Total or Selected Content ~nalysis I 
Julius Laffal 
Connecticut Valley Hospital, F~dd!et~n, Conn. 
The core of any system of content a~alysis is a dictionary speci- 
fying the content categories to be assigned to verbal units. The 
system described in this paper utilizes a category dictionary in which 
all the conmon words of English are classified into one or two of lid 
possible catec ories. 2 The associated cemputer method is at present 
capable of analyzing a text of 25,000 words in about 20 minutes of 
machine time, outputting a distribution of frequency of occurrence of 
the 116 content categories. Up to 15 separate texts may be analyzed 
at the same time, or the productions of \]5 speakers, mixed in any way, 
as long as each segment is identified with the appropriate source. 
The cc~puter analysis program will be discussed briefly, as will 
some of the considerations which went into the construction of the dic- 
tionary. An il\]nstration of the analysis of textual mater~a\] will be 
presented, and two studies applying the procedure will be described. 
Preparing, a ~ and machine analysis 
For any Enc!ish language corpus to be ane~!yzed, a simple editin C 
proce@Are is fo\]\]c~zed before the text is card punched. The editor 
I. Presented at the International Conference on Computational Linguistics, 
S~-S~y-,~ursgard, ~weden, Sept. I - 4, 1969. 
2. T~o of the categories (no score and hand-~core) are not actu~!\]y 
content categories. 
eliminates a number of functions words (articles, most conjunctions, 
auxiliary verbs such as J~ be and ~ ~, changes all verb forms into 
the present infinitive form, and reduces plural noun forms to the singu- 
lar. These editing changes are required because the dictionary contains 
only the root forms of the word. Illustrations 3 and 5 show a text and 
its edited form. Ultimately, as the dictionary expands to include 
common forms, many of these editing requirements will be reduced. There 
are 20,000 root forms in the dictionary at present. The editing pro- 
cedu~e reduces an original text by between 15% end 20%: a 25,000 
edited text would originally have consisted of around 40,000 words. 
The computer analysis is directed solely to content, and does not take 
grammatical form into account. The analysis thus shows what is being 
talked about, but provides no information that the grammar ordinarily 
conve~. It may show that aggression is a major theme of a text, but 
will not show who is aggressing against whom. The researcher may choose 
to analyze for total content, as represented by the 116 categories, or 
specify a limited number of content categories, in which case the pro- 
gram will disregard all other categories. 
The program is at present written in Fortran IV and MAP languages 
(ultimately to be entirely in Fortran IV). It consists of a number of 
subroutines capable of performing the following operations on a text: 
I. Print the text itself, numbering each line. 
2. Output the text words in order of occurrence, showing 
i 
speaker and line number. 
3. Output the text words in alphabetic order, showing speaker 
and line number. 
4. Output a list of words matched with the stored dictionary, 
shewing the categories which have been applied to each 
matched word, the speaker number, and the line number. 
5. Output a list of words which have more than one possible 
scoring in the dictionary (ambiguous words), showing 
either a selected number of the possible scoring, or all 
the scorings. Speaker number and line number are also 
indicated. 
6. Output a list of words which the dictionary cannot define 
because they are not represented by dictionary entries. 
7. Output profiles of the frequency of occurrence of each 
of the 116 categories for each speaker. 
8. Output a punched deck of category frequency profiles so 
that further statistical analyses may be applied. 
Th__~e Dictionar~ 
To find conceptual uniformities in the vocabulary of a language 
poses formidable problems. In typical content analyses, the researcher 
takes sentences or larger units, and evaluates them as to the presence 
or absence of a particular concept, dismissing as irrelevant to his 
purpose any material which does not bear on the concept. This approach 
is inadequate if one wishes to deal with all of the content of the sample. 
There are certain ampirical bases for developing a conceptional 
system which will deal with all possible contents. The steps in such 
a process may be described briefly. Numerous words are substantially 
similar to each other, and essentially synonymous in particular contexts. 
Dictionaries take advantage of this fact to introduce us to the meaning 
of unfamiliar words, by providing a synonymous word which we ~ow. 
Synonyms are of course not alwgys substitutable for each other without 
some change of meaning, for words may partake of more than one domain 
of meaning. 
The number of separate domains of meaning each occupied by a few 
synonyms is extremely large. As a step toward reduction one may group 
words similar in meaning, even if not synonymous. Words like walk, run, 
rid__~e, and ~o readily group together as sharing a c~on meaning. Appli- 
cation of the criterion of similarity along with that of synonymy leads 
to broader and more inclusive clusters of words. 
A further reduction in the number of clusters or domains of meaning 
may be effected by applying an additional criterion for grouping words: 
that of relatedness. If a set of words such as chair, stool, bench, 
seat, and couch is grouped by synonymy and similarity, considerations 
of relatedness may also bring tabl__~e into the same domain of reference 
s~ a common item of household furnishing. The merging of tabl..__~e and chai__~r 
into one group also projects the outlines of a category which could in- 
ciude other ~ like be__dd, dresse______~r, burea______uu, ches____~t, all of which also 
refer to ordinary household furnishings. These principles of synonymy, 
similarity, and relatedness, whe~ applied systematically to words, pro- 
duce a system of categories limited in number but still capable of 
describing meaningfully all the contents of t~e language. 
