heceuee it wee
felt thet one'e ehility to eetinete elepeed tine ehould not he in-
cluded er en experimentel veriehle.) Subjecte were then eehed if
they hed eueetione ehout the teet procedure. There were not nore
then two in eny group, end ee eoon ee ell euhjecte were ectiefied
with the inetructione they were told to begin on eeeeege nth-her
one. which wee the em in ell cheer-the control neeeege.

At the end of five einutee euhjecte were told thet their tile
wee helf pone. end it wee euueeted thet they try to get through
the acetate to tehe edwentege of the eeey hlenhe. Il'lle one minute
wereing wee given without current. At the end of ten ninntee the
euhjeete were told to finieh the word they were writing end etop.
After the firet teet euhjecte were egein eeked for queetione.
there were none.

in preperetion for the experinentel neeeege) euhjecte were
ceutioned not to he dieturhed hy the differ-t errengeeent of blank-.23

28Since the eeee worde were deleted in ell treetnente.
treeepoeing the eentencce gave come treetmte e eonewhet different
eppeerenee than the control meeeege. It is held thet thie wee not
e confounding variehle. einoe the within-sentence reletionehipe
rnein conetent.

29

After epproeinetely one einnte they were told to eontinee with the
eeecnd eeeeexe. the eeee timing procedure wee follewed ee in the
previoee teet.

After the eeeoed eeeeege hed been completed the ewbjecte were
eehedtotnreto the leetpexeof thepechetendeerhepointoe
the eeelee to repreeeet their feelinge towerd the neeeezee end the
teete. A few edjecte related to the «yeti-mentor in the die0
eeeeioe which followed thet they hed epplied the hey . . . Difficult
eeele to the teet. the Intereetinc . . . Unintereeties ecele to the
ewhjeet netter. end the ueefel . . . Iorthleee eeele to the projected
reewlte of the etwdy.

while the euhjeete worked the eeelee, the experieeeter die-
triheted red mile. the eebjette were then told thet they eoeld
to heck ever the teete eed eehe eey edditieee or eerreetioee they
wiehd. fieywerefwrther ieetrwetedwettoereeeeeyof their
previene enewere het to write eecoed choicee end fill oet blenke
thet hed not he. ecqleted duiee the tiled intervel. After ehewt
for wintee eleeet ell edjeete hed eteeeed writing end indieeted
thet they eowld do no were if riv- eore tile. (he eleeeroee no.
thet hed hed previowe experience with eloee precedere etopped
writing before the .d of the tiled period end refneed edditionel
tine. Since ell We were eewelly repreeeeted. the ecoree
of thie crew were reteieed.

Aw entreeeowe veriehle wee ineerted by chenee with one group
of twelve eehjecte. In the leet ten eeconde of the firet timed
period the lighte went out. After e elightly prolonged reet intervel.

30

the ewhjette indieeted thet they hed edepted to the eon-thet din
conditiene end were enrioee to so on. The experiment proceeded
without further interruption. In the dieeeeeion which followed
there wee eone ecu-lent eboet the voice of e teet of one'e ehility
to reed in the derh. but there were no epperent different“ in the
eeoree of thie group end thoee of eny other group. the mtedneee
of thie eitnetion did not ee- ewffieint eeuee for dropping theee
dete free the «merit-eat.
room 0! m DATA

[or the clone eeore only erect replee-ut of the deleted word
wee coated ee correct. liner irreewleritiee in epellin were not
counted off. but cheape- in o‘er or teeee of verhe were coneidered
ineorreet.‘ Since reylor hed indiceted there wee nothing to he geined,
no ettqt wee node to eweluete or enore eynenyee.29. For the
neeewre of entropy, the frequency of eeeh differut word wee re—
corded for eech hlenh. min. verietione in epellinp were not
eouted ee different unleee the verietione produced eeother poeeihle
word; e4" ehenge in water or form. Conletely ureedehle worde
were ecored no different "Iron." enewere ee were blenhe. the tined
ecoree were tek- ee the heet eetinete einee the tine linit
nininined reorpeninetion hy the efijeot. In the "entre tine"
period enhjeete hed been inetrueted to nehe correction end fill eny
reneining hlenhe with e red pueil. untined etoree were obteined
by eddinp the correct red enewu-e to the tined ecore end eubtrectinp

2"’Reoent Develop-onto in the Bee of Clone rrocedure," 48.

31

the out: of worde thet hed m incorrectly chenged. nu very
low proportion of the incorrect choosee (leee then 1% of totel
chengee) meet" high degree of certeinty ecconpenyins e

correct eloeue.

CHAPTER III

‘l‘hie chepter reporte reeulte of e etetieticel enelyeie of

the «Inn.

W838

hiricel evidence for or egeiut the theoreticel hypotheeie

of thie etudy ie to he obteined by teeting eevcn etetietieel

hypotheeee. Stored in null fore they ere:

1.

1.

3.

b.

5.

7.

the ween clore eooree on the control neeeege
for the eeven treetnent-aroupe ere the cone.

(i.e.. fi§=fi§= ' --= '3‘?)

The neen cloee ecoree on the eeven timed ex-
perinentel teete ere the cone.

(1.... F1: ‘-'-' fig : - n.3,- fig)

Interection between individuel ebility end
the etrweture verieble ie cero.

(Len fija“fij.'fi_,j + [‘7‘ c. O)

the neon cloee ecoree on the nntined ex-
perimentel teete ere the none.

“0". M1 3 F17. '3 "'=fi7)

'fhe neen entropy ecoree for the eeven
uperinentel neeeezee are the cone.

