The Significance of Vocabulary in the Interview
- Author:
Gladys C. Schwesinger
There are but few objective facts to offer relative to the vocabulary of the interviewee. With the exception of studies on the very young child, whose knowledge and use of words could not be measured through any other medium than that of oral speech, and with the exception also of Snedden’s2 work on adolescents, there are no outstanding researches on oral vocabularies. Most of the studies have been conducted by the written, pencil-and-paper, formal testing method, and conducted largely on groups of people at a sitting. The vocabulary of the interviewee, that is, his use and knowledge of the spoken word, has not received its fair share of attention, doubtless because of inherent difficulties involved in such.studies.
The best justification for reporting on what vocabularies may be expected to indicate is the available evidence 011 the significance of vocabulary. Studies have been carried out from time to time relating language growth to age, to grade reached in school, to the type of school attended, to studies of major interest followed in school or college, to place of residence, and to other environmental factors, such as extent of travel, contact with delinquency and the like. General vocabulary strength has been studied in relation to general knowledge, and special vocabulary strength in relation to special knowledge in any particular field. Experiments have been carried 011 to see how far vocabulary may be indicative of social background and the general cultural level of the home. In addition, the close parallelism between knowledge of words and general intelligence has been reported upon by several investigators. These various explorations of the field of word knowledge have not been formulated with the same objectives. I11 some instances, attention has been centered on discovering the size and range of a person’s word equipment; at least insofar as it concerned his knowledge of what words mean. In other cases, the investigator has studied the range of words actually used by the subject, or the quality of the words chosen by the subject. For instance, does 1 Read at the annual meeting of the National Vocational Guidance Association, Atlantic City, Feb. 21, 1930. 2 Snedden, D. S. A study in disguised intelligence tests. T. C. Contrib. Educ. No. 291, 1927.
he tend to express himself in slang or in good English? Is he inclined to use a “high-brow” or a “low-brow” word in a given situation? His pronunciation, articulation, inflection and other speech factors have also been observed and related to personality. The methods employed for measuring and weighing the use and understanding of words vary with the objective of the particular research. Even when pursuing one objective, for example, in the effort to compute the size of a person’s vocabulary, there is 110 one standard method. With very young children, it consists in counting the words actually used by the child. This means that for a representative period in a little child’s life, a few days, a few hours each day, perhaps for a few weeks, a watchful adult, who has cultivated an unobtrusive manner, must hover near the youngster to gather the pearls as they fall from his lips, and store them up in a list to be counted latex*. This technique has been called by Rugg,2 that of “the eye-witness analyst.” With older children who can read, and with adults, the method undergoes a change. Now the examiner is concerned not so much with the words actually used as with the words which are understood by the subject. Here the method is to apply a formal test, consisting usually of a certain number of words chosen from a dictionary. The score consists of the number of words whose meanings are known. This number is then multiplied by a constant (depending 011 the proportion of the number of words selected to the total number in the dictionary). In this way, the subject’s total range of vocabulary is estimated. To indicate that he knows a word, the subject is usually required to give a verbal explanation of its meaning, or in other words, a definition. Or he may be supplied with “multiple choice” answers, from which he is required to indicate the one word which in his judgment best explains the meaning of the stimulus word. Other experimenters present the stimulus word set in a sentence. The subject’s understanding of the word is shown by the appropriateness of his answer to the question. This last method of measuring word knowledge lends itself admirably to adaptation for use in the interview. It is the method which Sneddon employed in his disguised intelligence tests. It has its difficulties and complications, but its diagnostic possibilities are promising.
2 Rugg, H., Krueger, L., and Sondergaard, A. Study of the language of kindergarten children. J. Educ. Psychol., 1929, 20, 1-18.
The nature of the data accumulated in the various researches to which reference has been made may be briefly presented. The first point for consideration is the matter of sex differences. Up to the age of four or five, girls have a larger vocabulary than have boys; after this age, the boys outstrip the girls. This is explained by the fact that boys come earlier in contact with a wider environment, while girls stay more within the shelter of the home. This is mentioned, in passing as an early and interesting indication of the dependence of verbal ability upon environmental influence. Other facts collected about the speech of the pre-school child, need not be discussed as they are not immediately relevant. It is desirable, however, again to call attention to the method used, namely that of actually noting and counting the words employed by the subject during a given period of time. This method, with modifications, might advantageously be applied to a study of the words employed by the interviewee. In a set interview, standardized and controlled, more or less the same subject matter will always be covered. The language in which this material is couched will vary with the intelligence and training of the adult individual, perhaps as much as it does with the age and intelligence of the pre-school child. Further research work might well be done along this line. By way of illustration, reference may be made to cases encountered in a public employment bureau which is usually patronized by people of the lower social and intellectual levels. From among these can be selected two applicants; both women were just as dowdy and as untidy as any in the room; both were raggedly dressed, both more or less unkempt as to hair, nails, shoe laces, and the like; both decidedly restricted as to early schooling and later bookish interests, and so on. The first impression was, in each case, that they were little, if at all, above moron intelligence; but after talking with each woman, it was necessary to reverse this estimate. One, it was discovered, used fairly hard words rather easily; the other, although not inclined to express herself in speech, followed the use of the upper range words, quite understandingly. These observations led to a change in the rating of their intelligence and to crediting them with at least average mental ability. The conversation in each case was never allowed to stray beyond the limits set by the standardized interview. The sole reason for the shift in the estimate of their mental ability was the word knowledge indicated by the women.
