This piece works on two levels. First, Blumer is making some methodological statements about how to do sociology, second, he is making an argument for a particular type of social theory, which rests on the interpretive actions of 'selves'.
The methodological position of symbolic interactionism (SI) is "naturalistic". Science is a combination of exploration and inspection, which maps for SI onto depiction (presentation of a setting) and analysis (discussion of how the facts-on-the-ground) work together.
The key to a 'naturalistic' study is that it stays close to empirical
world, embedding investigators into the life-world of those being studied.
This is important, because of the different worlds formed by people.
To get to know a world, you have to get close to it. [JWM: Link this
idea with that presented by Simmel in The Stranger. If an
investigator is a stranger -- and it seems that if they ever planned on
leaving they would be -- then how possible is Blumer's proposal?]
There are 4 central concepts for SI studies:
Society as Symbolic Interaction
The substantive key is that we interpret each other's actions, not just react, based on the meanings embued in action. This implies that interaction is mediated by the interpretation of symbols.
Stimulus --> [interpretation] --> response.
3 key points:
These two factors combine to imply that behavior is not a result of pressures toward action (i.e. "social forces"), but instead arises through the interpretations of actors about things in their context.