When a certain degree of stability has been reached in construct- 
ing categories, new words may be considered in terms of how they will 
fit into the system. ~his does not preclude changing the system; it 
merely weighs all the prior Judgments against each new Judgment to be 
made, and accepts change Only on compelling grounds. As the system 
grows, changes are more likely to occur because two existing categories 
are seen to border too closely on each other and to require redefinition 
of their boundaries and differences, rather than because new categories 
are discovered. 
A primary consideration in construction of the present conceptual 
dictionary was that the scheme of language had to provide an adequate 
basis, at the very least, for discriminating the language of one speaker 
from another. ~his meant that each entity in the scheme had to con- 
tribute to such potential. In practice this militated against units 
which were heavily loaded with tokens from everyone's speech and against 
units which rarely contributed a token to anyone's speech. Such global 
categories as "nature," "man," "animate" and "inanimate" provide very 
little basis for discriminating between the word predilections of sepa- 
rate speakers, since they inevitably contribute large numbers of tokens 
to everyone ' s speech. 
Relatively simple criteria for the acceptability of a category in 
the overall scheme were (I) that it not draw ~oo many or too few tokens, 
and (2) that it contribute to discriminating between separate speakers. 
~ere, in the present scheme, a category turned out to be large by these 
criteria, what was also readily apparent was some basis for splitting 
the category along one or more lines to produce separate categories. 
And where a category seemed too narrow, it was possible to merge it 
with another neighboring category, the domain of reference than swell- 
ing to include both meanings. The area of hostility and aggression was 
one which required subdivision along lines which differentiated violence, 
anger, and simple disagreesnent. The area of sex, in which at the out- 
set a discrimination was provided between heterosexual and homosexual, 
ended by becoming a single area, since very few references occur to 
homosexuality as such as opposed to sexuality in general. 
One other criterion which was applied is related to well ~own 
ideas of reliability and validity. The domains of reference in a scheme 
of language must be sufficiently discriminable from each other so that 
separate judges considering a word will be in agreement as to which 
categories are involved. If the dcemins of meaning are too closely re- 
lated, this will be a difficult Judgment and an unreliable one. To the 
extent that it is unreliable, its validity and its pertinence to other 
behavior, will be dubious. Distinctiveness of a domain of reference in 
a dictionary scheme is therefore an important desideratum. 
Over the years, as texts from various sources were analyzed, with 
a continuing effort to fit new words into existing groups or to redefine 
the word groups to make them more congruent with word meanings, a dic- 
tionary of words with their appropriate categories emerged. The dict- 
tionary and its mode of application are illustrated on the following 
pages which shc~T: 
!. Definitions of a few categories. I 
2. A page of dictionary entries showing associated categories, 
and illustrative sentences. 
i. A detailed description of the categories is contained in Laffal, J. 
Pathological and normal language. New York: Atherton Press, 1965. 
3o 
4. 
5. 
6. 
A selection from Jonathan Swift's Gulliver's Travels. 
Handscoring of the Gulliver's Travels selection. 
Selection from Gulliver's Travels, edited and punched on 
cards for machine analysis. 
Profiles derived by computer and by hand analysis of the 
Gulliver's Travels selection. These profiles also show 
an alphabetic listing of the categories by heading word. 
In the computer analysis the frequency profile is updated 
by the different categories in the first two meanings for 
each ambiguous word, whereas in the hand analysis, only 
the categories in the relevant meaning are applied. 
I. Definitions of a Few Categories 
CHANGE: The kinds of words which are scored here are: 
a. ~.~ords which refer to deviation: ben___~t, croo~_____dd, ~, tortuous, 
~, derail, avoid, evad_.__~e, distract. "He turned (CHANGE, GO) 
the corner." 
b. Words referring to alteration either of form or position: canvert, 
mutation, revise, transform, shift. "The dress was altered (CRANGE)." 
c. Words referring to instability or to inconstancy. "His opinion 
fluctuates (C~L~NGE) from mament to m~ent." 
d. Words referring to transience, temporariness, or momentariness. "I 
caught a ~ (CHANGE, FAST) glimpse of him as he ran." "~he 
co~Antry was ruled by a provisional (CHANGE) government." 
CLEAN: The types of words scored here are: 
a. Words referring to personal cleanliness and gro~ng: toothbrush, 
comb, wash, towel, so_~. "Scrub (CLEAN) behind your ears." 
b. Words which refer to elimination of dirt: purification, cleanliness, 
dustin_____~, la~dry. "She finished ~ (CLEAN) the floor." 
CLOTHING: All references to wearing apparel and thin~s pertinent to 
being clothed or dressed are scored here. In addition, relsted aetiv~- 
ties or references, such as se__~w and darn, are scored here. Some clnthing 
is clearly male, female, or infant. In such cases the additional scoring 
of KALE, FEFJtLE, or YOUNG is applied. "His trousers (CLO~NO, MALE) 
I had to be altered." "She wor___~e (CLOTHING) a beautiful dress (CLOTHING, 
FEV~LE) ." "She changed the baby's diaper (CLOTHING, YOUNG)." 
COLD: All words in which the idea of coolness or coldness is of impor- 
tance are scored here. "It's freezin K (COLD) in here." 