(t-.0e Fl; 3 Hz: ‘ ‘ ‘3 R7)

lenh order correlntion between "trenepoeition"
renhe end "clone" renke equle eero.

lenh order eorreletion between "trenepoeit ion"
renhe end "entropy" route of the eeven treet-
nnte equele zero.

32

the .05 level of eipnifieence wee choeen to teet thoee hypotheeee.

rheneenecoreeonwhiehthieenelyeieiebeeedere thee-intehlei.

TABLI 2.--'ihe men decree on the Control end merino-tel neeeesee

for See- Groupe of Mjeete .

 

ctorr econte
Control

tiled

Inperinentel

tined
utined
name!

IILIIIVI
INTRO?!

hlLKIIVl

IILNTIVI
HUBDIRICTIOI

23.3
It.’

21.:
21.7

.888
.512

.138

I!

22.8

23.8

19.: ’

19.7

.536
.414

.100

II!

11.8

83.1

17.8
18.3

.371

.. 572

.428

.133

25.:
26.1

18.8.
19.1. '

.562

' .Q38 »

.118

22.8
88.3

18.3

18.8

.381

.333

.467

.157

"VI

33.!
83.8

18.4

19.2

.383

.520

.480

.166

711

21.3
12.9

'15.3

13.8

.319

'.611

.389

.146

ANAL?!“

Before teetinp for differencee enong neene with the enelyeie
of verience it wee neeeeecry to teet the eeeuption of honoaencity
of new. the hertlett tent for honogencity me epplied to
tb timed clone come for the control neeeege.‘ the chi-centre
obteined (table 3) did not per-it th rejection of the hypotheeie
thet the veriencee ere eqwel.

mu 3.»!ertlett ten for lonoguity of Verience Anon.
“moteleeeleeeeeenthemetrelleeeepe

A A

v—ri —v

 

 

 

 

atone N1 N1 n1 51 1 81" 103.. 5." 71,-;
1 20 19 135.15 23.25 1.11913 .05233
1 2o 19 ' 195.31 11.33 1.32130 .0523:
3 20 19 310.19 33.11 1.52133 .0523:
1 20 19 311.39 13.12 1.21531 .0523:
5 20 19 555.13 29.22 1.13553 .0523:
3 20 19 333.02 33.53 1.52303 .0523:
.L 29 15 A 73242 31.91 141331 .0g_2_3_3_
n- 110 133 , 1130.05 10.35113 .33311
C = 1.02005 ' B' = 1.9213 B: 1.32 [711.95 ’ (3) : 12.3]

1.L.__. - .___._ 4.‘.. 'Ai .L

lhe next etcp wee to teet hypotheeie ember one-"the nenn clone
ecoree on the control neeeege for the eeven treetnent-proupe ere the
cone-with en enelyeie of verianco. einale verieble dceizn. the

reeulte ere reported in Table I1. ' The obteined g feile to reject

the hypotheeie thet the neene ere equel.

35

TABLE Mr-Anelyeie of Verience of the Control Mceeege

 

 

 

 

81- of Keen
. 811111130 gf m - F E
1...... Hunt 169 6 . 18.17 1.09 -
Within ' 35;; 1;; 25.73
Totel 3594 139

 

Ae fer ce the overell perpoee of the experinent ie concerned.
the feilure to reject thoee hypotheeee indicetee thet the gronpe
were effectively notched in chility.

In the enelyeie of the crperinentel ecoree it wee epein neeeeeery
to tent for homogeneity of verience. the tertiett tent of humanity
wee epplied to the experinutel ti‘d ecoree. Although the reuse of
veriente ie eonuhet lerper then on the control neeeeee the obteined
ehi-eqeere ie not eipnifieene et the .05 level end eeein feile to
reject the hypotheeie thet the variencee ere eeeel (table 5).

Given thie evidence of honogeneity it in poeeible to cent the
eecond hypotheeie--the neen .1... ecoree on the eevu tined er-
perinentel teete ere the eene-mwith en enelyeie of verience deeign.

A einple verieble deeign yielde en 2 of 1.67 end the hypotheeie ie
rejected ct the .05 level of confiduce (rehle 6).

‘l'AhLl Sulertlett rent for Honogeneity of Verienee Anon; Seven
‘ Croupe of Clone Scoree on the Irperinentel fined Heeeege

 

Group Ni 11; n1 5;," 51" 1.031051" -%
1 20 19 295.00 15.523 1.19103 .05233
2 20 19 113.15 10.123 1.31090 .05233
3 20 19 590.55 31.031 1.19219 .05233
1 , 2o 19 152.55 _ 23.313 1.31390 .05233
5 20 19 210.20 . 12.1112 1.10131 .05233
3 20 19 , 120.10 22.111 1.31531 .05233
1 19 10. 1 9 33

3- 110 133 3393.01 9.33225 .33311

C = 1.02005 8’: 9.31213 9: 9.19 [)11L .95 (31:12.33

 

'fAhLl 6--Anelyeie ef Verienee «1 hperinentel . fined ecoree

 

 

 

 

 

I. of then

£12.11! m F .1 p
rumor 407 .95 67 .9! 3. 67 (.05
within 3393.05 133 2 21.3;
I'otel 3801 .00

Since clone team on the tined control neeeese were fond to

correlete .58 (< .01) with tincd expcrinentel ecoree. it wee thought

edvieeble to rueove thet portion of verience which could be ettrir

huted to individuel ebility to produce clone ecoree. In the two

wey cleeeificetion need in thie teet it wee poeeihle to tect three

eeperete hypotheeee 1

37

(e) the row neene ere eqwel (i.e.. there to no difference
ettributehle to the control verieble).