From, many different studies we have evidence to show that vocabulary grows as knowledge grows. There is a progressive increase in growth of language ability from grade to grade. There is a definite and positive correlation between one’s score on a general vocabulary test and a general knowledge test. This correlation is so large that, to save time, a vocabulary test can be given in place of a general information test. The same holds true for measuring knowledge within a special field. One investigator, Babbitt,3 as far back as 1907, found that he could use a vocabulary test to classify students in a German class. Later he found that this held also for English. Kennon4 has offered a vocabulary test for teachers of high school English which in itself provided the best single measure for predicting success within the field of teaching English literature. Whipple5 has devised a word list to measure specific knowledge of any one of several fields. Familiarity with American history, for instance, will be tapped by the subject’s knowledge of “Anthony Wayne”; or of French by ” aujourdhui”; of social usage by “R.S.V.P.”; of golf by “midiron,” and so on.
Again to illustrate the significance of this for the interviewer, consider the case of the student who presented herself as an applicant for training in the psychological clinic of a hospital. Ker only groundwork was a half course in the elements of psychology, and a most old-fashioned brand of psychology at that. The psychiatrist Avho interviewed her used technical words of which she had never heard, such as “reactions,” “orientations,” “I.Q.,” and the like. Needless to say her ignorance of the fundamental terminology of the profession convinced her that she was not ready for the practical training which she sought, and she was glad to return to the college classroom to learn some modem psychology. The psychiatrist’s discovery of her inadequate preparation was doubtless based on her own frank admissions, but if she had chosen to obscure this lack of background, he could easily have found her out by this other method?that of using technical words and gauging her reactions, to note whether they indicated professional comprehension or lack of it.
a Babbitt, E. II., A vocabulary test. Pop. Sc. Mo. 1907, 70, 378. * Kennon, L. II., Tests of literary vocabulary for teachers of English. T. C. Contrib. Educ. No. 223, 1926. 5 Whipple, G. M., Manual of mental and physical tests. Part II. Baltimore, Md.: Warwick and York, 1915. (Pp. 308-332, 674-688.)
A fourth general line of evidence on factors behind vocabulary comes from research studies which attempt to relate vocabulary to social and cultural background. There is a definite and fairly high correlation between knowledge and use of words and the family social and cultural level. Investigators recognize that good social upbringing is associated with good general intelligence and that good intelligence is associated with good vocabulary. However, the relationship between social background and vocabulary is not entirely accounted for by the factor of general intelligence, inasmuch as when intelligence is held constant by the partial correlation technique, the correlation is still significantly high (.54) .6 This indicates the presence of something else, presumably home training, which accounts for the vocabulary development. Van Alystine7 found that the vocabulary of three-year-old children is slightly more related to a composite of environmental factors (.70) than it is to intelligence score (.61). The difference is not large but is better than chance. Burks,8 in California, studying foster and own children of an older age suggests that vocabulary may be more closely related to environment than intelligence is. The writer,6 in testing grade and high school children, found that the relation between vocabulary and social background, measured by two different scales, was .54 and .72 respectively.
To illustrate the powerful force of home training on vocabulary development, an extreme case may be cited. This young man of twenty comes from an unusually cultured home. His father is an author of international fame; the other members of the family are consistently superior intellectually and vocationally, but this boy’s intelligence, according to the Stanford-Binet, is about on a moron level. In appearance, he somewhat resembles the Mongoloid type. His grasp of school subject matter such as arithmetic, spelling, even writing, is on a third to fourth grade level, but his word knowledge and general command of language is at least ninth or tenth grade. He has been well-brought up, is well mannered, 6 Schwesinger, G. C., The social-ethical significance of vocabulary. T. C. Contrib. Educ. No. 211, 1926.
