Open main menu

Popular Science Monthly/Volume 70/February 1907/A Vocabulary Test

By Professor E. A. KIRKPATRICK


OF all the inventions of the human race nothing compares in importance, as regards mental development, with language. In the development of each person also, nothing exercises a greater influence in molding and developing thought and feeling than his language environment. The vocabulary of a person represents in a condensed and symbolic form all that he has experienced and imagined. The breadth of his mental experience is indicated by the number of words that have for him a meaning, while the accuracy of his thinking is shown by the constancy and exactness of meaning with which he uses words. The study of vocabularies ought therefore to be an important branch of psychological investigation.

Studies have been made of the number of words used by great writers, and by children a few years old. The latter studies have shown that a child may not use words that are perfectly familiar to him for months merely because he has no occasion to use them, e. g., words frequently uttered in the summer or when in the country may never be used in the city or in the winter. Adults are familiar with many words that they have rarely, perhaps never, used. The difficulties in the way of counting accurately the number of words used by an adult or even by a child over three years of age are almost insurmountable.

When we attempt to estimate the number of words that have a meaning for an individual, the difficulties are less although the number of words is much greater. The writer long ago estimated the number of words in his own vocabulary by going carefully through an unabridged dictionary and counting the number of familiar words on every tenth page (see Science, O. S., Vol. XVIII., pp. 107-108). Since then he has often had his students estimate the number of concepts that they possessed by counting the number of words that had for them a fairly definite meaning, on a few pages of the dictionary, and then calculating from the proportion of familiar words the total number of words they knew.

When a student began, say on page 2, and counted all the words in bold-faced type and the number of these known on every fiftieth page, and then did the same beginning with page 20, the results were so nearly the same as to convince me that the method was fairly accurate. Some preliminary tests were then made that showed that a hundred words taken by chance from various parts of the dictionary might serve as a fairly accurate measure of the size of one's understanding vocabulary. The words used in the final test consisted of fifty words taken from the first four words on every fiftieth page of Webster's academic dictionary and fifty words from the first of other pages leaving out different forms of the same root word (e. g., photograph, photographer). This was done with the thought that older persons might be able to infer better the meaning of unfamiliar words than younger persons. The results were negative and the author now considers that the best list of words is obtained from Webster's academic dictionary (which contains about 28,000 words on 645 pages), by taking the first, second, or last word, or any other definite word on every sixth page. For general purposes and for all ages this is probably better than to take a hundred words from an unabridged dictionary which contains so many various and obsolete forms of the same words, along with rare words, and technical terms not found in the smaller dictionary. Estimates based on words from the academic dictionary give less than half as many words in the vocabulary as those based on data from the unabridged, but they are more representative of fundamentally different concepts.

The method of using the test was to place the printed list before the subjects and ask them to mark the words that they knew with a plus () sign, those that they did not know with a minus () sign, and doubtful ones with a question mark (?). The tests which numbered about two thousand were made chiefly upon pupils from the fourth grade up through the high school and university, although a few were made upon younger children. Control tests showed that if the same test was given orally, there was some difference in the words marked as known and unknown. This difference was of course very great in the second and third grades, where a few tests were made, and became less with age, yet it usually amounted even in the case of adults to from one to three per cent. In a few individuals the difference was quite marked.

The reason for this is that some words are more often heard than others, while others are more often seen, hence in one case the auditory stimulus arouses familiar associations while in the other case the visual stimulus is more effective. In general the auditory stimulus" is more effective for children, but, as they read more, the visual stimulus becomes more effective and later many words are seen that are rarely or never heard; hence for such words the visual stimulus is the most effective and sometimes the only stimulus which will produce the reaction of familiarity. The test is more accurate if both forms of stimuli are used, i. e., the words pronounced as the pupils look at them.

There is another cause of difference and also of inaccuracy. In the auditory test unfamiliar words are often mistaken for familiar ones having a similar sound, e. g., barque for bark, baron for barren, and in the visual test similarity of appearance plays a similar part. A striking case of this form of error was made by a third grade boy who marked the word amaranth as known. I said to him, 'You don't know that word, do you?' He said, 'Yes,' in a tone that implied surprise that I should question it. I then said, 'What is the word?' He replied, 'Arithmetic' Another boy for similar reasons, partly visual and partly auditory, marked 'eschar' as known and when questioned called it 'sister.'

On the other hand, young children often do not mark words that are perfectly familiar to them, because the sounds and forms without any other stimuli of suggesting words or circumstances are not sufficient to immediately arouse the sense of familiarity. One second grade boy who marked only eighteen words in the test, when questioned, showed by synonyms or definitions, or illustrations, that he knew the meaning of thirty of the words.

Individual habits of thinking or judging is probably the largest factor in tending to make the marking of words an unreliable index of the actual mental furniture of the subject of the test. Some mark as known every word that arouses the feeling of familiarity, while others mark as known only those for which they are confident they can give a correct definition. The differences in this respect are, however, most shown in the doubtful marks while the plus mark usually means the arousal of a specific idea by the word form. This idea may be vague or distinct, narrow or broad, general or detailed, correct or incorrect, but it is the idea usually aroused by the word.