(b) the col-1 neene ere equel (i.c.. there in no
difference enen. erect-cute).-

(c) Interection ie nero (i.e.. high ebility end low
ehility euhjecte ere equelly effected by the en-

perinntel veriehle).
Ihe enperinentel-tined clone ecoree were divided into two

eectione. One eection wee conpoeed of thoee individuele who ecored
nore then 13 on th control tent. ("high" creep). end the other of
.11... .1. ecored 23 or 1... ("13."). 111. 11.1.1... g... epprooiinetely
eqeel frequenciee in the fourteen cub-elected. The conpntetion
eeedtheneene oftheeeh-eleeeeeee ecoree. thereeulte ofthe

teet (table 7) reject hypotheeee (3) end (b) but foil to reject (c).
the conelueione ere: (1) the come of the high grow ere diffcrnt
frontheneene of the lewproq. (2)Iecneprodueedhytheeewen
enperinentel treetnente ere different. (3) there in no eignificent

interection between individuel ehility end neeeeze etructure.

mu 7.--Anelyeie of Verience of prerinentelo'rined Scarce Divided
on the Deeie of accrue mode on the Control Heeeege (23 end below I

 

 

 

 

nore then 23)

8. of linen
.. Wf mu F J-
Individual
Differencee 82.09 1 81.09 110.13 < .01
treetnut 39.62 6 6.60 3.25 < .01
Interection 1.56 6 .76 .37 --
totel 126.21

Irror 356.07 116 1.03

 

38
The cone operetione were perforned on the mined enperinutel
ecoree with elnoet identieel reeulte. the mince were nee on
the heeie of the wntinnd control ecoree. one group being thoee*who
ecored 19 or leee en the attend control end the other thoee who
ecornd wore then 13.

run t.--Anelyeie of Vericnee «1 lrperineeul-Uetined ecoree

 

 

 

 

Bun of linen '
‘ £322!!! df' W_5II2" F' P
front-Int, 211.11 3 32.33 2.13 < .05
31.11. 3153.00 133 23.35

fuel 38 13 . 1‘ 139

SAIL! 9.«Anelyeie of Verience of leperinentel-htind leoree Divided
entheneeie ofteoreenedeontheIhtinedControllcee-epe (Bend
helew/ lore then 35)

A LA

 

 

 

 

t. of linen

m 4* 3239. F P -
Individual
0111...... 59.51 1 59.51 11.31 < .01
rte-nee. 33.02 ‘ 3 3.00 1.11 < .01
Interection 7.10 6 1.19 .83 ..
2.1.1 102.11

trror 181 . 36 136 1 .M

 

 

 

39
In on effort to pinpoint the coerce of the differcntee ewe-s
fleet. g-teete were epplied' to the edjeeent .nne of the ennui-entel
ecoree. the reeulte of the trteete eppeer in ration 10 end 11. the
totel m end the two eubgroupe were tented individully. Hone
of the _t_'e ere eigeificent et the .03 level.

run 10.--t ~‘feete Between Adjeoect fleece-onperiuntel 'fined Clone

 

 

ecoree

1.9323 W 1» W , W 35...;
1 e 2 1.31 33 -- .19 11 «- .31 19 --
2 e 3 .11 33 -- .52 13 W -- .51 13 --
3 3 1 - .30 I 33 -- ~ .31 2o -- .13 13 «~-
1 3 5 .11 33 -- 1.01 20 -- -1.11 13 --
5 3 3 - .15 33 -- - .33 I) 11 -- ¢ .53 19 --
3 c 1 1.33 33 -- .11 13 -- 1.12 20 ~-

 

rim 11.--1-re.t. letween Adjecut inane-denounce“). noticed

 

 

Clone Scored
m W . W . We. a
1 h 8 .85 38 u .60 19 or .96 17 ~-
3 d: 3 J15 38 .. .08 17 r- .841 19 --
3 e 1 - .39 38 -- .21 20 .. .33 l6 ..
6 I. 3 .31 38 u .86 20 u .97 16 ~-
3 e 6 - .38 38 -- - .55 16 -- ~ .34 20 ~-

6 ‘ 1 In“ 38 .. 1o68 17 no. le:6 19 --

__L 4—.

40

(he of the prinry hypotheeee of thie etudy wee that noon
entropy (freedom of choice) ecoree for the eeven expermntal
ordere would lncreeee in direct reletlon to the number of tune--
poeltlone tron e ”heteuee” order. (The null version of thie
hypotheele 1e thet the neon entropy ecoree for the men ex-
pert-entel treatmente ere the em.) The enalyeie of werlenoe
wee egele need to teet for different“ anon; mot. A two-way
tleeelfleetlon, by hlenhe end by treeteente, wee need. there
were forty-eight blank- end eeven treeteeote. tech blank to the
eeee for yell eeven treeteente. Each blend: in eeeh treetnent
repreeeete the reeponeee of twenty «Moon. The reoulte reported
in Teble l: ledioete thet there were different“ among blanks and
no” treet‘te whloh would octetleee then 11 of the tine by
theme. the null hypotheele 1e rejected.

mu l!.--Anelyele of Valence of Entropy Scoree -- bro-way
Cleulfleetloe by Blanke end treat-ate

 

 

 

 

81. of linen

£29.22 di ms- F P
llenhe 17.5631 47 .3737 40.18 < .01
treat-eat .5088 6 .0343 9.12 < .01
leeueel 2.3313 133 .0033
Totel 20.6932 335

 

11- to the 31:33:13. of theee dltferenoee, 11;. 2 then :11- neen
reel-deeoy ecoree for the eewen ordere end llleetretee the relation-
ehlp between reletlve redundancy (Rm. close proportion (C P). and

eleelreetloe (M). (Relettve redundancy to l - relative entropy.)

