Volume 4 Number 2
January 1965

By Kenneth Stafford

B. L. Whorf speculated from his descriptive studies of American Indian languages that cognitive processes in general are significantly affected by the language variable.’ The relative meditational efficiency (mental representation and manipulation) of different languages, however, has not been experimentally determined. It was the aim of this study to test the hypothesis that problem-solving behavior varies as a function of degree of exposure to two different languages, Navajo and English.

In order to minimize any cross-cultural bias, comparisons in this study were made of language groups within one culture, the Navajo. The language variable was controlled by selecting two groups of children differing in amount of exposure to Navajo and English. One group learned English during preschool years (before six years of age) and knew little Navajo; the other learned English after starting to school (after six years of age). Ratings of teachers regarding proficiency in English and Navajo corroborated these groupings. Seventeen children were in the group who knew English well and Navajo little; 16 were in tile other group. All of them were attending Phoenix Indian School and were from the Navajo reservation in northern Arizona, presumably coming from similar homes (teacher evaluations). The groups were alike in sex ratio, age (14-16 years), and intelligence (teacher ratings). Subjects in the English-speaking group were 9^th graders; subjects in the other were "ungraded."

A problem-solving apparatus was used to test problem-solving ability. It was so designed that its problems presumably were equally fair to English-speaking and non-English-speaking Navajos, and it was assumed that problem difficulty call be systematically varied and that the exact number of trials necessary to solve a problem can be determined.

On the face of tile apparatus (see drawing) was a screen divided into quadrants. Beside each quadrant was a control button to be operated by the subject; above the screen was a reward light. A square and a triangle were flashed on the screen in separate quadrants. The subject pressed a button. If it was the correct one, the reward light flashed. Each time a button was pressed the figures changed position. In this experiment ten consecutive reward light flashes were construed as a solution, and 80 trials were allowed for each problem before presenting another.

The experimental task consisted of six progressively difficult problems, the relative difficulty of which was determined empirically by ascertaining the number of problem solutions and trials per problem solved in a preliminary study of similar subjects. The problems were arranged in order from least to most difficult and presented to each subject in that order. An easy problem, for example, could be solved by pressing the button by the triangle (which required only representation by images). A difficult problem could be solved by pressing the button by the square when the square and triangle were side by side and by pressing the button by the triangle when they were diagonal on the screen (which required more complicated representation and verbalization and placed greater demand on cognitive processes). The solution of each problem except the last involved pressing a button contiguous to a figure, and the solution of every problem was different. The subject was required to discover these facts for himself and thus each succeeding problem made an increased demand on his memory and reason.

The problems were given to each subject individually at the Phoenix Indian School under standardized conditions. All directions were given by pantomime.

In order of difficulty the problems were:

Findings were as follows (statistical tests omitted in this report):

1. There was a small difference, in the number of problems solved, in favor of the children who learned English early. It was defensible to suppose that the difference would have been greater had it been practicable to compare only English-speaking with Navajo-speaking subjects instead of making comparisons between degrees of bilingualism.

2. No significant differences were found in the number of solutions or the trials per solution for easy problems (problems 2 and 3).

3. There were small differences for the hard problems (problems 4 and 5) in favor of those who learned English early. However, (he number of trials per problem solved was less for those who learned English late.

4. Few solutions were made of the very difficult problem 6; however, the early English-speaking group excelled the other in number of solutions, although there was no statistically significant difference. Had there been a larger number of subjects and a more rigorous separation of the groups regarding exposure to English, a significant difference might have appeared. The difference in number of trials required to solve problem 6 was significant, favoring the early English-speaking group.

5. A rank-difference correlation coefficient of .30 was obtained between length of verbal statements of problems and problem difficulty.

Because of lack of rigorous control, this study is largely heuristic and suggestive. However, these results can be construed as supporting the hypothesis that there are differences in problem solving which are traceable to the language variable. It may be that English is more suitable for this kind of problem solving, perhaps offering a more effective means of symbolic representation.

It would appear difficult problems are more amenable to verbal (symbolic representation) solutions but that some, nevertheless, can be solved by representation by images-sometimes even more readily. The present study corroborates previous findings2 in which congenitally deaf children were compared with normals Perhaps the congenitally deaf possess a relatively less adequate symbol system and must rely more heavily upon an imagery system—just as the Navajo-speaking children do. Complexity levels of problems may be more readily solved by representation through images, and others may be amenable to solution by verbal or symbolic representation which facilitates syntactic manipulation. This may lend support to the notion (fiat, even though problems can be solved by a more rudimentary mode of representation, there is a level at which it is expeditious to switch to symbolic or linguistic representation, and, conversely, representation by images is more efficient at simpler levels.

Before starting this experiment it was thought there might be a relationship between problem difficulty and length of verbal statement required to state the problems in Navajo and English. Although the statements in English were somewhat shorter than when they were put in Navajo, the order of increasing length was essentially the same. The low correlation found between length of statement of problems and difficulty of problems may ]lave resulted from some solutions occurring by means of representation by images, not verbalization.

Further efforts are being made to explore in a similar fashion (1) the possibility of mediational interference resulting from two languages in contact and (2) the effects of early bilingualism and late bilingualism on problem solving.

1. Whorf, B. L., Technol. Rev. 42 (1939)

2. Stafford, K., J. sp. hring. Res. 5, 169 (1962)