IN THIS EXPERIMENT ERRORS WERE SYSTEMATICALLY INTRODUCED INTO A SERIES OF SENTENCES TO STUDY THEIR EFFECT ON RECOGNITION AND RECALL OF THE SENTENCES. THE SUBJECTS, 32 ADULTS, WERE SHOWN SLIDES OF SENTENCES CONTAINING A SINGLE ERROR ("THE IMPATIENT COMPOSED A SYMPHONY FOR THE ORCHESTRA") AS WELL AS CONTROL SENTENCES WITHOUT ERRORS. THE VIEWING TIME REQUIRED TO PERCEIVE AND REPEAT EACH SENTENCE WAS RECORDED. EXPERIMENTAL VARIABLES INCLUDED--TWO BASIC SYNTACTIC STRUCTURES, THREE ALTERNATE POSITIONS FOR THE ERROR (SUBJECT NOUN, VERB, OBJECT NOUN), TWO PARTS OF SPEECH FOR INTRODUCTION OF THE ERROR (ADJECTIVE, ADVERB), AND PREVIOUS KNOWLEDGE OF THE CORRECTNESS OF THE SENTENCE TO BE VIEWED. IN GENERAL, THE EFFECT OF AN ERROR WAS TO INCREASE THE VIEWING TIME FOR THE SENTENCE WHILE KNOWLEDGE THAT A SENTENCE WOULD BE CORRECT OR INCORRECT HAD NO EFFECT ON SUBJECT PERFORMANCE. ERRORS OCCURRING IN THE MAIN VERB WERE MORE DISRUPTIVE THAN WERE ERRORS IN EITHER THE SUBJECT OR OBJECT POSITION. ADVERB ERRORS AND ADJECTIVE ERRORS HAD APPROXIMATELY THE SAME EFFECT ALTHOUGH THERE IS A COMPLICATED INTERACTION BETWEEN TYPE OF SENTENCE, FORM OF ERROR, AND POSITION OF ERROR. PERFORMANCE ON THE TWO SYNTACTIC TYPES OF SENTENCES DIFFERED WITH INTRODUCTION OF ERRORS. THIS ARTICLE APPEARS IN THE "JOURNAL OF VERBAL LEARNING AND VERBAL BEHAVIOR," VOLUME 6, NUMBER 5, OCTOBER 1967, PUBLISHED BY ACADEMIC PRESS, 111 FIFTH AVENUE, N.Y., N.Y. 10003. (JD)