41

1001

80

70

60

50

”cu-«“0.

30*

30

20

 

10

 

tip. 1.- merino of eee- eeletiwe rednndency (RR).
clone proportion (CF), end eiedireotioe (r1) for the eewen
treat-tote

42

The" eeeeeee. treetnente one be renhed eeeerding to ditticnlty
by (1) ”er of trenepoeitione free e "beceoee" order. (2) obtained
eloee ecoree, end (3) reletive tedundeney ecoree. Couputetion of
rent order correletion eoeffioiente ehowe e dignifieent reletioe be-
tween (l) end (2) (g: .75, p<.05). there in no evidence of e
«1.113. between (1) end (3) (g: .29) or (2) end (3) (q. .54).

To eoqplete the enelyeie of the date. e hereon 3; wee eompnted
to deter-line the reletion between e eubjeet'e clone eoore end hie
ecore on eeeh of the five-point eweleetioe eoelee. the melee were
coded S for the noet fevoreble reepoeeeu-Beey, Internetinz, end
Ueefnlnend l for the noet enfevorehle reepoeee. The reenlte re-
ported in teble 13 provide no eviduee of e reletion between e
enbject'e eveloetion of the eeeeeae on theee three dineneione end
the clone ecore which he prodeted.

mu 13.-~ Correletioe between Clone ecoree end teelee of teen.
Intereet, and Utility

41_.

 

£9.11: A - 3 A V P
Beey . . . nitfieelt .063 --
Intereetila . . . mun; .0511 --
Eeeful . . . Worthleee .151 --

Ae reported in reble 13. there were no eignifieent differueee
none the eeven treete-t-zroepe on neee retinge of eeeineee.

intereet. or ntility.

43

TAIL! lA.--Anelyeie of Veriente among treeteente for Three

Jude-at Ste lee

 

Scele 1 l A 1 5 5 6

 

3-3 3.0 2.2- 2.3 2.1 :2.1 2.3

I - 0 3.9 - 2.9 3.6 3.6 3.6 3.8
U ‘ W 3.5 _ Se‘ 3.5 ’05 3.6 ‘ ,e6

. - 7 F. P
1.6 1.10 --
13.0 1.54 ~-
3.0 .91 -

 

MN

Thin chepter inclwdee e mry of the reeeerth, conclneione,
dieceeeion. cone obeervetione on Infomtion theory, end mention
for further reeeerch.
cm

the preeent etndy in en etteqt to enelyee the releticeehip
between centence order end neeeege cocpreheneion. lore eeectly,
it ie en inweetigetion of the reletionehip between etretterel
werietion end orgenieetioe. A neeeege wee prepered which net ell
the requires-tn of e "becenee" ordu' oetline. Bin elternetive
eeeeesee were prepered which were eeeigned predicted renhe of
difficulty in terne of trenepoeitione of eentencee. for ample.
the noet difficult. or eevuth renting eeeeege. wee 43 trenepoeitione
iron the firet renting neeeege. ‘l'he eeven treetnente were then
eetileted end prepered eccordinx to the require-note of clone pro-v
cednre. Clone ecoree were eubeeqnently obteined for the eeven tren-
nente free eeven rendonly eelected grown of enbjecte. An eetinete
of Cb reletive -tropy of the eewnn treeteente wee obteined by e
procedure outlined by Hilton L. Taylor. Judgmente of the perceived
difficulty. intereet. end ntility of the treetnmte were eleo obteined
from the eubjecte. Mean cloee end entropy ecoree were tempered by
eeene of enelyeie of rerience, end correletion coefficiente computed
for clone ecoree end the jungInte of difficulty, intereet, end
utility. leek order correletione were eleo computed between the

Al;

65
neen clone end entropy ecoree end the predicted renhe of difficulty.
calcium-13

l. there in e eigeitieent reletion between clone ecoree end
trenepoeitione iron the "bounce" order of eenteneee. 'lhie coe-
ninth. in heeed on the harm-um of 3133111333: diiiereneee in
eeen clone ecoree none order-mute end eisnilicent rent order
correletioe between clone reehe end trenepoeition renhe. however.
thereletioedoeeleeweeceethingtohedeeiredinthetrteete
hem-“Mannereeoteienitieentendtherenhcrder
cotteletion coefficient ie only .13.

3. there ere eignificqt 312m...- nee. eeen reletive
entropy ecoree (or the eewee treeteente. bet the reek-order.
correletion hone. reletiwe entropy end treeepoeitione tron e
W“ order it not eigeilimt.

3. ihere u no menu of e reletiee um... e enhject'e
eleee eoore end hie jedgneete of diffienlty. intereet‘ or etility.
nor in there e treetnent difference neon: neen jndznnte on eny
of the mice.
aircrew:-

the reletion between clone ecoree end the trenepoeitim hon
the "heeeeee" order. thoezh etetieticelly eignifieent. done not
per-it the coeclneion thet eny trenepoeitice of eeetencee will here
the cone effect on the eoqrehueion of e eeeeeze. Inch e oonclneion
would neeeeurily nee-e thet there in no poeeihility of two ordere
of the teen eenteneee being eenelly well etrectnred. end thet the
one degree of dep-doey obteine bet-en ell peire of edjeeent

e6

eateneee. the-titer oitheeeeeewatiene ie deeigdby theaw.
ietaee of 113 poeeihle ”beau" erdeee of the eataeee need in
nu. aperineet- the eeccnd is denied by «name... 2- thet the
”heceeee" order penite W eataeee which eey he. trenepoeed
withoet nodilyinu the etrnetwre e! theaeeese. It would eea thet
the teehine of coating traepoeiticee iron 21 dednetive or in-
dectiwe order woeld in feet control th. ehowe-eatienedteotore; ,
however. the pertienler ordere developed (or thie reeeerch epeciiied
the W e! the trenepoeitieee ad did not pee-it elterettiwe
cedere to occur.

