JOURNAL OF VERBAL LEARNING AND VERBAL BEHAVIOR9, 444-449 (1970)
Word Associations to Homographs DONALD H . KAUSLER AND SYLVIA F . KOLLASCH l
Saint Louis University, St. Louis, Missouri 63103
Word associations to 40 homographic stimuli were scored in terms of the semantic features serving as S's apparent functional stimulus, and hierarchies related to the separate sets of features of homographs were determined. For most homographs the dominant meaning in the associativehierarchy was the more frequently occurring meaning in semantic counts, a finding in agreement with a spew-likeprinciple of perceivinghomographic stimuli. A homograph is a word which has two or more distinct meanings, each of which has its own etymology. Homographs are usually homophones as well (e.g., bat, date, mine), although there are instances in which a differential pronunciation accompanies the variation in meaning (e.g., lead, sow, wind). Homographs function semantically as if they have been encoded and stored as two (or more) distinct words, each having its own unique set of semantic features (Katz & Fodor, 1963). As such, homographs represent a useful population of words for testing hypotheses related to the priming of meaning and the subsequent effect of priming on such activities as word association, clustering during free recall, false recognition, and so on. Recent studies by Cramer (1968) and Kausler and Kamichoff (1970) attest to the heuristic value of homographs in this capacity. In addition, homographs offer a valuable research tool for testing hypotheses relating psychopathology to verbal behavior (e.g., Benjamin & Watt, 1969). Other than semantic word counts (e.g., Lorge & Thorndike, 1938), there has been little research directed toward the characteristics of homographs per se. Of special interest We wish to acknowledge the aid of Joan T. Erber, Bayla M. Myer, and Gayle A. Olson in serving as judges. We have a limited supply of the complete normative data. A copy may be obtained by writing to the senior author.
is the nature of word associations to homographic stimuli, in that these associations provide an effective source material for selecting priming words and for analyzing the differential salience of a given homograph's meanings. Many associations may be viewed as response words which share a large number of features, and which offer a minimal feature contrast, with the stimulus words (McNeill, 1966). Thus, for a single S the association to a given homograph is likely to reflect the particular meaning, or set of features, perceived at the moment of responding, and for a group of Ss the associations to that homograph should be distributed in proportion to the frequencies with which the separate meanings are perceived. The best available predictor of the differential frequencies for the separate meanings is the information provided by semantic counts. A spew-like principle (Underwood & Schulz, 1960) leads to the prediction that the more frequently occurring meaning will also be the more frequently perceived meaning in the elicitation of word associations. Although normative studies on word association contain a scattering of homographic stimulus words (e.g., Bousfield, Cohen, Whitmarsh, & Kincaid, 1961), their concentration is insufficient to permit a systematic analysis of the distributions of associations to distinct meanings. The primary purpose of this study is to provide this kind of systematic analysis. In addition, the study makes normative data avail-
444
WORD ASSOCIATIONS TO HOMOGRAPHS
able for a relatively large collection o f h o m o g r a p h i c stimuli.
445
psychology classes at St. Louis University, serving as Ss. An additional 50 men and 50 women, also students in general psychology classes, served as Ss for the sentence construction test.
METHOD
Homographs Word associations were collected for a total of 40 homographs. The reference source was the Thorndike Barnhart Handy Dictionary, a source which'lists meanings of homographs in order of frequency of occurrence in the Lorge-Thorndike (1938) semantic count. The order listed in this source overlaps completely with that given in the Random House Dictionary of the English Language (1967). The homographs were selected so as to include a range of Thorndike-Lorge (1944) values (four per million to AA), a range of word lengths (three to seven letters), and, hopefully, a range of fractionation between component meanings (near equality of separate meanings to high dominance for one meaning). The homographs and their component meanings are given in Table 1.
