THE PURPOSE OF THIS EXPERIMENT WAS TO COMPARE THE LANGUAGE BEHAVIOR OF STANDARD AND NONSTANDARD ENGLISH SPEAKERS WHEN ASKED TO REPEAT STANDARD AND NONSTANDARD SENTENCES. THE SUBJECTS (47 THIRD AND FIFTH GRADERS AT A NEGRO SCHOOL IN WASHINGTON, D.C., AND 30 OF THEIR WHITE COUNTERPARTS AT A SUBURBAN MARYLAND SCHOOL) WERE ASKED TO REPEAT 30 TAPED SENTENCES AS BEST THEY COULD. FIFTEEN OF THESE SENTENCES WERE IN STANDARD ENGLISH AND 15 WERE IN NEGRO NONSTANDARD DIALECT. WHITE SUBJECTS DID SIGNIFICANTLY BETTER THAN NEGRO SUBJECTS IN REPEATING STANDARD ENGLISH SENTENCES. LIKEWISE, NEGRO CHILDREN WERE SIGNIFICANTLY BETTER IN REPEATING NEGRO NONSTANDARD SENTENCES. THE AUTHOR FEELS THAT THE FACT THAT BOTH GROUPS EXHIBITED SIMILAR BEHAVIOR PATTERNS WHEN CONFRONTED WITH SENTENCES OUTSIDE THEIR PRIMARY CODE INDICATES THAT THE LANGUAGE DEFICIENCY THAT HAS BEEN ATTRIBUTED TO THE LOW INCOME NEGRO CHILD IS NOT A LANGUAGE DEFICIT SO MUCH AS A DIFFICULTY IN CODE SWITCHING WHEN THE SECOND CODE (STANDARD ENGLISH) IS NOT AS WELL LEARNED AS THE FIRST (NONSTANDARD ENGLISH). IN A SECOND TASK INVOLVING IDENTIFICATION OF THE RACE OF THE SPEAKER, HIGH PERCENTAGES OF WHITE AND NEGRO CHILDREN IDENTIFIED A STANDARD ENGLISH SENTENCE AS BEING SPOKEN BY A WHITE MAN AND A NONSTANDARD SENTENCE AS BEING SPOKEN BY A NEGRO. (DO)