1b trenaoeitionel ida oi weryiu etreetere wee eeleeted tron
e o‘er of elteenetiwe nethode heeeeeet (1) it hed th potatiel
odaerdieeleeeleed etrateleenda) iteeaedtohetheeoet
lihelytoheoteaeeeeeteidethqeri-etel (tea. nth
heeie of teeth: thinking. end en aelyeie of tho teeelte ct thie
reeeereh. en elternete nethod ie propoeed on e tool for further
reeeerehaieheonldreenireaehleeeeflortonthepertotth
enperinater end elininnte the bin oi tree-poem; eeleeted eatencee.

letceeeiwe eppreniaticee to e .iwa order conld he obtehed in
thie wey. ltertin; with e given order of neeeege aite, every n“
alt (letter. word.- eatenee. uperegreph) ie deleted free tb not
end reineerted in M order. the “and epprneihtlon it oh-
teined by rendceiein. every (13-11“ ait. .... ”n. eiee of 3 would
he deternined by the naber of unite in the eet end the ember ot
epproninetione deeired. the (incl order would of conree be e tendon

errensenent o1 ell unite. of cine 1... in the neeeege. the thence of

47

reproducing the originel order ie coneidered negligible. end thie
procedure hee the edded edventese of being epplicehle to neny hinde
of exietm neeeegee. Such e procedure would, by definition, pro-
vide en ordinel eeele of etructure providing only thet cone
etructure obteined in the origin-l neeeepe.
80M! marine (I Nocturne 'flIORY

hiring the proceed of enelyeina end reporting the preeent reeeerch
the euthoe hee ade tone obeervetione ehoet Monetion theory. cone
of which ere directly releted to the preeat etndy end cone of which
ere of e nore generel neture. ween obeerwetione ere reported here
for whetever benefit they any be to other co—nanicetion reeeerchere.

'l'he firet of theee obeervetione ie thet ‘reylor'e epplicetioe
of Inforation 'lheory ohteine e bieeed eetiate of the reletive
entropy in e eyeta. in thet hie eetinete of axin- atropy ie
depend-1t on the amber of eubjecte giving reepoeeee. the biee ie
releted to the feet thet, in Teylor'e celculetione, "choice" in
defined differently for the conputetione of ebeolute end £1.92.
entrgpz. Choice in defined for ebeolute entropy ee e differat
word; i.e.. the number of choicee ie the nunber of different worde
ectuelly eupplied for e given blenk. tor nonin- entropy, houewer.
choice in defined no the number of eubjecte; i.e.. the number of
choicee ie the amber of different worde thet hypotheticelly £0315

occur g all .93 eubjecte geve different none; .30

”1111»: um reylor. "Applitetioe of ‘Cloee' end Intropy
lleeeuree to the Study of Conteetuel Conetreint in Senplee of Con-
tinuoue Prone" (unpubliehed lh.D. dieeertetion, The Univereity
of Illinoie. 1956) pp. 17-18.

48
In no example. neing en 3 of 100. 'i'eylor pointe out hie own
error.

If 96 out of 100 eubjecte chooee the eene word. e co?-

peretiwely lerge aount of orgenieetion [redundeney_
nuet be inferred even if the other four ell chooee

different other worde end the outer of elternetivee
heconee five. However. if eech of the five hinde

were choeen by twenty enbjeete. ebeolute entropy

would reach ite axia- welue for thet nutter of hinde. 31

thie enelyeie eeae to egree with Inever'e etetaent.
If one rechone, for thie once, [- two choice citation]
the naericel welee of 13, it turne out thet ! hee ite
lergeet velue. nenely one. when the two neeeegee ere 31
eenelly prohehle; thet ie to 'eey wha Pp'Pz =7 ’5 . . . .
In epplicetion. however. teylor ignoree enother etetaeet of
Ieever'e.
thathenaherof‘eeeee in fined.wehewejueteeen

thet tha the infornetion it greeter, the lore neerly
eeuel ere the prohehilitiee of the verioue ceeee.

m__ it enoggr gotten; m g; Heine dig I, m

JW 2; er of £532. More eccuretely,
gag—:17 W.¥t+¥%_.%‘tfifi.%§”
the prectied of defining "choice" differently for the naeretor
end dacnineter of the proportion "reletiwe atropy" reeulte.
tha. in e low eetiate of the reletiwe entropy of e eyeta.
‘l‘o daonetrete the eetat of thie hiee the following prephe
hove been prepered. tn theee figured. the proportion of reletire

entropyiecoauted foreatofhypotheticelceeee. 'lheeeeee

 

,1”. g ’3 l’e

”Claude I. thennoe end Inna Hawer. 1!! Methennticel
15:35.0 of (imitation. (Bthenea the Univereity of Illinoie
u.

. 1.
”gig” p. 103

49
ere the none in both figures (3 end 6). Figure 3 1: obtained
neing 'feylor' e fornnle. end figure 4 neing the corrected fox-mule
which definee M! in ell ceeee en the outer of differut worde
"plied hy the euhjetten 'l‘he terticel dineneion of the grephe in
the proportion of reletive entropy. the horieontel dineneion ie e
hypotheticel clone ecore. Linee A; 3, end c 1.5. figure : ere
computed for a...‘ eieee a «.91.. 3 quel- 10‘. 3 «an to.
.‘n 3 m1. 100. m an». im- the anaemia.“ thet
only ten different enuere ere given, one of which ie "right" end
one "wrong". the A in fignre 4‘ nehee the nose neeqtione end ie
the eene for ell ye.“ Linen x. I. end I. in both ceeee. eeeI-e
e given clone proportion end thet til the wrong enewere ere
different once; i.e.. en the clone proportion increeeee the nun-her
of thoitee decreeeee. the principle 4mm in to he oheerved
between linen A. I. end c (fig. 3) end line A (113A), where the
clone proportion eenele 501..