2. A Page of Dictionary Entries Showing Associated Categories 
and Illustrative Sentences* 
COED N FEMALE 
COEDUCATION N JOIN 
COEFFICIENT N NUMBER 
COEQUAL J AGREE 3 
COERCE V COErFL. I 
COEVAL J AGREE 3 
REASON 2 An attractive coe d won the contest. 
REASON 2 Coeducation is now co~suon in college. 
A coefficient is a multiplier. 
The soldiere were ~ in rank. 
DC~. i Bandits coerce the travelers. 
TIME 4 Roosevelb and Churchill were coeva_...~l 
leaders. 
LIVING Nations must coexist pescefu!ly. 
VEGETA. He drank coffe_~e in the morning. 
Coffee is a brown color. 
She kept her Jewels in a coffer. 
HOLLCW 2 A flag was draped on the coffin. 
The c_~ of one gear fits the other. 
H~s unfulfilled promise was a co__~. 
JOIN The carpenter will ~ the boards. 
TRUE The argument has great ~. 
T~JE The debater presented a co~ent argument. 
It is wise to ~ before acting 
VEGETA. Co~ is a French brandy. 
JOIN Father and rater are co~ate words. 
Learning is the basis of cognition. 
The court took ~ of the dispute. 
COEXIST . V JOIN 
COFFEE N DRINK 
CO.~FEE N COLOR 
COFFER N HOLL~ 2 
COFFIN N DEAD 
COG N RILL 
COG N FALSE 
COG V HILL 
COGENCY N ~EASON i 
COO~{T J REAS~I ! 
COGITATE V ~EASON i 
COGNAC N D~NK 
COCANATE J AGREE 3 
COGNITION N REASON I 
COGNI~-~NCE N P~EASON I 
*J, adjective; N, noun; V, verb. 
iO 
3. A Selection from Jonathan Swift's Gulliver's Travels 
My father had a small estate in Nottinghamshire; I was the third 
of five sons. He sent me to Emmanuel College in Cambridge at fourteen 
years old, where I resided three years, and applied myself close to my 
studies; but the charge of maintaining me (although l had a very scanty 
allowance) being too great for a narrow fortune, I was bound apprentice 
to Mr. James Bates, an eminent surgeon in London, with whom I continued 
four years, and my father now and then sending me small smm~ of money, 
I laid them out in learning navigation and other parts of the mathe- 
matics useful to those who intend to travel, as I always believed it 
would be some time or other my fortune to do. When I left Mr. Bates, 
I want down to my father, where, by the assistance of him and my uncle 
John and same other relations, I got forty pounds, and a prmmise of 
thirty pounds a year to maintain me at Leyden; there I studied physic 
two years and seven months, knowing it would be useful in long voyages. 
Soon after my return from Leyden, I was recommended by my good 
master D~. Bates to be surgeon to the Swallc~, Captain Abraham Pannell 
Commander, with whom I continued three years and a half, making a voy- 
age or two into the Levant and same other parts. When I csme back I 
resolved to settle in London, to which Mr. Bates, my master, encouraged 
me, and by him I was reco~aended to several patients. I took part of 
a muall house in the Old Jewry, and, being advised to alter my comdition, 
I married Mrs. Mary Burton, second daughter to 'Mr. Edmund Burton, hosier 
in Newgate Street, with whom I received four hundred paunds for a portion. 
But my good master Bates dying in two years after, and I having 
few friends, my business began to fail; for my eonsciance would not 
II 
suffer me to imitate the bad practice of too many among m~ brethren. 
Having, therefore, consulted with my wife and some of my acquaintances, 
I determined to go again to sea. I was surgeon successively in two 
ships, and made several voyages for six years to the East and West 
Indies, by which I got some addition to my fortune. My hours of lei- 
"sure I spent in reading the best authors, ancient and mode~n, being 
always provided with a good number of books, and when I was ashore in 
observing the manners and dispositions of the people, as well as learn- 
ing their language, wherein I had a great facility by the strength of 
my mmaory. 
The last of these voyages not proving very fortunate, I grew weary 
of the Sea, and intended to stay at home with my wife and family. I 
removed from the Old Jewry to Fetter Lane, and from thence to Wapping, 
hoping to get business among the sailors, but it would not turn to 
account. After three years' expectation that things would mend, I 
accepted an advantageous offer from Captain William Prichard, master 
of the Antel~e, who was making a voyage to the South Sea. We set 
sail from Bristol, May 4, 1699, and our voyage at first was very 
prosperous. 
