Lecture Outline: The social network perspective I and II

• Introduction: Roots of the social network perspective
  • British family studies and anthropology:  Bott's different social networks by class and gender
  • Mathematics and mathematical sociology; Duncan Watts an example of a math type
  • Sociometry: small group and community studies - Scott's "Harvard synthesis"
    
• What makes a network a social network?
  • A network is a mathematical formalism of nodes and edges
  • Nodes are social actors - individuals and bounded constructs
  • Edges are social relations: culturally mediated interactional exchanges
  • Actors have "attributes" which are often group memberships (e.g. gender)
    
• Watts' Chapter 1:  The science of networks
  • Graph theory has been here since 1736 - so what's new?
  • Dynamics on networks and dynamics of networks
  • The idea of "emergence": complexity and self-organization - "sychronization"
  • "Connection topology"
  • Random graphs, clustering and "small-worlds"
    
• What is distinctive about the network approach to the social?
  • Multi-level embeddedness, rather than macro or micro reductionism, holistic
  • Behavioral and empiricist bias: focus on observable behavior of individuals, not variables
  • Focus on relations, not attributes: the anti-categorical imperative
  • Focus on network positions and roles
  • Focus on "partitions" of group membership

   

• Social networks and the main sociological traditions
  • Interactionism and symbolic interaction
  • Rational choice and exchange theory
  • Conflict theory and Marxism
  • Structural-Functionalist systems theory
    
• Watts' Chapter 2:  The Origins of a "New" Science
  • Traditional social network analysis is comparative statics
  • New work in networks focuses on theories of causes - dynamics
  • Math and physical science developments from random graphs
  • Two key social science contributions: why social networks aren't random
    • "Homophily" -- groups and culture matter (reciplrocity)
    • "Clustering" (or triadic closure, or transitivity biased nets)

Review Questions

1. The contemporary field of social network studies has diverse roots in anthropology, mathematics, and community studies. Can you describe the main interests of each of these "pre-networks" fields of study?

2. Scott argues that there was a Harvard "synthesis" that brought these elements together. What does he mean?

3. Social networks study relations among "social" actors. What do we mean when we say an actor is a "social" actor?

4. Social network analysis studies social relations. What do we mean by a "social" relation?

5.  What is a "random graph?"  a "clustered graph?"  

6.  What's the difference between "dynamics on a network" and "dynamics of a network?"

7.  Explain the idea of "emergence" and "self-organization."  Provide some sociological examples.

8.  What does the term "connection topology" mean?

9. What is the "anti-categorical imperative?" Why do some argue that this imperative makes social network analysis different from conventional sociology?

10. What does it mean to say that an actor is "embedded" in a network of relations, and that one network can be embedded in another? Why is the notion of "embedding" and multiple levels of analysis so important to the social network perspective?

11. What is macro reductionism? What is micro reductionism? How does the social network perspective seek to avoid reductionism?

12. Social network analysis is often said to be a "perspective" rather than a theory. That is, it can be used to interpret, and to study the ideas of the major theory groups. Review how network theory fits with interactionism, systems theory, conflict theory, and rational-choice/exchange theory.

13.  Explain the difference between "comparative statics" and "dynamics."

14.  What is the "largest connected component" of a graph, and why does it matter?

15.  Explain the ideas of "critical point" and "phase transition."  Give an example from physical science, give and example from social science to illustrate the concepts.

16.  What is a "random graph?"  How does a "biased random graph" differ?

17.  What is "homophily?"  What is it's significance in sociological network dynamics?

18.  Explain the concepts of "transitivity" and "triadic closure."  How do these relate to the idea of "clustering" in social networks?

Application Questions

1. Without invading anyone's privacy, or intruding, observe a small group of people interacting in a public place (like a cafeteria, a mall, etc. Notice which actors initiate interaction with which other actors, and whether these initiatives are reciprocated or not. Is everyone in the group equally likely to interact with everyone else? Are some persons more likely to initiate interaction than others?

2. Think about our class as a population of actors connected in pattern of regular and repeated interaction. Suppose you did not know that there were such social roles as "student" and "teacher," and could only observe interaction patterns. Do you think that you could conclude that there were distinctive differences among the actors, based on observing patterns of interaction? What might you notice that distinguishes "teachers" from "students?" How is this exercise an illustration of the "anti-categorical imperative?"

3. Think about our class again. A network analyst might insist that the interaction patterns in our class was "embedded" in a larger set of network relations. How would you describe this pattern of "embedding?" In what ways do the larger structures in which our interaction is embedded offer "opportunities" and "constraints" on what happens in the classroom?

4. Is the Internet a "social network?" who and/or what are the nodes? what are the relations? If we could study interaction on the internet, would it be "random" or would it display repeated and differentiated patterns?