Procedure The word associations were collected by means of a group test that followed exactly the procedure of Palermo and Jenkins (1964). Two different forms (A and B) of the test were employed. Form A contained 11 homographs plus 29 filler words (nonhomographic words selected from the Palermo-Jenkins stimulus words), and Form B contained 29 different homographs plus 21 filler words. Two different random orders of words were used in preparing booklets for each form. The associations to each homographic stimulus were scored independently by five judges, the two Es and three graduate students in human learning. The judge's task was to interpret the meaning of the homograph that served as S's functional stimulus for each association and to assign a meaning designation (i.e., first or second meaning as listed in the reference source) for each association. For validity purposes a group sentence construction test was administered to a separate group of Ss. The test contained 20 words, the 11 homographs of Form A plus 9 filler words. The S's task was to write a sentence for each word that would contain that word in any capacity. Each sentence containing a homograph was then scored by the five judges in terms of the meaning designation implied by the sentence.
Subjects For each form of the word association test there were 100 men and 100 women, students in general
RESULTS AND DISCUSSION A n associate was considered related to a given m e a n i n g o f a h o m o g r a p h only if there was c o m p l e t e a g r e e m e n t a m o n g the five judges. A l l other associates were scored as being a m b i g u o u s . Table 1 gives the percentages o f m e n a n d w o m e n who r e s p o n d e d to the separate meanings o f each o f the 40 h o m o graphs. The p r i m a r y a n d s e c o n d a r y associates for each h o m o g r a p h are also listed in Table 1. The f r a c t i o n a t i o n o f h o m o g r a p h s a c c o r d i n g to meanings m a y be seen to v a r y widely, f r o m those which have virtually equal p r o b a b i l i t i e s o f meanings (e.g., pitcher a n d seal) to those which have a clearly d o m i n a n t m e a n i n g (e.g., al-m a n d date). F o r p u r p o s e s o f c r o s s - v a l i d a t i o n the m e n a n d w o m e n o f this s t u d y were treated as separate p o p u l a t i o n s . C o r r e l a t i o n s between percentage scores o f m e n a n d w o m e n were then c o m p u t e d for b o t h the p r i m a r y a n d s e c o n d a r y meanings (as indexed by the reference source). The degree o f a g r e e m e n t between the sexes was very high for b o t h the p r i m a r y a n d the second a r y meaning, r ( 3 8 ) = . 9 7 a n d .96, respectively, p < .01. A d d i t i o n a l v a l i d a t i o n o f the percentage scores was p r o v i d e d b y the d a t a f r o m the sentence c o n s t r u c t i o n test. The degree o f agreement between the percentage o f p r i m a r y meanings as m e a s u r e d s e p a r a t e l y by w o r d association a n d b y sentence construction was m o d e r a t e l y high, rho(9) = .80, p < .01, for men, a n d .73, p < .02, for w o m e n . Parallel correlations for the percentage o f s e c o n d a r y meanings were .88, p < . 0 1 , a n d .75, p < .02. The results offer fairly convincing s u p p o r t for the spew-like principle. T h a t is, the semantic c o u n t was m o d e r a t e l y successful in predicting the d o m i n a n t , or m o r e frequently r e s p o n d e d to, m e a n i n g o f h o m o g r a p h s . T h e
446
KAUSLERAND KOLLASCH
dominant meaning in word association was also the more frequently occurring one in semantic counts for 30 of the 40 words in the male sample and for 31 of 40 in the female sample. These proportions clearly exceeded chance expectancy, X2 (1) = 9.02, p < .01, for men, and X2 (1) = 11.02, p < .01, for women. In addition, as may be seen in Table 1, primary associates were generally related to the more frequently occurring meaning in semantic counts (31 of the primary associates for men and 37 for women). This trend extended through the tertiary associates as well, with 32 of the tertiaries for men and 27 for women being related to the more frequently occurring meaning. Nevertheless, a sizable number of homographs yielded associative hierarchies for which the less occurring meanings were
dominant. Post-hoc interpretations of these exceptions are readily available. Words like pitcher and poker may reflect the uniqueness of the culture sampled--the baseball pitcher Bob Gibson is a potent attraction in St. Louis, poker is likely to be a popular pastime among college students, etc. Words like can and like receive their primary semantic counts through their use as functional words. Such words are unlikely to have distinctive semantic features which would encourage associative responding on the basis of matching the features. Consequently, S's associative responding is directed to the alternate meanings, meanings for which semantically rich features are available for matching purposes. Testing these and related hypotheses presents an interesting area for future investigation.