12 
4. Hand Scoring of the Gulliver's Travels Selection 
DOM.2 EARTH 
SR* MALE POSSESS LITTLE POSSESS IN PLACE 
My father had a small estate in Nottinghamshire; 
JOIN DOM.2 HOLLOW I 
SR NUMBER NUMBER MALE MALE GO SR REASON 2 
I was the third of five sons. He sent me to ~anuel College 
HOLL(14 1 
IN PLACE NUMBER TIME 4 TIME i PLACE SR LIVING NUMBER 
in Cambridge at fourteen years old, where I resided three 
• TIME h WANT i SR NEAR SR REASON i CONF.3 MONEY 
years, and applied myself close to my studies; but the charge of 
DUPABLE HELP 
HEIP SR CONF.3 SR POSSESS SHARP 2 LITTLE MCNEY 
maintaining me (although I had a very scanty allowance) be- 
SHARP 2 BIG AGREE 3 
ing too great for 
CURE 
ALL 2 REASON 2 
LITTLE MONEY SR WANT I SUB. i 
a narrow fortune, I was bound apprentice 
CURE 
YALE SHA~RP I DOM.I SHARP I IN PLACE JOIN SR 
to Mr. James Bates, an eminent surgeon in London, with whom I 
DOM.2 
DUiABLE ND~BER TIME 4 SR MALE TIME 2 GO SR 
continued four years; and my father now and them sending me 
*Self Reference 
13 
OUT 
LITTLE ALL i MONEY SR MONEY SEPARATE REASON 2 GO 
small s~was of money, I laid them out in learning navigation, 
CONF.3 SOME NUMBER HELP WANT I 
and other parts of the math~uatics, useful to those who intend to 
AGREE 2 
GO AGREE 3 SR DURABLE REASON I TIME 4 
travel, as I always believed it would be s~ae time or 
END GO CURE 
CONF.3 SR HELP TIME 4 SR SEPARATE MALE SHARP I SR 
other my fortune to do. When I left Mr. Bates, I 
DOM.2 DOM.2 
GO DOWN SR MALE PLACE HELP MALE SR 
went down to my father; where, by the assistance of him and my 
JOIN 
MALE SOME~ CONF. 3 JOIN SR POSSESS NUMBm~ MONEY 
uncle John, and sOme other relations, I got forty pounds, 
DURABLE 
AGREE 2 NUMBER MONEY TIME 4 HELP SR NE~ PLACE 
and a prOmise of thirty pounds a year to maintain me at L~den: 
SR REASON i REASON 3 NUMBF~R TIME 4 NUMBKR TIME 4 
there I studied physic two years and seven months, 
BIG 
REASON i REASON 3 HELP DURABLE GO 
~icwing it would be useful in long voyages. 
NEAR BACK AGREE 2 
TIME 3 BACK SR GO SEPARATE PLACE SR GOOD 
Soon after my return frm, Leyden, I was reco~uended, 
CURE CURE ANIMAL SEA 
SR GOOD DOM.2 MALE SHARP i SHARP I UP ~gANS. 
by my good master Mr. Bates, to be surgeon to the Swallow (ship), 
DCH.2 D~.2 JOIN SR DURABLE NUMBER 
Captain Abraham Fannell cmmmander; with whom I continued three 
NUMBER 
TIME h SaME GO GO NUMBER IN PLACE 
years and a half, making a voyage or two into the Levant, and 
DURABLE 
SQME CONF.3 SC~E TIME h SR GO BACK SR WANT I LIVING 
some other parts, When I came back, I resolved to settle 
CURE 
IN PLACE MALE SHARP I SR DGM.2 HELP SR 
in London, to which Mr. Bates, my master, encouraged me, and by 
CUPm ACL~EE 2 
SHARP I SR GOOD SaME ILLNESS SR POSSESS S~ME 
him I was recommended to several patients. I took part of 
GROUP HELP 
LITTLE HOLLOW I IN TIME i LAW PLACE REASON I 
I a small house in the Old Jury (Place); and being advised 
JOIN F~ALE 
CHANGE SR HAPPENING SR SEX F~£%LE JOIN NUMBER 
to alter my condition, I married Mrs. Mary ~, secomd 
n 
to ~. ~ S~r~o,.hosler 1,1 aowp~~.(r~e), 
~ttb ~ I received four h~dred ponds for a.~m"ticm' ~ .*~:* 
CONF.3 S~ GO~ IDa,2 ~ 1 ~ ~ I~\[~Z-:I~.~ 
But, wS good mast, e~ Rstes ~ in t~o- ~:,.\]mars ~. 
after, and I bavln~ few friends) ~- business ~ began to" fa~l; 
- GO(U). ._.' .At~l~i~2 ~:3 : : - 
• :~' REASON l NEGAI"ZON UPSET SR F~ BAD -. 
for .~-emsselencewould not • :. anxfTer .. me to Imlt~te, tJse bad -- 
JOIN 
DURABLE- SHARP 2 ALL 2 J(\](IN SR ~ , ;~ 7 " 
practice of too many amonZ ~ brethren. P.~rlng therefore 
HELP . )L~ DQH\[.! 
H~ASON 1 JOIN SR-. S~X SOH~ SR .N\]~R : , :,,, ,. ;SR HEASON 1 
eonmL1ted lrltb ~ srJ~e, and same of ~ aequalntanee, I dotendned 
._-~ ~ .... ~ .: .: .:, '. "~ 
• GO &~3 SEa ~Sa. ,Sm~Url ~ ,, .I~ nn~..~ms. 