Upon defining a list of words to a class of normal students after they had marked them, it was found that the errors in marking words as known and unknown usually cancelled each other, so that the number finally reported as known and unknown was for most members of the class about the same as when they were first marked.

Instruction as to what shall be the standard for deciding whether a word is known, such as "Count as known all words that you would not, as to their meaning, need to look up in a dictionary if you saw them in a sentence," helps to render the marking more uniform. Another and more accurate method of bringing about uniformity of standard is to ask the pupils to define or put in sentences some of the words, then to mark the rest according as they think themselves able or unable to indicate their meaning.

If students are asked to define a certain proportion of the words as accurately as possible, giving all meanings where there are more than one, depth and accuracy as well as breadth of knowledge may be tested. In college classes where twenty of the hundred words were defined, 114 out of 246 students were found to have denned the same proportion of words that they marked as known and only seventeen showed a difference of as much as three words of the twenty from the corresponding proportion of the hundred words marked. The overestimations slightly exceeded the under estimations.

The author is convinced that one hundred words selected as has been described and marked with care gives sufficient basis for an approximate estimate of the size of the understanding vocabulary of college and high-school students, and of the higher grades of the grammar school. In the author's own classes where students were ranged in three grades according to the number of words marked as known in one list of words, other lists of words similarly selected resulted in 60 per cent, to 80 per cent, of them being again in the same grade, while none were changed from the lowest to the highest grade.

Using Webster's Academic Dictionary as a basis it appears from averaging about two thousand papers that the size of vocabularies are likely to approximate the following:

Grade II 4,480 Grade III 6,620
Grade IV 7,020 Grade V 7,860
Grade VI 8,700 Grade VII 10,660
Grade VIII 12,000 Grade IX 13,400
High School.
Freshmen 15,640 Sophomore 16,020
Junior 17,600 Senior 18,720

The average for normal school students is 19,000 and for college students 20,120. The colleges represented in this test were Bryn Mawr, Smith, Columbia, Brown University and Pratt Institute, while the grades and high schools were mostly in Massachusetts cities.

There seems to be no constant difference between the sexes. On only a part of the papers was age given, but there is reason to believe that vocabularies increase up to thirty. In Pratt Institute where students varied greatly in age, those above twenty-five knew from five to ten per cent, more words than those in the same classes who were below twenty years of age. It is not likely that the growth of vocabulary is great after thirty, when deeper specialized and executive activities have taken the place of general advancement into new fields of knowledge and many words once known are forgotten.

One important factor in the growth of vocabularies was investigated by accompanying the list of words with a request to write names of papers and magazines frequently read and of books read since the beginning of the year. It was found that in general those who named the most books and magazines had the larger vocabularies, regardless of their grade.

The individual differences in size of vocabulary were very great, some ninth grade children falling to the rank of second grade children, while some third or fourth grade children ranked with the average of those in the ninth grade or high school.

Sometimes a very small vocabulary was accounted for by the fact that the child was of foreign parentage and did not hear English at home, but the mere fact of being of foreign parentage was no assurance that the vocabulary would be small.


The relation of size of vocabulary to school standing was considered, but owing to the scarcity of data and uncertainty as to its reliability (only a small proportion of the papers were accompanied by the class records or teacher's estimate of ability), no conclusive results were reached. In the grades there was no clear proof of relationship though in one room, where there was reason to think the teacher's estimate had been carefully made, the grading corresponded almost exactly to the size of the vocabularies. In one normal class nearly all of those who had been named by the faculty as belonging to the lower third of the class had small vocabularies. In another class there seemed to be little or no relation between size of vocabulary and estimates of teaching ability. In two colleges, one for women, the other for men, the marks given to the women in English and to men in all subjects were secured for the freshman class and compared with the number of words known. The average number of words known by the men who in general ranked in the various subjects above the average of their class was 5 per cent, greater than for those ranking below the average; while the women who ranked highest in English, averaged nearly 4 per cent, better in vocabularies than those who ranked lowest in English.

In the case of individuals there was often a wide divergence between the marks and the size of the vocabulary. In some instances exceptionally poor definitions indicated a difference in the standard used in marking words as known, but not always. This divergence is not, however, greater than between marks in different subjects, e. g., students have honor marks in some subjects and fail to pass in others.

Is size of vocabulary any indication of attainment or ability? An affirmative answer to this can not readily be proved by experiment, because we have no reliable standard of ability and attainment by which the value of the vocabulary test may be determined. It is well known, however, that persons who do well in one subject often do poorly in others and that success in life after school bears little relation to success in school. It has recently been shown by Dr. Thorndike that entrance examinations bear little relation to college marks.

From the side of experimental psychology, no accurate measure of intellectual ability has been established in spite of many persistent and painstaking researches. The various tests used are found to be special in their character. There are also indications that what are good tests at one age or stage of development may have no significance at another stage. Sensory and motor tests are probably valuable indications of mental ability in young children, memory and imagination tests in older children and reasoning tests in youths.