7 Van Alystine, D., The environment of three-year-old children. Factors related to intelligence and vocabulary tests. T. C. Contrib. Educ. No. 3G6, 1929. 8 Burks, B. S., The relative influence of nature and nurture upon mental development. Twenty-seventh yearbook, Nat. Soc. Study Educ. Part I, 1928, (pp. 219-31G).
well-spoken, and unless one penetrates into any special subject of conversation, one would not discover his intellectual limitations. The discrepancy between his attainments and his verbal ability can be accounted for almost entirely by his superior home environment, where rich and wide vocabularies have always been in daily usage. If a group of such subjects could be gathered from feebleminded institutions and elsewhere, that is, a group of people whose language responses are considerably superior or considerably inferior to their general intelligence level, and if the relationship of social background and vocabulary in this group could be studied, it would doubtless throw more light on the degree of dependence of language growth on environmental supplies.
This inter-relation of word knowledge and social background has been studied from still a different angle. Here the quality rather than the extent of one’s vocabulary has been considered. For example, Wyman9 in California studied whether children were socially, intellectually or actively interested by their response words to her free association test. The writer10 once undertook to look for possible group differences in knowledge and use of slang and “underworld” language. The tests were applied to two groups, equated for social and intellectual traits, but differing only for the factor of delinquency, the one group being in a state reformatory, the other at large. The experiment was repeated on two other groups, also equated for background and intelligence, but not so widely separated from the standpoint of conduct. In the second experiment, the differences proved to be not so marked as in the first group. If this same experiment could be tried on different cultural levels, the results might be illuminating.
This same tendency can be noted in children’s responses to the word “curse” in the Stanford-Binet vocabulary. Young children from the lower socio-economic groups almost invariably, whatever their other limitations, know the meaning of this word. Children of the same, and older ages, in a high grade private school, representing a high socio-economic level, whatever their intellectual attainments, almost invariably fail on this word. “Wyman, J. B., Interest tests of a group of gifted children. Terman’s Genetic studies of genius, Vol. 1, Stanford Univ., Calif.: Stanford Univ. Press, 1925. (Pp. 455-483.) 10 Schwesinger, G. C. Slang as a factor in character. J. Appl. Psychol. 1926, 10, 245-263.
A fifth, and more fruitful line of evidence on factors back of word knowledge, comes from the data gathered on the relation of general intelligence to language. This relation is so well supported by experimental findings that it need not be discussed at length. Terman11 states that his vocabulary test is the best single measure of whatever the whole Stanford-Binet test measures, and claims that, in a large majority of cases, the vocabulary test alone would give an I.Q. within 10 per cent of that secured by the entire test. Other psychologists have added more evidence to this. Still others have pointed out that this high relationship is due to a spurious factor, because intelligence tests are heavily loaded with vocabulary elements. But even so, the fact remains that a vocabulary test measures whatever a general intelligence test measures, and makes that measurement with almost equal effectiveness.
Hence, if it is possible in the course of an interview to get an estimate of,the vocabulary of the interviewee, one will, at the same time, have secured a fairly reliable guage of his general intelligence. People will often consent to taking a “vocabulary” test, who would balk at an “intelligence” test. Van Alystine resorted to this device most effectively when she had mother and child each taking a “vocabulary” test. In her experiment it was possible to use the pencil and paper method. But in an oral interview, when interviewer and interviewee meet for perhaps the only contact which they may ever have, the technique devised by Sneddon is suggested, llere words of graded difficulty are introduced in simple sentences in such a way, that if the interviewee understands the meaning of the key word, his knowledge will be revealed by his answer. He can then be led through a scale of words which become harder and harder until the subject’s limit is reached.
The chief defect of this method is that the interviewer must be sure that the interviewee actually knows the meaning of the word; or that, in other words, he is not “bluffing.” Questions which might be answered by “yes” and “no” are to be avoided. These and other complications render the “disguised” method of testing intelligence somewhat impracticable, but with further experience, the difficulties should be overcome. 11 Terman, L. M., Measurement of intelligence. Boston: Houghton, Miflin, 191G (p. 230.)
Summary
With the exception of the work on pre-school children very little has been done on the oral vocabularies of children or adults.
The importance of pursuing such research is indicated by the significant results obtained from studies on groups or individuals to whom formal vocabulary tests have been applied.
Methods have varied from that of counting the words actually used in a given time, to formal definition requirements, and more recently to presenting the test word in a sentence and requiring an understanding answer. This last method, which has most promise for the interview, has not yet been perfected.
The influences at work in determining the quality and extent of a person’s vocabulary are: general intelligence; chronological age up to adult mental levels; environment, consisting of such determiners as education, general knowledge, social and cultural background, special interests, and special experiences. Knowledge of what good and poor vocabularies signify may guide the observation of the interviewer in regard to these factors.
Much more research is needed before these factors can be designated as facts.
Disclaimer
The historical material in this project falls into one of three categories for clearances and permissions:
Material currently under copyright, made available with a Creative Commons license chosen by the publisher.
Material that is in the public domain
Material identified by the Welcome Trust as an Orphan Work, made available with a Creative Commons Attribution-NonCommercial 4.0 International License.
While we are in the process of adding metadata to the articles, please check the article at its original source for specific copyrights.