to go a~atn to sea. I was surseo~ succesetvel~ in teo ships, 
-~,AI~ "~. ~;- LANGUAGE2.:-.. DOM~I/\] .~- .~.~ ~:~, 
hours of leisure I spent t,~7~-eading the best authors, 
TI~.~.:. ~TIM~ 2 :.'~-. DI~RKBLE:. HELP,~ ~.~==, .-~GOOD ~ NUI~ER 
ancient!and mbd~r~ bet~ alw~y~= pz~Ir~de~wi~h 'a: good. n-umber of 
• /: ~- ". ~' • . EART~ ,. : ~ :_ =. ;" ?i DURABLE ~- 
LANL~AGE 2:: : TIME h~" SR "~:~ SEA::--: '~'- :SEE ,.'::~, - HA.P~'N(3 
book~; and.when I was~mhore, in observing the manners and 
WANTLI: 3 =..~-" GROUP KGREE 3 --.~ASCN 2 GROUP -LANGUAGE I 
d~spos~ti~n~ o~'J ~he people, a~well as ~. :' their-- language,. : 
...... ' ," / / 
SR POSSESS " BIG .LF~SY ~C ~-~ ". ~ D~,I '. "- >~ SR REA~: I~ 
vher%~ :I" ~:nhaT ~i ;"" a ~Teat~ ~i\]~ ' ~5 ~':the ~st~nEd~h Of:~ ' memory.: 
-.::- The last of the%e VbFages,rmt prbving very :,_ = ,5..: 
":-- -:---~ - . -.'-SffB.I' ~-:: ~"." " :: • " .'- -, '" 
HELP SR CHANGE UPSET ;SEA WANT 1 ~ NEAR 
, -" ~o~tuDa-te, ~ grew ~;tm~r3r 0% the)sea, an~ intended to stay at 
HO~L~; 1 JOIN;SR SEX. ~. 5".~ JOI#7,,. SR SEPARATE SEPARATE-:- 
~P 
TIME i LAW . PLACE CONFINE GO FLACE SEPARATE 
Old Jury (Place) %0 Fetter - Lane (Place), and from thence 
PLACE WANT 1 POSSESS COI~CE JOIN SEA CONF.3 
to Wapping, hoping to get business among the sailors; but it 
"~'"".. . ."-:L, 7~ i~.7%7.£-{ ~0 J.L,--~S' .i\]..:? YL~L ~~. i ::_ .7~ ~50LU ;L~lE~ 
CHANGE 
?";~\] ,S~ ...... .~, ,~L-"-~ '.. i .71,£F£ :\].-- ,'--',, ..... --,v~ ...iTk..L \]Y ~"~-~ ..~, v~::,...=~. UR.Ci 
NEGATION GO HELP BACK NUMBER TIME 4 WANT i 
~" 77~"'£-1 -',~L;F.-~ ,t-..?.'.i" F..'5"Y~ >, ~-: '~=~"~>"'-.~ ,~- : . ,~'~ OT ~'~" ~-~;"~,, ~ ~ :--A &~)3£ ~'j" - 
would not turn to account. After three years expectation 
AGFRE 2 
...... ~II?.0\[%AI iiL:,l\[ .... '""'~ .... ,,..,'D~, ~{~! ilia. .... idlE' ':-'.'\]'iY 
HELP SR POSSESS HELP HELP 
that thinls would mend, I accepted an advantageous offer from 
(V},I I :T u(T2!\]! }U" .-c-.E-LI~ R0 ,,,~,t/7 ~,~,t i~! C.TJC,; 17 ~''~-~ 
SEA 
DON.2 DOM.2 A~CI MAL TP.~NS . GO 
Captain William Prloherd, master of the Antelope (ship), who was m~lin I 
GO 
~0<,'I :'..(Y: ..W ~V.~£ ,\]i : ;£-.-1 ;'~\] ~:'.7 -•-r Y".',-_ "-.~Z ~. 1~-\] +~I ..?~ \[2Ni\]~ 
GO ICE SEA SR BEGIN SEA SEP~%TE FLACE TIME h 
a voyale to the South - Sea. We set sail from Bristol May 
NUF~ER FO?~WARD DOM.I ~F ~-.~ ~ .......... ;':-"; "= ..... '"~ ~,~i~..~ ~'('.~.~ L~£. ~ 
NUMBER TIME 4 SR GO NUMB~ SHAd2 2 HELP 
",-.-'LL :£~'- ~!'£'". \[-..'I'Tq'U ,~)7 ~."~ u-'-: : . . . S_"~-!:'\]-::..'£kRL~,. WL~.'_ ~ ,.~ 
4, 1699, and-our voyale at first was very prosperous. 
~_~ ~ >~ &~;g-/~'?,~ ,~."~F~?~ ~'.~ .T+~ ' ~. ~.~;{.~ .~'~'~ 
5. Seleetlen ~ O.111w~r~s 
~ted and l~mcbed -- Oards for Macbiee ~is 
CLOSE TO ~ STIII~, IRJT THE CHAR(Z ~' MAINTE\]~ B AL~H I(P) 
~,~x v~r sc~rr Au~ow~os sE~o~v) Too oanT F0e NaTm0W FOa1~e~, I(P) 
~LS B00ND ~CE TO ~R. JAMES-E%TES S~EON, AN ~ S~ROE~ IN 
LOm~ PLACE, WITH ~:t(P) oun'Im~ Fum ~a~6 Am) ~ PA~m No~ AND 
S~D ME ~ SUN OF ~0RE~, I(P) ~ 0OT LEARN NA~IC, ATE, AND 
OTHER PART OF THE MATHemATICS. U3~I~UL TO THOSE WHO ~ TO ~JkVEL, AS 
I(P) LEAVE ~. RATES S~R0~N. I(P) GO DOWN TO ~ FATHE"~ 
B~ THE ASSIST ~ FATHE~ AND I~ UNCLE JO~N, AND ~ OTH~ }~LA~,, I(P) 
GET FORTf POUND. AND PROMISE OF ~ POOND ~ TO MAINTAIN ME AT 
L~q~EN PLACE. THEEE I(P) STODT PHISICS TWO \]~AR AND SEV~ 14\[NTH, KNG~ 
PHYS!C3 WOULD BE ~ LONG VOYAGE. 