TABLE 1 WORD ASSOCIATIONS FOR 40 HOMOGRAPHS : PERCENTAGE OF ASSOCIATES FOR EACH MEANING AND THE PRIMARYSECONDARY ASSOCIATE FOR EACH HOMOGRAPH
~oAssociationsa Homograph
Men Women
Primary associateb Men
Women
Secondary associateb Men
Women
Arm
1. upper limb 2. supply weapons Base 1. bottom; station 2. mean; selfish Bass 1. deep in sound 2. fish Bat 1. club 2. flyingmammal Bowl 1. rounded dish 2. a game
93 1
97 2
Leg (35)
Leg (39)
Hand (21)
Hand (26)
85 1
88 3
Ball (54)
Ball ( 5 3 )
Baseball(5)
Bottom (6)
57 40
52 44
Guitar (12)
Drum (9)
Fish (35)
Fish (38)
62 36
68 29
Ball (52)
Ball (62) Man ( 1 9 )
A n i m a (5) l
65 22
73 13
Soup (11)
Cereal (10)
Soup (11)
84 6
89 2
Cardboard (18) Square (12)
Square (7)
Container(10)
73 19
64 29
River (17)
Water (13)
Water(16)
Cereal (14)
Box
1. container 2. strike; fight Bridge 1. structure 2. card game
River (18)
447
WORD ASSOCIATIONS TO HOMOGRAPHS Table I--continued 9/00Associations a Homograph
Men Women
Primary associate b Men
Women
Secondary associate b Men
Women
Can
Do (9)
Will (9) Opener (9)
Boy (25)
Today (12)
Today (7)
Water (13)
Quack (25)
Goose (10)
Water (13)
Hair (10)
Blonde (9) Dark (9)
Cheat (9)
65 28
Cabinet (28)
Cabinet (30)
59
66
Good (17)
2. sum of money
30
24
Fit 1. suitable; proper
83
91
42 46
35 58
Food (10)
Tin (12)
99
97
Girl (37)
1
2
Duck i. bird 2. plunge; lowerhead
78 6
84 4
Fair 1. just; beautiful
78
70
10
17
File 1. set of papers 2. steel tool
70 18
Fine l. excellent; small
1. to be able 2. metal container Date 1. time 2. fruit
2. display of goods
Nail (7) Good (20) Money (12)
Wear (9), Tight (9)
Clothes (18)
Nail(s) (20) Bad (6), Well (6) Money (6), Ticket (6) Tight (13)
4
6
Hide 1. keep from sight 2. skin
85 6
93 5
Seek (44)
Seek (51)
Find (13)
Find (17)
Host 1. receives guests 2. large number 3. Eucharist
79 0 8
83 0 14
Hostess (33)
Hostess (47)
Guest (16)
Guest (16)
Jam 1. press tightly 2. preserve o f fruit
21 77
5 94
Jelly (43)
Jelly (54)
Bread (8), Butter (8)
Toast (11), Bread (11)
Lead 1. show the way 2. metal; graphite
34 54
40 54
Follow (18) Pencil (29)
Pipe (13)
Lie 1. not true 2. prone position
57 30
62 33
Truth (27)
Like 1. similar 2. pleased with
12 65
7 81
2. disease
Love (21)
Follow (26)
Truth (19)
Love (40)
Down (17)
Down (13)
Dislike (16)
Hate (30)
448
KAUSLER AND KOLLASCI-I Table 1--continued Associationsa Homograph
Men Women
Primary associate b Men
Mean 1. have in mind 2. low in quality 3. halfway
3 57 29
8 77 8
Mine 1. belonging to me 2. hole in earth
68 28
77 21
Yours (49)
Mint 1. plant used for flavoring 2. money
71 25
87 10
Candy (32)
Miss 1. fail to hit 2. young woman
10 83
8 85
Net 1. open fabric 2. profit; gain
73 20
Page 1. one side of paper 2. boy servant
Women
Nice (21)
Secondary associate b Men
Women
Angry (10)
Nasty (8)
Gold (11)
Coal (11)
Average (12) Yours (65)
Green (12)
Candy (46) Money (21)
Mr. (17)
Mrs. (32)
Mrs. (14)
Mr. (17)
84 12
Fish (40)
Fish (39)
Catch (6)
Hair (13)
85 8
91 6
Book (56)
Book (57)