Another obeervetion thet ten he nede ie thet whet rnylor celle
W in Jnet enother eoeroe of entropy. In tent. if re-
mining reletive entropy ie counted eocording to reylor'e.formle.
neing the corrected figure for millet entropy. the Inn of re-
neining reletive entropy end the clone proportion ney encud unity.
lhet night he coneidered niedirection, neing the corrected fornnlee.
ie the proportion of reletive redudency ninne the clone proportion.
In the onee where clone enceede redundancy, niedirection would.

 

“file ie the eene curve which Shennon preeente for e two-choice
eituttnn with prohehilitiee of (p) end (l-P). 2 .. p. 20.

~anon-inn neuannn

 

o to
omzosoeosocomooeomot
mm

 

L

Fig. 3."!nylor’e teltnlntion of Intropy for two idenl ceeee
end 3'e of 10. 80. end 100

IC'UOII-II' C‘HHDFI'4

1““

70

50

10

31

 

 

\

Y
2
F
A
0 IO‘N‘JO'UM‘GO'7OIOW

Dig. uncorrected eeleeletion of Intropy for two ideel

ceeee endg'e of 10. to. end 1”

1m

52
then, heve e negative velue; i.e., direction. In the 336 can.
putatione in thie etudy. the cloze proportion did exceed the re-
dundancy figure in e majority of the canoe £11521 _t_l_z_g corrected
fmlee m 9393. Fig. 5 shone three lines, the mean scores for
the eeven treatmente, ehowing the relationship between the clone
proportion and the average relative redundancy as computed by the
two nethode previonely described. It'will‘be noted that the averege
niedirection, or defined ebove, has disappeared in the revised com.
pntetione.

Fig. 5 eleo ehowe that there ie mch cloeer egreenent on the
relationship omens treatmente between cloze proportion and the re-
vieed estimate of reletive redundancy. Rank correlation between
clone end redundancy neeeuree in now significant et the .05 level.
the correlation between redundancy ranks end predicted renke, however,
etill done not ettein eignificence.

the two~way cleeeificetion enelyeie of varience wee repeated for
the revieed entropy ecoree. end egein differencee were eignificent
between blenhe end treetnente (9 4 .05). There ie. therefore. no
cheese to be node in the conclueione of the etudy ee eteted en pege
d5 of thie peper. If. however, one viehee to ettribute come
ebeolnte'velne to the proportion of relative entropy or redundancy
in e eituetion. the differencee in the reeulte of the two methode
of conpueetion could become extremely eignificent.

the final obeervetion to be made hot e purely theoreticel
velne (if eny) end may eeen inconeietent with the preceding die-

cueeion. Entropy. or information, ee defined by Shennon hes

”1002‘

90.

70

_Ta.

\‘ .

 

 

10"

 

o J _n 4‘ A L A ‘ 1

1.1 3.‘ 5 6 7

Fig. 5.. «Coqerieon of clone proportion (GP) end reletiwe
redundancy ee computed hy 'reylor'e method (TR) and the corrected

WCCR)

54
nothing to do with denotetive neenina, but ie concerned only with
the uncerteinty of the neuron ee to whet the next M in going

to he.35

low. if it cen be eeeued thet the eonrce hen e pnrpoee.

e neening to comicete (we epeeh now of e hinen eource), thie
pnrpoee will in neny teeee linit the choicee eveileble to him, the
reducing the entropy of the neuron. whet ebont the one. halvever.
inwhichtheeource in fecedwithmme?
thin in e eituetion in which it eeene e neefnl dietinction could be
nede between redundeney end orgenieetion. rent it, “tithe eitnetion
1m deecrihed, redundency be e velne of eero. .m... could be
eeid thet orgeninetion in unity. m 3.3 w m 9222;.

mangemmmmmm. If th- vie-point
ie ewitehed to that of the receiver. the receiver in the proceee of

receiving e neeeege ney be quite note of the neening intuded hot
nerinelly “detain ehont the eynhol to he need. In feet, it ie on
the heeie of thie diecrepency thet e receiver in eble to "correct"
e ecoree. Another eitution in which thie dietinction holde--be-
tween redinldenoy end orgenieetionuie efter e neeeege ie gmleged.
Once the eynbole ere eelected (for the ecoree) or received without
interreptionChythereeeiver) thereieggmimeboutwhet
the enhole 3” (no entropy). hut there any be e greet deel of n-
tertninty ebont their intended neeninse (low orgenieetion). For
theee reeeone. it ie qneetionehle whether the fignree obteined in
thie etudy. end celled atropy end red-ninety. ere in eny erect

 

,’Mep ”e ’9‘IWe

55
eenee letter-en of infatuation in the naming intended by the de-
velopere of Inforution Theory. They ere epperently, however,
eeeenree of eoee peychologieel phenomenon which werrente careful
urttiny end exteneive invettigetion.
8mm '0! rm: rum

1, the reletinnehi, hetveen "entropy" end "organization"

neede to he thoroughly explored” hath. theoreticelly'end experimentelly.
' 1. A eeriee of camperetive ethdiee of the verione methode oi '

errencie; eeeeegee could he conducted elm, the linen of the preeent

etedy. loch e eeriee ie needed to never the Quention, "meet ie

the heet my to errenge e me?"

