THIS STUDY SOUGHT (1) TO CLARIFY THE TERM "LANGUAGE APTITUDE," (2) TO BETTER DEFINE ITS RELATIONSHIP TO INTELLIGENCE, AND (3) TO DETERMINE THE ROLE OF SOCIOECONOMIC LEVEL IN THIS RELATIONSHIP. SUBJECTS WERE SIXTH-GRADE PUBLIC SCHOOL STUDENTS, 93 OF WHOM WERE MIDDLE CLASS AND 39 LOWER CLASS. SOCIAL CLASS WAS DETERMINED BY THE OTIS DUDLEY DUNCAN SOCIOECONOMIC INDEX. LANGUAGE APTITUDE WAS MEASURED BY THE MODERN LANGUAGE APTITUDE TEST (LONG FORM), AND INTELLIGENCE WAS ASSESSED BY THE COOPERATIVE SCHOOL AND COLLEGE ABILITY TESTS. FINDINGS SHOW THAT LANGUAGE APTITUDE "IS NOT A UNIFIED DIMENSION OF THE COGNITIVE DOMAIN AS IS INTELLIGENCE." HOWEVER, SUBJECTS FROM DIFFERENT SOCIOECONOMIC LEVELS USE DIFFERENT PROCESSES IN THINKING ABOUT LANGUAGE. MIDDLE-CLASS CHILDREN TEND TO APPROACH ALL TASKS INVOLVING INTERRELATIONSHIPS AMONG MEANING, SOUND, AND SYMBOLS IN THE SAME WAY BUT TO USE A DIFFERENT APPROACH TO SENTENCE STRUCTURE. LOWER-CLASS CHILDREN APPEAR TO USE DIFFERENT APPROACHES FOR TASKS INVOLVING STRUCTURAL RELATIONS, SOUND-SYMBOL MEANING RELATIONS, AND SYMBOL RECOGNITION. THE EXPLANATION FOR THESE DIVERGENT APPROACHES LIES IN THE FACT THAT LOWER-CLASS CHILDREN USUALLY USE TWO DIFFERENT LANGUAGES--"PUBLIC" FOR HOME AND THE NEIGHBORHOOD, AND "FORMAL" FOR SCHOOL. SOME CONFUSION BETWEEN THE TWO LANGUAGES MAY DEVELOP, OR DIFFERENT SETS OF REFERENTIAL MEANINGS MAY BE USED FOR THE LANGUAGE LEARNED AT SCHOOL. THIS PAPER WAS PREPARED FOR THE 1968 ANNUAL MEETINGS OF AMERICAN EDUCATIONAL RESEARCH ASSOCIATION (MARCH 19, 1968). (NH)