Archer, M., 2000. For structure: its reality, properties and powers: A reply to Anthony King. The Sociological Review, 48(3), 464-472.
“As Bhaskar wrote, ‘the real problem appears to be not so much that of how one could give an individualistic explanation of social behaviour, but that of how one could ever give a non-social (ie, strictly individualistic) explanation of individual, at least characteristically human behaviour! For the predicates designating properties special to persons all pre-suppose a social context for other employment. A tribesman implies a tribe, the cashing of a cheque a banking system. Explanation, whether by subsummption under general laws, advertion to motives and rules, or redescription…alw3ays involves irreducibly social predicates’” (464).
Realist social theory uses analytical dualism and not philosophical dualism, as stated by King. It is analytical for very instrumental reasons. The interaction: Structural Conditioning -> Social Interaction -> Structural Elaboration never allow for structure and agency to interact simultaneously, but they do provide for analytical tools for exploring the behavior of the micro/macro.
Archer re-presents the three kinds of emergence: numerical emergence, relational emergence and bureaucratic emergence.
Numerical Emergence: It is not simply that individuals can change things if they want to and if they possess enough information, as stated by Watkins. It is also the result of the social structure surrounding them. For example, if a small part of a population is literate and tries to improve the literacy rates for others, it will take longer than if a larger part of the population was in their shoes.
Relational Emergence: uses the example of the division of labor and how collective productive activity was on aggregate more than individual action. First order emergence, more efficient pin making, leads to second order emergence, wealthier nations, leads to third order emergence, current advantages in education, for example.
Bureaucratic Emergence: the emergent nature of roles.
UPDATE: Distinguishing emergence from aggregation: "has the generative capacity to modify the powers of its constituents in fundamental ways and to exercise causal influences sui generis. This is the litmus test which differentiates between emergence on the one hand and aggregation and combination on the other" (466)