3. A eeriee oi etndiee ie needed to eetehlieh the reletion
honed-mend interectioelot-inrder effecte'et the verioee levele of
neeeege eoeetruction. where levele ere defined in tenu oi else of
unit under inneeuetiee.

e. Order eiteete coeld he inveetigeted ecroee leagues“. or
non; eehjeete of different celtnrel hechgrounde'vho epeek the "cane"
linen-rd.

the pretent etedy hee ehnve thet the deletion-coupletion method
(clone procedure) in eeneitive to the order veriehle. Goehinetion
of thie tool with the rendoeieinsrevery- n“ wait eethod of ohteining
epproxinetione to etrnctnre euggeete e ire-e within vhich the

muted inveetigetione one he conducted.

 

‘emxomem

thorn. Iurrey. endluhenetein. lerhert. ”Interaction theory
. end lunediete hecell. ”‘ 123ml o urinate; ~
mm. xuv (Sept-her.1952. 260-266.

heighley. K. c. "An leperinentel etudy of the “tented roer

Speech Varietlee on Lintunr Wei“, " &
£322.11: (love-her. 1952). ”-158.

Iryent. loneld e... ... ueueee. lerll. Wfiw

m. “ York: Appleton-Outer, Country. the”
1947.

Getter. licherd 1. "writing controvereiel Itoriee 2..

Wm." magma. um (en-r.
1955). 319-313.“

hele. Idxer. end then, Jeemn I. "A tornule for Predicting
“ml-him! 'Mmganmmm am:
(“17,194“). 11-20

________. "A lornnle {or hedicting Inedehilitya lnetrnctione. "
““W mum mu: (him-m "#8).

lleech. Molph. "A I. lendehility terdetich,"i§ggl 95
mm. mm (June. 1948). 1-133.

Gilhineon. Board. Mean. DJ... end tithinh. DJ.
"Ifiect of Order end Authority in en Arguentetive speech,"

my 12m}. 2! m. n (mu. 19“). 183-191.

Grey, Willie: I. end teary.- Ietnice l. 14an g as
w. .chicego: he wherein of “ego here, 1935.

Galley. Hebert 1.. end Ierlc, Devid r. "lftecte oi lntereellnler
end Intrncelluler Speech Structure on Attitude Chen” end
Kenning." m 140003922. mu (love-her. 1956). 288497.

lumen. Irwi- H- m mm min m3 "mm.- of
byeholou." Vol. I. no. 9; rev rare: 6; Science rem. 1907.

Rock."- Wflu 1- 10m- of A W 22251 2!. W.
by clende I. thennnn end Vertex Ueever. We, XXIX (1953).
69.939

56

57

Hovlend, Carl 1., 2531. 23.29.25 93. meantetion in Eurasian.
In: York: Yele wivereity Irene, 1957.

lime. Devin 1!. "the Definition end Hector-rent of Word Probability,"
Unpubliehed ’hJ dieeertetion. Dept. of Peychology. .Barverd

, end Solonon, hicherd 1.. "Viral Duretion threehold en e
function of Hord~rrobebility." gonrne; 95 Eminent“
w. rm (June, 1951), 401-410.

Killer, George A. M :93 gmicetion. lee York: accru-
Iill My, Inc., 1951.

, "flhet ie lufornetion Heeenrenuu?" £1.68! "190103“ ,
VIII (Jennery, 1953). 3-11.

, end Iriednen. llinebeth A. "The reconstruction of

Mutileted Inslieh Tente," "interaction £13 gontm . I
(Bert-her. 1957), 38-55.

lille, Glen I. W the. m. lee York: Prentice-3e11,
Inc.. 1951. '

Mn. Idvin I. "Conpntetionel Kethode Ueeful in Anelyein; e

eeriee of hinery ante," Anergen gmml 35 W,
1..le (April, 1951), 251-162.

Ongood. Cherlee I.. end Inheoh, thoeee A. (ede.). [excholigguigticeg
A Emery of theory end neeeerch Irohlene. Mleeent to

gournel 2; marine; 93 eerie; m. an (October,
1954 .

roe-rm. Ila-nor u. £22222 2.: mama .12 52-. ____1.mu

usage. In: York: Doreen of Publicetione. 'reechere
College. Cole-bin Univereity. 1954.

Selfridse, Jeniffer. "lnveetiaetione into the Structure of Verbel
Content." mpubliehed nonore theeie, Dept. of reychology,
Radcliffe College. 1949.

Shennom Cleude I.,end fleever. Ilerru. 113 Hnthggtigel
2; W Urbenn: the Univereity of Illinoie
Irene. 1949.

Iponbera, Berold. "The leletive litectiveneee of cum end
Anti-Cline: Order in en Arsmeetetive Speech." m
W. X!!! (lo. 1, 1946) 35-44.

_sa

‘l'eylor. Hilton 1.. "Applicetiae of 'Cloee' end lntropy Heeeuree
to the Study of Contertnel conetreiet in Benplee of
Continuoue Drone." unpubliehed Ih.b. dieeertetion.
the Univereity of Illinoie, 1954. Sunnerined in gieeegtetgon
We. XV (lurch 1955) 464.465.

”'Clone' lendebility ecoree ee Indicee of lndividnel
Difference in Coqreheneion end Aptitude," gm 3;
mm. m (lo. 1. 1957). "-26-

-———.- "30°.“ Mloeneete in the lee of Clone Procedure,
“Io-mum. mm (Um-r. 1956). 42-43.

Arman: A (Experimentel neeuse in outline fore)

mum

eonething ehonld be done to inforu more people ebout the rceecrch
thet hee be. done on reedehility.