SO~ AFTER l~ ~ FRC~ LE~DE~ PLACE, I(P) WAS ~, ~f l~ 
GO~ ~ M~. BATES S~RG~, TO BE S~R~ TO T~E S~LL~ SHIP, 
*~et+~ers in parenthe~s in.care part of speech. Where word may have 
more than one part of speech, spectficatian of the part selects the 
appropriate dictionary - entry. 
19 
I(P) COME BACK, I(P) RESOLVE TO SETTLE IN LONDON PLACE, TO WHICH 
MR. BATES SURCEON, MY MASTER, ~TCOUI=LAGE ME, AND BY S~RGEON I(P) WAS 
RECONMEND TO SEVERAL PATII~?. I(P) TAEE PART OF SNAIL HOUSE IN THE OLD 
an~ PLACE, AND BEIHO(V) ADVIS~ TO ALTE MY CONDITION, I(P) MAm~ mS. 
MARX-BURTON DAUGHTER 2 SECOND DAUGHTER TO MR. EDMUND-BURTON HOSIER, 
HOSIER IN N-E~GATE STREET Ira.ACE, WITH WH~ I(P) RECEIVE FOUR POUND FCR 
PORTION. 
BUT, MY GOCD MASTER BATES SURGEON DIE(V) TWO YEAR AFTER, AND I(P) 
HAVE F~ ~, MY BUSINESS BEGIN TO FAIL, FOR(C) MY CONSCIENCE WOULD 
NOT SUFFER ME TO IMITATE THE BAD PRACTICE OF TOO MANY AMONG MY BRETHRen. 
THEREFORE CONSULT WITH MY WIFE, AND SOME OF MY ACQUAINTANCE, I(P) DETer- 
MIME TO GO AGAIN TO SEA. I(P) WAS SURGEON SUCCESSIVE IN TWO SHIP, AI~ 
MAKE SEVERAL VOYAGE, FOR SIX YEAR, TO THE EAST AND WEST INDIES PLACE, 
BY WHICH I(P) GET SCME ADDITION TO MY FORTJNE. MY HOUR OF LEISURE I(P) 
SP~D READ THE BEST AUTHOR, ANCIenT AND MODERN, BEING(V) ALWAYS PRGTI~ 
WITH GOOD NUMBE~ OF BOOK, AND ~ I(P) WAS ASHORE, CSSERVE THE MANNER 
AND DISPOSE OF THE PEOPLE, ASWELLAS LFJuW_N PEOPLE LANGUAGE, WHEREIN I(P) 
HAVE ~°~EAT FACILITY BY THE STR~GTH OF MY M~OI%Y. 
THE LAST OF THESE ~OYAGE NOT PROVE VERY FORTONATE, I(P) ~OW WE~Y 
OF THE SEA, AND I~ TO STAY AT H(~E WITH MY WIFE AND FAMILY. I(P) 
REMOVE FROM THE OLD JURY PLACE TO FETTER LANE PLACE, AND FRaM TH~CE 
TO WAPPING PLACEj HOPE TO GET BUSINESS AMONG THE SAIL(R, BUT IT W(N/LD 
NOT TURN TO A~T. AFTER THREE YEAR EXPECT THAT THING WOULD M\]~4D, 
I(P) AC(~PT AN ADVANTAGE OFFER FRCM CAP~ '~/LLIAM-PRIC~, MAST~ OF 
THE ANTS~OPE SHIP, WHO WAS MAEE 9U~AGE TO THE SOUTH SEA. WE SET SAIL 
20 
FROM BRISTOL PLACE MAY FOUR, SIX, AND OUR VOYAGE AT FIRST WAS VERY 
FROSPK~. 
"7 
21 
6. Profiles Derived by Computer and by Hand Analysis 
of the Gulliver's Travels Selection 
(SI is the machine output profile, $2 the hand scoring profile). 
