Publishers of basal reading series claim that beginning reading materials must contain primarily high frequency words and that new words must be introduced gradually. Inherent in their argument is the position that reading is based on the student's ability to recognize words and that short, phonetically regular words are easier to read. To examine this position, a study compared first-grade students' reading of a traditional basal story from a skills type basal series to their reading of student-dictated stories. The Spache Readability Formula was used to determine the readability levels of all stories. The Goodman and Burke Miscue analysis procedure was used to analyze oral reading performance, and story retellings were used to assess comprehension. Results showed that the basal story miscues at the sentence level were semantically and syntactically acceptable, but were coupled with meaning change. The interrelatedness of syntax, semantics, and meaning change on the story level was ignored by students as a basis for correcting miscues to maintain meaning. On the other hand, student-dictated story miscues produced interrelationships that allowed little meaning change and no loss in comprehension. Although they were confronting a more sophisticated vocabulary and more complicated sentence structure, the children used more efficient strategies when reading the dictated stories. (HOD)