Re: Serious thoughts about objectivity

Michael Cahill (MCBlueline@AOL.COM)
Sun, 6 Oct 1996 15:36:48 -0400

John McCreery writes:

<< Think of the difference between "She said, 'Stalin believes in God'" and
"She said, looking pale and furious, 'Stalin believes in God'" and "She
said, with a small, wistful smile on her face, 'Stalin believes in God'" or
"'Stalin believes in God,' she said, barely suppressing a giggle." I
frequently find myself wondering how many "beliefs" described in
ethnography were originally statements spoken in faith, in anger, in

Brilliant insight........ (Stalin, indeed!) (Paralinguistics go here :-)

Andrew Petto replies:

<<The problem is -- and John's example shows this exquisitely -- that without
"paralinguistic" stuff we have no context for organizing the observations.
is the classical "signal-to-noise" problem. We don't know what is signal and
what is noise until we put it into a context.>>

Brilliant analogy.

First, I'll add a more mundane observation. In the very act of measuring
social phenomena, we may alter them. This is the anthropological equivalent
of the Heisenberg Uncertainty Principle. It's a notion that I thought less
about when I was in the field, but that I think more about now, for a variety
of reasons.

Second, while mistrusting ease of measurement for precisely the reasons JM
outlines, I want to try to make use of it here where I think its use is due.
When I spoke of "events observed" and "events composed," I was referring to
a distinction which is, I think, relatively easy to grasp. "Events composed"
are those which are fictional. The characters and the specific incidents (as
opposed to classes of incidents) are made up. They may be conjured in order
to tell a truth about the life of a community that no real character can
tell, but they are imaginary nonetheless.

Here's an example of what I mean. I did my fieldwork in two social worlds.
One was a tough, inner-city neighborhood. That's where I lived. The other
was a social agency charged with policing family life -- in my neighborhood
and across the rest of the county. That's where I worked. Over the years,
I have come to feel that while the agency helped the neighborhood in many
ways, it deformed it in others. For some reason, it's the deforming that I
think a lot about now. But how to convey the nature of the process? It
occurs in thousands of little ways and it's not always easy to see. I have
considered inventing characters -- a powerful commissioner with an unhealthy
interest in a local crack-addicted teenage girl, for example -- and a
dramatic event -- the commissioner's rape and murder of the girl, and the
attempted coverup that follows -- which would symbolize in a dramatic way
certain underlying realities about the abuse of power and the need for
inclusion and for love that exist within and between *both* of these social

But, clearly, this would be an "event composed." It might be "grounded in
ethnographic, anthropological facts," perhaps, as Myles Richarson suggests.
But it'd still be fiction. If I were to include it in my text, I might call
the product "narrative anthropology," but I should not call it "ethnography."

This may not clear up much, but it does, I hope, clear up something.

How to characterize incidents that occur in the field as being of one sort or
another based on the stated or implied motives of the actors, for example, or
on the outcomes those incidents produce, another alternative, or on our own
preconceived, ethnocentric notions, yet a third possibility, is a lively
question indeed. In fact, as has already been noted, how do we fully
distinguish the one from the other?

Got to run. I'll be back later.

Mike Cahill