SI $2 SI $2 Sl S2 
ABSURD O O CO~RERCE 5 4 FAST 0 0 
AGREE i i I CONFINE 6 i FEA ~JRE 0 0 
AGREE 2 14 6 CONFLICT i 0 0 F~tALE 5 5 
AGREE 3 15 6 CONFLICT 2 I 0 FORWARD 2 2 
ALL i i i CONFLICT 3 IO 8 FUNCTION i 0 
ALL 2 2 3 CRIME 0 0 GO 25 22 
ALL 3 0 0 ~JRE 9 9 GOOD 9 7 
ANIMAL 2 2 DEAD I i GROUP 7 5 
ART O 0 DIFFICULT 0 O HAND SCORE 2 * 
ASTRONC~YI 0 0 DIRTY 0 0 HAPPENING 2 2 
ASTRONC6S2 0 0 DOMINANCEI 15 6 HEAR 0 0 
BACK 6 5 DOMINANCE2 12 12 HELP 19 17 
BAD I I DOWN i I HILL O O 
BEGIN 1 2 DRINK 0 0 HOLLOW- I 9 4 
BIG 5 4 DURABLE J3~ i~i HOLLON 2 O 0 
BLURRED 3 0 EARTH 2 2 HOT 0 O 
CALM I I EASY i I HOUSEHOLD 0 0 
CHANGE 3 3 EAT I 0 ILLNESS i I 
C!~AN 0 O END 5 2 IN 8 8 
CLOTHING 3 2 ESSENTIAL I O INDIVIDUAL 4 O 
CG~D O O FALSE i I JOIN 18 16 
COLOR O O FAR 0 O LANGUAGE I 2 i 
22 
LANGUAGE 2 3 3 POSSESS 13 II TIME I 5 4 
IAW 3 2 REASON I 14 9 TIME 2 2 2 
LITHE 5 5 REASON 2 4 4 TIME 3 I i 
LIVING 2 2 REASQN 3 2 2 TIME 4 21 17 
MALE 13 13 SACRED 5 O TRANSPC~TA 3 3 
MATERIAL 1 0 SEA 9 9 TRIVIAL I O 0 
MEASUREM~ 3 0 SEE 2 2 TRIVIAL 2 0 0 
MEC.~ANISM 0 0 SELF REFER 58 58 TRUE 1 1 
MONEY 6 9 SEPARATE 9 7 UNREAL 0 0 
NE~ 12 5 SEX 3 3 UP 1 1 
NEGATION 3 3 SHARP I I0 9 UPSET 4 2 
NO SCORE 162 . SHARP 2 i0 5 VEGETATION 0 0 
NUMBE~ 24 23 SLOW i i W~T I 13 8 
OP~ O 0 SCME 13 12 WANT 2 2 I 
OUT I I STRUCTURE 5 0 WORK i 0 
PLACE 19 19 SUBMISS i 3 3 YOUNG 0 O 
PLAY 2 0 SUBMISS 2 0 0 
PROFILE TOTALS 731 446 
*Hand Score and No Score items are not tabulated when a profile is 
scored by hand. 
23 
To illustrate the application of the computer analysis, two stud- 
ies are described, one, of the free flowing language of children, the 
other, of scientific writings. 
A book by Evelyn G. Pitcher and Ernst Pre!inger (1963) presents 
stories told by boys and girls from ages 2 to 5, taken down in short- 
hand by one of the authors. These were a group of economically privi- 
leged children attending a nursery and kindergarten of a private school. 
~/ost of the children were of superior or high average intelligence. 
When the occasion presented itself and the child was for the mament 
playing alone, the researcher asked, "Tell me a story." ~ly those 
stories were rejected which were a retelling of a familiar fairy tale 
or television show. 
The stories taken from the Pitcher-Prelinger book were transcribed 
onto 13~ cards end analyzed by computer to obtain profiles showing the 
distribution of the content of each age group. In order to sharps~ the 
subsequent analysis, all categories showing a frequency of less than 
1% of the total responses within each profile were elim~ated. Also 
eliminated was the one category (ANIMAL) which showed a uniform high 
frequency (over 7% of the total) for each profile. These extremely 
low and extremely high frequency categories are eliminated because 
they contribute only to raising all correlatiens between profiles and 
do not contribute to discriminating between profiles. ~he profiles, 
now consisting of 52 categories were then correlated with each other, 
and the matrix of correlations factor analyzed. I will not present 
the actual factor analysis, but only describe the outcome. 
24 
The first factor in the analysis is a "female" factor. The girls 
in all age groups load very highly on it. When we look in detail at 
the categories which contribute most heavily to this factor we find 
that YOUNG, F~MALE, and HELP are the three outstanding contributors. 
In the category system, F~ALE and HELP are the two scorings which 
apply to the word mother and its variants. These findings suggest 
that the girls of the study had as central themes in their fantasy, 
ideas relating to mothers (feminine helpers) and children. 
The second factor is more clearly relevant to boys, with ages 3 
and 4 showing predominance. The most significant categories here are 
GO, MALE, and TF-~NSPORTATION. Thus, the boys in the group, particularly 
those in the 3 and 4 year bracket, had prominently in their fantasies, 
references to maleness, physical movement, and vehicles of transporta- 
tion. 
The third factor applied primarily to girls age 4 and boys age 5. 
Primary categories were MALE, GO, and FEATURE (bodily features). The 
F~UP~ and ~LE categories would be consistent with a psychoanalytic 
view that at this age both boys and girls are in the oedipal period 
and are becoming preoccupied with sexuality, and particularly with the 
relationship to father. 
The fourth factor seems most clearly applicable to boys age 2, 
although to a lesser degree it is characteristic of boys age 3, suggest- 
I 
ing a decay process with age. Here the outst~ndin C categories are 
c~rFLICT I (physical harm), GO, and YOUNG. Like the other boys, the 
t~o year olds were also occupied ~th physical movement. Li~ the 
25 
F 
girls, they were occupied with ideas about children; but violent aggres- 
sic~ was their most prominent concern and distinguishes them most clear- 
ly from other groups 
A second study by Hartsough and Laffal explores the question 
whether scientists who have been described as visual imagists and 
scientists who have been described as verbal imagists (Roe, 1952) will 
also differ in the types of content ~mployed in their writing. The 
four sciences which Roe had studied were Social Science, Biology, 
Experimental Physics, and Theoretical Physics. Based on a series of 
projective tests and interviews with scientists in these fields, as 
well as their self-reports of problem solving activities and typical 
modes of thinking about problems, she concluded that Social Scientists 
and Theoretical P~sicists were verbal imagists, while Biologists and 
Experimental Physicists were visual imagists. Verbal imagery is think- 
ing ~In words or talking to oneself, while visual imagery is thinking 
in p~ ctures. 