The function of the nervous system is to respond in an appropriate way to the various phases of the stimulating environment. The most common phase of environment to which human beings respond is the word environment, first to auditory words by movements, then to auditory and visual words by images and concepts. The number of words that are known by any person depends upon two factors, the variety in his word environment, auditory and visual, and his own readiness to respond to the various elements of this environment. It is perfectly natural therefore that children who are surrounded by intellectual people or who read a great deal should have large vocabularies and yet that the size of individual vocabularies should vary with their readiness to respond to this word environment. The accuracy of response or quality of knowledge can be judged not by the number of words but by the accuracy of definitions or use of words.

The question naturally arises whether size of vocabulary and ability to define and use words is not a sufficiently accurate measure of the intellectual ability of youths to justify the use of vocabulary tests in examinations for entrance to college. College work is supposed to be general in its character, demanding general ability, of which the vocabulary test ought to give an indication. Of course if students should devote their time to a special study of the dictionary, the test would become special and valueless, since size of vocabulary would not then be an accompaniment and indication of experiences and intellectual advances, but of special study of modes of defining words in terms of other word symbols.


A study of the kind of definitions given by persons of different ages is an interesting indication of the sources of word knowledge and of the modes of thought at different ages.

The first words are of course obtained from direct association with acts and objects and this continues to be a source of vocabulary growth. A large proportion of words, however, come indirectly from experience through the medium of words that have already become familiar. These new words are sometimes received as equivalents of other words, because of synonyms and definitions or of special descriptions. The greater part of them, however, gain their significance from their association with familiar words in various situations, just as the original words were gained from association with various real situations.

These truths may be illustrated by the definitions of gourd given by college students. 'A drinking cup made from the gourd vine.' 'A vegetable which grows in the ground having a hard shell and many seeds.' 'A vessel for holding water or other liquid.' 'A receptacle for carrying water about, usually of skin.' 'A water bottle made from a pumpkin or squash.' 'Vessel sometimes made by scooping out, for example, making a vessel by scooping out a pumpkin.' Evidently most of these definitions represent ideas gained from sentences in which the word, 'gourd' is used, though those who speak of them as 'pumpkins' or as a 'summer squash,' may have seen the real thing without the discriminating eye of the gardener or botanist. The idea that it is a vessel of some kind evidently predominates and this idea is sufficient for interpreting most sentences in which the word occurs.

It is interesting to notice the various forms of the subordinate idea of the object itself as the various persons picture it under the stimulus of the context. 'A shell of certain nuts, fruits and vegetables, or of the cocoanut, squash, cucumber, etc' l In many countries it is used as a receptacle for food and drink.' (A fruit on a tree whose shell is used for carrying water.' 'The dry fruit of some sort of tropical tree.' 'It is hard and round, and some are the size of an apple and rattle when you shake them.' 'A species of dried melon.' 'An old style wooden drinking vessel.' 'A hollow piece of cane.' 'A fruit characterized by the fibrous outer shell similar to the cocoanut.' Few of the writers of the above had a sufficiently correct idea of the article to be able to identify it if it were shown them. They react satisfactorily (to themselves) to the book situation though they would be laughed at by the gardener and botanist. It is an interesting fact that in a prominent college for women the word e decemvirate,' which only readers of Roman history would be likely to encounter, was correctly defined by most of the young ladies, while some could give no definition for gourd, and many others gave such definitions as have been quoted. This is a striking illustration of the difference between the word environment of scholastic halls and that of the industries and the literature of to-day.

The following definitions of gourd are inexplicable until one realizes that one word form has been mistaken for another. 'To spur on' (goad). 'To plunge a weapon into some one, to make a jagged wound' (gored). 'An animal' (goat?). 'A greedy person' (gourmand). 'A chasm or piece of land that is very much lower than the surrounding land' (gorge).

The definitions thus far quoted are by college students, and though most of them are exceptional rather than characteristic of the definitions of college students, they are surprising as well as amusing.

One English teacher was so astonished at the 'depth of ignorance' displayed by the definitions of his freshman class in English that he had all the papers looked over by his assistants, who all agreed that the results were 'shocking.' They, however, saw no relation between the definitions and the scholarship of individual pupils. (As has already been stated the figures show that those ranking high in scholarship knew on an average about 5 per cent, more words than those ranking low in scholarship.)

Character of the definitions changed greatly with age. Descriptions which are so common in the high school and college papers are rarely or never given by children in the kindergarten and primary grades. The same is true of definitions by synonyms and inclusions under larger terms. The younger children nearly always define by mention of some specific incident, e. g., 'A chair is to sit on'; 'Baby stands up by a chair'; 'A bee goes around a piazza and makes a noise.' What anything can do, or what can be done to it, or with it, is of most importance in early knowledge of all things, hence we find the definitions of children expressing action and use more than anything else. Reference to personal experience of self and friends is also common. These facts are of great significance to pedagogy, strongly endorsing the change now being made from the old descriptive 'object lesson' to the better forms of nature study in which use is made the center of interest.