Until now I thought it was handy to have a textbook. It gave me and the students a kind of road map for how to procede. It was easy to make tests for because I could more easily quantify and qualify what we had covered, and so could the students.
This semester I used another method of study in addition to a textbook. Now I am regretting that I required my students to buy the text. First, I started using a method that would get students communicating in English based on their own experiences. I found this idea on Duane Kindt's site on TIPS (Tools for Increasing Proficiency). His method included having students tell other students about themselves, starting from who they are as people and then progressing on to their families and lifestyles.
This process convinced me of some points:
a. Students do not naturally approach grammar as it is presented in textbooks in any skill area, listening, speaking, reading or writing. (Available texts don't approximate what my students want to say.)
b. Traditional testing patterns are not sufficient for helping students learn a language. (Classroom events should steer testing, not artificially imposed criteria.)
c. All class content can be generated from student communication.
d. The teacher's role, at least in the university setting where I find myself, is to introduce a method or methods for learning, to encourage and guide communication, to provide criteria for evaluation, to contextualize student needs and language requirements, and occasionally provide evidence based intercultural perspective.
Textbooks are not written to provide this kind of support.
There are limitations to this approach. It is time consuming, and necessitates a teacher be available to students. That means that this may not be useful for teachers who are paid to teach by the hour. Their jobs are to be with a class for a period, not to plan, or evaluate outside of the class. For those teachers texts are probably an essential resource.