In our study, we selected three outstandin o - living sc~@nt~sts 
from each of these four disciplines, the Soc~! Seiences being repre- 
sented by Anthropologists. For each Scientist one re~esentative book 
was chosen from his technical publications addressed to colleagues, snd 
one from his popular writings for the laity. The key words 
scientist, ~, .Anthropolosy, and ~ were used to identify 
contexts for comparison. A context included the sen~nce containing 
the key word, the preceding sentence, and the subsequent sentence. In 
26 
addition to the books of the scientists, an Anthropology, a Biology 
and a Physics textbook of this time were also selected. The content 
of the textbooks was sampled by taking three sentences alternately 
from the top, middle and bottom of every other page, without regard to 
any key words. The writings of the various scientists, and in the texts, 
were analyzed by the com~rater program described, and content category 
profiles were obtained. These profiles were then correlated with each 
other and factor analyzed. I will not present details of the factor 
analysis, but only the results. 
Highly represented on the first factor were all of the Anthropol- 
ogical writings, and the writings of the Theoretical Physicists. Since 
these two groups together constitute what Roe had called the verbal 
scientists, the first factor could be considered a verbal factor. The 
actual content which was stressed in the writings of these scientists, 
however, evokes more the idea of a "humanistic': factor, since the 
particular categories contributing most importantly to the factor are 
R~SG~ 1 (kn~;ledge and thinking), GROUP, HELP, and GOOD. 
Thus, the content analysis tells us that what manifests itself as 
a highly developed verbal co~anicative s~ll, also reflects a concern 
with h linen affairs and human issues. Since language is par excellence 
the social skill, it is natural that those who use it best would also 
be those whose attention is directed toward the ht~mn condition. 
The second factor was clearly related to Physics, the writinEs of 
the Theoretical and Applied Physicists loading highest on it. The 
categories which contributed most to this factor were : REASON 3 
% 
2 
27 
(science), NAT~IAL, REASON 1 (knowledge and thinking) and SOME. Other 
categories which were stressed were ESSENTIAL and TIME 4 (general time). 
The third factor was clearly a biological factor, since the biological 
writings loaded highest on it. The categories of greatest importance 
were: FUNCTION (bodily functions), BEGIN CHANGE, MATEP~IAL and FOR- 
WA~D. 
To be noted is that no "visual" factor as such appeared in the 
analysis. This suggests that the contrast which had been drawn by Roe 
between "verbal" and "visual" may actually be a contrast between those 
with a central humanistic concern inclined to be strongly verbal in 
their thinking, and those who do not have this same central interest, 
who are more likely to stress non-verbal processes in their thinking. 
The method of total content analysis which I have described is 
applicable to many language content problems. Beside the applications 
described in the present paper, the technique has been used in studies 
of the symbolism of key words in the language of a psychotic patient 
(Laffal, 1960); changes in the language of a patient in treatmen~ 
(Laffal, 1961); comparison of the language of therapists in tal~ng 
about different patieuts (Watson & Laffal, 1963); comparisons of the 
content given by subjects in single word association, continuous word 
association, and free speech in response to the same stimulus word 
(Laffal & Feldx~an, 1962, 1963); comparison of the language of two and 
three speakers in conversations about a variety of topics (Laffal, 1967). 
In broadest terms the method may be applied to comparison of one 
individual with himself (for example, over time); one individual with 
2B 
another; a group with itself or with other groups; the contexts sur- 
rounding one topic with those surrounding some other topic. The unique 
contribution which it makes is in the use of a category dictionary 
which classifies all the ccmaon words of ~glish into one or two con- 
tent categories, so permitting rapid display either of all the content 
or of selected contents, in a text. 
29 

References 

Laffal, J. The contextual associates of sun and God in Schreber's 
autobiography. J. abnorm, soc. P~chol., 1960, 61, 474-479. 

Laffal, J. Changes in the language of a schizophrenic patient during 
psychotherapy. J. abnorm, soc. ~., 1961, 63, 422-427. 

Laffal, J., & Feldman, S. The structure of single word and continuous 
word associations. J. verb. Learn. and verb. Behav., 1962, i, 
54 -61. 

Laffal, J., & Feldman, S. The structure of free speech. J. verb. 
Learn. and verb. Behav., 1963, 2, 498-503. 

Laffal, J. Characteristics of the three person conversation. J. verb. 
Learn. and verb. Behav., 1967, 6, 555-559. 

Roe, Anne. The making of a scientist. New York: Dodd, Mead, 1952. 

Watson, D. L., & Laffa!, J. Sources of verbalizations of psycho- 
therapists about patients. J. gen. P~hol., 1963, 68, 89-98. 
