A study examined the effects on 80 first graders' reading achievement when direct instructional phonics is incorporated as a supplement to a whole language approach. Two first-grade classrooms used the Houghton-Mifflin Integrated Literature Program, and two other classrooms supplemented the program with direct phonics instruction. Subjects completed a school readiness inventory and word recognition and fluency pretests, and were given posttests on word recognition, fluency, and the Houghton-Mifflin Student Progress Survey. Results indicated that the direct instruction group: (1) did significantly poorer in the school readiness inventory; yet (2) scored as well on the word recognition and fluency posttests as the other group; and (3) scored higher on the student survey. A second study of teaching and learning of reading in seven upper elementary reading resource rooms was conducted using the premises of grounded theory research. One finding of the study is relevant to a possible explanation of teachers' willingness to consider new instructional ideas: an underlying routine seemed to bind together the more obvious routines into stable interlocking networks. Traditional teachers who resist a whole language orientation often raise concerns associated with routines. In order "to do" whole language, a traditional teacher must relinquish the notion that a transition to a whole language orientation simply means a change of materials and the inclusion of a few new procedures. For major changes to occur efficiently teachers need to appreciate that change in one routine may strongly impact others. (RS)