Required reading:

1. early calls for ethnographies of schooling
   1. in the 1960s Phil Jackson made a call for such ethnographies after a long career in school psychology

      [As an educational psychologist] I began to realize that the instruments that I knew how to use best, paper-and-pencil tests of one kind or another, created artificial environments... They were like little cages in which people sat while the investigator poked at with questions... I began wondering what my usual subjects of investigation, students of all ages, looked like in the wild, ... as students whose natural habitat was a school with its corridors and classrooms. (Jackson 1968 [1990]: xi)

   2. in fact there already existed quite a number of significant school ethnographies (though they mostly focus on high school organization)
      1. Lynd, Robert & Lynd, Hellen  Middletown: A study in modern American culture (1929)
      2. Hollingshead, A. B.  Elmtown's Youth: The Impact of Social Class on Adolescents (1949)
      3. Henry, Jules  Culture against man (1963)
2. the convergence between anthropological linguistics (sociolinguistics) and anthropology of education (schooling, really):
   1. The confirmation by sociologists that there is a statistical relationship between initial class position of a child and adult class position after the passage through schooling
      1. early work on class
      2. Coleman "family background" as most predictive of school success (1966: 325)
      3. Bourdieu and reproduction ([1970] 1977)
      opens a new set of question: what is happening in school that produces such relationships between family background and school achievement, particularly give the historical attempt to have school precisely break "birth privilege." Sociological work demonstrating post-test inequity leaves the school as a kind of black box. sociological methods cannot tell us what actually is happening in schools.
   2. the impact of conversational analysis and ethnomethodology, as foundations for ways of directly observing what is happening in classrooms. After all, what do teachers and students do with each other: they talk, they make each other talk, they converse, and they they mete consequences. Whatever will be one's interpretation of the relationship between social differentiation, schooling, learning, and the reproduction of the differentiations, the mechanisms involve a moment where practical activity is mostly a matter of language (talking, writing, etc.).
   3. thus early works like those by
      1. Cazden et al. Functions of language in the classroom (1972)
      2. McDermott et al. (mostly published between 1977 and 1983)
      3. Erickson et al  (1982a and 1982b)
   4. and ongoing about many sub-issues made salient by the early work
      1. Mullooly on teacher authority and student challenge (2006)

"What time is it?"

3. Mehan and the structuring of interaction within "learning lessons"
   1. focusing of that part of the total life of teachers and children in school that are most marked for "teaching/learning": the testing sequences
      1. of course, this is but a moment in the total life of adult (teacher?) and child (student?) in school.
      2. and thus it is not clear when "learning" may be happening in classrooms, and who controls this learning (rather than the assessment of the learning.
   2. Structure: the room as set up; the people assembling in particular ways; the kind of talk (questioning/answering/assessing).
   3. Structuring: when looking at the testing sequences, Mehan emphasize fact that teachers & students are producing the scene together:
      1. this is the great advance in that it moves attention away from the child and on to the dance (interaction), its properties, and, in later works, on the properties of the more encompassing dance (interaction) within which a testing sequences makes sense and it is in fact required that it be publicly known that the testing has happened (even if it has not, quite)
      2. the roots in pragmatism, particularly G.H. Mead on the conversation of gestures;
         1. meaning is determined by the next gesture--probably not performed by the original actor, and in fact often not even performed by the original dancers but by a "third" (observer, evaluator).
      3. the roots in Garfinkel's ethnomethodology (as one of the inspiration for Mehan, McDermott, etc.) even as it continued developing
         1. "determination" here must be understood as a temporary, though "immortal" (existing before and after participation), process:
            1. temporary given that any further "third" can propose that what happened is something else than what the original actor(s) may have intended it to mean, or had agreed that it meant
            2. immortal given that all determinations of meaning ("this is what is meant") will constrain next speakers--even if they are most powerful. In other words, all "third" acts are bound to acknowledge the existence of the first two--if they want to be known as relevant to what originally happened, whatever the "third" may be attempting to determine it must have meant.
   4. modeling an interactional pattern to highlight its consequences for ordering polities of practice (i.e. reproduction of a "social structure" ensuring that some will not get as much resources/power than others)

   This contrasts with McDermott work. However, in McDermott's perspective, any consideration of the consequences depends on the power of the model

   5. No exploration of the practical consequences of the organization of this structure
      1. for the people in the group as group (e.g. ongoing work of maintaining the frame)
      2. for the individuals in the group as what happens while in the frame makes a difference in other frames (whether a student is shifted into another track; whether the teacher gets a headache, or abandon the profession)
      3. for the polity as a whole as it debates how to organize schooling
4. Methodological note
   1. finding a structural link within a set of repeated observations in arguably related settings.
      1. emphasize the need to argue that a bunch of observations not randomly chosen from a set of events actually belong to a set. This cannot quite be determined a priori
   .