I.

Ileny of thoee who could echo the bent one of Ireedebility re-
eeorch ere unewere of ite erietence.

A.

B.

Pinch of the reeeerch on reedehility it effectively
hidden tron teechere in the traditional dieciplinee.

l. teeth of the reeeerch on readability ha been
pnhliehed in peychology journele. ‘

3. any of the reyorte on nedebility reeeereh hove
been publiehed in e very ureednhle fore end
cluttered with netheneticel eelcnletione.

heedere with linited ebility ere not aconreged to leern
ehout reedebility.

1. hey cennot reed the technicel reporte which ere
smerelly filled with eyntectic Synneetice end
fifty cent worde.

3. Reedehility hee generally be- preeuted on e eecret
weepon deeigned eepecielly for writere.

.m howledee now eveilehle ehoet reedehility een he need to
fecilitete co—nnicetion.

A.’

Iritere who only the principlee of reedehle writim
find their writing to he more effective. '

1. A tending eedience underetende better, leerne teeter.
end reteine nore from writing that ie rether eeey
to uderetend.

2. leople senerelly enjoy reading nore if it ie not
too difficult.

leedere who. uderetend the fectore thet influence rude-
bility ere chin to reed nore efficiently.

1. Given echoiee of neteriele. e render who known
the principlee of rudehility cen. quickly end
eeeily. eeleet boohe within the rage of hie
reeding ehility.

3. If required to reed e difficult book or erticle,
rendere find thet knowing the eource of the difficulty
in e etep towerd overcoming it.

59

NAM! AGE SEX

 

Year in School Grade Point Average

A_.

 

Your presence here indicates that you are willing to act as a subject
for an experiment. This experiment istart of a thesis that is being done
in the College of Communication Arts. Your scores will not count on your
grade, but the usefulness of this research is dependent on your doing your
best.

In the following messages you are asked to fill in the blanks with the
word that seems most appropriate to you. These messages have been carefully
checked for mistakes, and if the right word is filled in, every sentence will
be "good English" complete with punctuation. Each blank requires onc'word
and only one word.

Start work on the next page and work on "Impromptu Speaking" until thus
is called. You will have ten minutes on message one. Do not go beyond the
stop sign. work as rapidly as possible, but pay close attention to every clue
that might suggest the "right word" or eliminate a "wrong word." You should
try to get through the whole massage in the time allotted to take advantage
of any easy blanks. If no word seems exactly right, GUESS.

Ton minutes will also be allowed on the second message.

IMPROMPTU SPEAKING
Impromptu speaking often confused with extemporaneous

, not only by students A; A also by dictionaries. It

 

inadequate to state that speaking is done without

 

It is more accurate say that it is without

specific preparation. An a. speaker relies exclusively upon

V

 

general preparation. A speech if is given without either

 

or specific preparation, if be possible, is not

 

listening to. Some learning in speaking without specific

 

is extremely valuable. It us for the otherwise occasions

when we are to "make a few‘g " without recourse to a

speech. There are undoubtedly situations which lend

 

themselves well to impromptu speaking even an in-

experienced speaker do quite well. Perhaps ggh remarks of

another speaker serve as a challenge must be answered.

For less favorable situations, however, is well to be

for the short notice . Look and listen carefully

 

 

 

that you can adapt what is going on.

facility in the use some simple patterns of , and

. 61

.2-

 

 

 

 

 

 

 

 

 

 

 

 

practice the control A stage fright. Despite all , the
practice of impromptu has limited usefulness. Any
who succumbs to the ' rationalization that he can his
best without specific __ is deluding himself. It A . ‘well to
learn to .4. on one's feet, but ‘ is more important to
to think before getting to speak.
3 T 0 P

 

(1) mum

Something should A A done to inform more A_ about the
research that been ’done on readability. V of those who

could the best use of research are unaware of

 

existence. Each of the on readability is effectively

 

from teachers in the disciplines. Mach of the on

readability has been in psychology journals. Many

 

the reports on readability have been published in very
unreadable form and with mathenatical calculations. Readers

limited ability are not “F to learn about readability.

 

cannot read the technical which are generally filled

 

syntactic gymastics and fifty words. Readability has

 

generally presented as a secret designed especially

for writers. knowledge now available about can be used
to communication. Writers who apply :k principles of
readable writing their writing to be effective. A
reading audience better, learns faster, and more from
writing that rather easy to understand. generally enjoy

reading more it is not too . Readers who understand

 

(1) -2-

 

 

 

 

 

 

 

 

 

 

the that influence readability are to read more
efficiently. A a choice of materials, reader who knows
the of readability can, quickly easily, select books
within range of his reading A g . If required to read
difficult book or article, + find that knowmg the
of the difficulty is step toward overcomihg it.

 

 

For the first message, "Impromptu Speaking,” mark an x on'theJ'

following scales to represent your feeling about this experimental test.

Interesting : : : : ,Uninteresting
Easy : : : :: Difficult
Useful : : : é IiWorthless

For the second message, "Readability,” make an x on the following scales

to represent your feeling about this experimental test.

Interesting : : : : Uninteresting
Easy : : : : Difficult
Useful : : : : Worthless

‘5

ROOM USE ONLY

R-QM USE ONLY
MM17 fifilf’f
....-«.' 30". l, V".

      
     

DE 297 '
meufi - '
SNAG ,. 7 ‘

W‘-

  
    
 

     

 

 

     

F Luisa a

W m o

.47 ‘

 

 

 

At

HICHIGAN STQTE UNIV. LIBRQRIES

I II) I! II II! l|||l III” |

31293010 36764