Number (6)
Number (10)
Peer 1. equal; man with title 2. look closely
46 28
56 25
Friend(s) (20)
Equal (16)
Look (17)
Perch 1. roost for birds 2. fish
62 33
60 38
Fish (31)
Fish (34)
Pitcher 1. container for liquids 2. baseball player
42 55
52 45
Water (24)
Water (25)
Poker 1. metal rod 2. card game
8 86
8 87
Riddle 1. puzzling question 2. make many holes
82 3
Ring 1. circle 2. sound
Look (15) Bird (30)
Bird (33)
Baseball (14)
Baseball (16)
Cards (24)
Cards (30)
Game (15)
Game (23)
96 0
Joke (22)
Rhyme (35)
Rhyme (20)
Joke (29)
78 14
82 9
Finger (19)
Finger (19)
Wedding (13)
Wedding (15)
Rock 1. stone
54
74
Stone (30)
Hard (14), Stone (14)
Hard (15)
2. sway
42
23
Roll (24)
46 42
51 41
Close (12)
Stamp (8)
Animal (12)
Seal 1. design on wax; shut 2. animal
Zoo (10)
449
WORD ASSOCIATIONS TO HOMOGRAPHS Table 1--continued Associations a Homograph
Men Women
Second 1. next after first 2. unit of time
67 23
59 33
Sole 1. only 2. bottom of foot 3. fish
24 42 21
20 55 11
Sow 1. scatter on ground 2. female pig
44 35
55 29
Spade 1. shovel 2. black figure; cards
64 32
74 23
Tire 1. weary 2. wheel
2 94
1 98
Wind 1. air 2. twist or turn
88 4
80 2
Primary associate b Men
Women
First (44)
Secondary associate b Men
Women
Minute (16)
Minute (15)
First (39)
Only (11) Shoe(s) (32)
Shoe (50) Fish (15) Seed(s) (16)
Pig (29)
Pig (24)
Shovel (32)
Shovel (27)
Seed(s) (23)
Hoe (14) Card(s) (18)
Car (40)
Car (55)
Wheel (10)
Wheel (14)
Blow (34)
Blow(s) (30)
Cold (9)
Breeze (7)
"The remaining associations were ambiguous with respect to the meaning referent. b The numbers in parentheses refer to the percentage of Ss giving that association.
REFERENCES BENJAMIN,T. B., & WATT, N. F. Psychopathology and semantic interpretation of ambiguous words. Journal of Abnormal Psychology, 1969, 74, 706714. BOUSEIELD,W. A., COHEN, B. H., WHITMARSH,G. A., & KINCAID, W. D. The Connecticut free association norms. Technical Report No. 35, 1961, University of Connecticut, Contract Nonr-631 (00), Office of Naval Research. CRAMER,P. Mediated priming of polysemous stimuli. Journal of Experimental Psychology, 1968, 78, 137-144. KATZ, J. J., & FODOR, J. A. The structure of a semantic theory. Language, 1963, 39, 170-210. KAUSLER, D. H., & KAMICHOFE,N. C. Free recall of homographs and their primary associates. Journal of Verbal Learning and Verbal Behavior, 1970, 9, 79-83.
LORGE, I., & THORNDIKE, E. L. A semantic count of English words. New York: Teachers College, 1938. McNEILL, D. A study of word association. Journal of Verbal Learning and Verbal Behavior, 1966, 5, 548-557. PALERMO, D. S., & JENKINS, J. J. Word association norms: Grade school through college. Minneapolis: University of Minnesota Press, 1964. THORNDIKE,E. L., & LORGE,I. The teacher's wordbook of 30,000 words. New York: Teachers College, Columbia University, 1944. UNDERWOOD,B. J., & SCHULZ, R. W. Meaningfulness and verbal learning. Philadelphia: Lippincott, 1960.
(Received February 24, 1970)