I'm the product of state schools, so I'm comfortable with large lecture classes. I'm a good lecturer, too. I get a lot of student evaluations that say I'm witty and make "a boring subject more interesting." (gee, thanks.) Here at Tiny Liberal Arts College, though, lectures are looked down upon. Even in "large" (meaning 70-120 student) introductory courses, there are professors who never lecture, but teach entirely through readings, activities, and discussion.
I've not yet learned the knack. I do a lot more activities than is typical for a big state school (and always did, even when I was teaching at a big state school), but I really can't wrap my head around teaching an intro course through discussion alone. I've been assured by my colleagues that students learn the material much more thoroughly and remember it for longer if they "learn through doing". The problem, as they acknowledge, is that you cover a lot less material. I'm sure students would understand the principles of evolutionary theory much better if they discover them for themselves. At the same time, it took Darwin decades to publish on natural selection. Why have students re-invent the wheel?
OK, natural selection might be a bad example, since a strong understanding of Darwin's theories would be worth the extra time. But what about topics that are less conceptually fundamental, like the details of the Neolithic revolution in North China, or of variation in primate social organization? There is a real trade-off here: do we teach more facts, allowing students to become familiar with the breadth of anthropological material, or do we teach fewer topics but more thoroughly? Note that this is a wash for most students. The average student will learn less of more material in the first scenario, or learn more of less material in the second. The changes come on the ends. Poor students probably do better with more dynamic class structures, while strong, self-motivated students don't learn as much as they could.
My personal preference is to lecture a great deal in intro courses, with activities and discussions for the most important, foundational concepts. In upper-division courses, I've moved away from lectures. I find that I need a lot of structure to those courses to make sure the important topics are covered and the discussions and activities are useful. Free-form discussion on the reading just doesn't do it for me.
Here's one example of the type of structure I've imposed on an upper division course. The course is an over-view of the culture history of the region-that-shall-not-be-named.
The class is organized around weekly thematic topics (diet, agricultural systems, government, trade, religion, etc.). Each student in the class chooses a sub-region or cultural group at the beginning of the semester. Every week, they bring a single-paged summary of the theme as it applies to their sub-region. For example, if this was a North American archaeology class (it's not) they would bring and present information to the rest of the class about the agricultural techniques that have been documented for, say, the Hopewell, or the Missisippian, or the Ancestral Puebloan area. This forces students into research (although they're still using secondary summaries, not primary sources of data). It also gives every student something to say, and helps move discussion along.
If you have an interest in implementing this structure with undergraduates, I have a couple of suggestions: 1) the grading structure must punish students who don't show up. The first time I taught this class, participation per se didn't count (just the quality of the research), and I frequently had half the class absent. 2) walk students through the research process the first time or two, because students who have no clue what you want will not tell you until it's too late. 3) have exams, not just weekly or semester-long research papers, and include all of the students' presentations on the exam. In other words, all students have a vested interest in coming to class, taking notes on other students' presentations, asking clarifying questions, and pushing their colleagues to produce high-quality work.