This is the eigthth in a series of notes to fifteen lectures for my class ITSF5001: Ethnography and Participant Observation.

    Observation supplies fact. Induction ascends from fact to law. Deduction, applying the pure logic of mathematics, reverses the process and descends from law to fact. The facts of observation are liable to the uncertainties and inaccuracies of the human senses; and the first induction of law are rough approximations to the truth. The law is freed from the defects of observation and converted by the speculations of the geometer into exact form. But it has ceased to be be pure induction, and has become ideal hypothesis. Deductions are made from its syllogistic precision, and consequent facts are logically evolved without immediate reference to the actual events of nature. If the results of computation coincide, not merely qualitatively but quantitatively, with observation, the law is established as a reality, and is restored to the domain of induction. (Benjamin Peirce, Ideality in the Physical Sciences 1881: 163-4)

  1. The scientific cycle:

    some anthropologists and many others outside the social sciences reject the postulate that there can be any science of humanity, that all statements about human beings are, at best, interpretations, and can only be evaluated as any other artistic production can be evaluated, through "criticism"

    1. as ideal-type (Weber 1949): From observations to analytic statements and back
      1. the hypothetico-deductive cycle: ("theories," "laws," "rules), deduction from statements about verification or the recognition of limitations in the statement, leading to further observations and inductive sharpening founded on a better discursive representation of some aspect of experience for some purpose. (predominently used in sociology and other "data-driven" disciplines). (aka "quantitative" research; "empirical";).
      2. the "grounded theory" cycle: observations -> substantive theories -> formal theories to be tested through hypothetico-deduction (a grounded theory remains to be proven)
      3. the structural cycle: abstracting to statements of properties with predictive power for the shape of the future (predominently used in linguistics and related fields such as communication, conversational analysis, and sometimes anthropology).
    2. as model (at it will be drawn from studies of the production of new knowledge and its acceptance as new knowledge)
  2. The limit of ideal-types and models, including Glaser and Strauss's:
    1. they are "reconstructions" of the process (Kaplan 1964: 8) somewhat useful in epistemology and in ... teaching classes about methodology. They are presented in a linear fashion where:
      1. proposal (deduction)
      2. is followed by data collection (observation)
      3. is followed by analysis (induction)
      4. is followed by write-up (and the elaboration of a "grounded theory" to be tested). 
    2. This reconstruction must precisely not be taken as a description of the process itself which, to be successful must an iterative (circular) process in which the ethnographer is continually involved in the various types of activity into which the process can be analyzed. (proposing, observing, analyzing, writing).
  3. Which is why it is mostly not a good idea to delegate any part of the process to someone else: the ethnographer is involved in all the activities. 
  4. Thus the importance of what Glaser and Strauss call "constant analysis" which involves different steps at different stages in the research:
    1. "notes to yourself": early writing about your observations (etc.) to organize your first (second...) impressions, criticize what you have been doing
    2. possibly redirect (that is re-propose) your investigation both (theoretical sampling G&S: 45ff)
      1. on issues of sampling both at the level of
        1. persons
        2. settings (constant comparison)
      2. on issues of what to privilege in your observations
    3. comparison of observations and cases (G&S 101ff)
  5. The process of analytic abstraction: formal categories and theories vs. patterns and models

    "Written down" (trans-scripted) observations, by themselves, have little value unless re-written one more time within a theoretical framework in which they become part of the argument that is being made and that they justify. 

    1. Abstracting categories and theories: a process that, in Glaser & Strauss, leads from a "substantive" to a "formal" category (G&S 42).  These categories and related hypotheses are generated inductive but, in their further use for comparison and theorizing they belongs to a world of ideal-types (Weber 1949) (e.g. authority, social value, etc.) that allows the linkage of the observation to the general theories of society that they favor.  Thus they present their work in medical sociology as rediscovering ("grounding") some classic hypotheses about the relationship of class (income, education, etc.) to treatment by professionals and, in the process, arguably discovering new relationships (e.g. age).  Thus the importance of
      1. categories
      2. theories
  6. Abstracting properties for further manipulation:

    a different process of deconstruction identifying parts and their properties which, together make an object (e.g. a classroom) for the participants. an object that they themselves construct (constitute, institute) through their actions.  The goal of the analysis is a model (Lévi-Strauss 1962) highlighting parts and properties for particular purposes.

    1. One example is actually Glaser & Strauss own analysis of the social process of dying (1965)
    2. One classic example of such modeling in anthropology of education is Mehan analysis of the basic classroom "lesson" structure (1979: 52-54)
  7. In both cases, theory in the broadest sense is central. 

    Without a theory there is no way to differentiate the wheat from the chaff.  Conversely, the research should be a contribution to theory in a slightly narrower sense, pointing out what earlier statements did not re-present for the field within which the research places itself.