INTRODUCTION to CHAPTER SEVEN
This chapter introduces the ideas of group selection and multi-level selection, and defines the interactor.
Thanks to the work of David Sloan Wilson, Elliott Sober and others, the concept of group selection has been rehabilitated in mainstream biology. To understand group selection it is vital to make a distinction between objects of selection (interactors) and outcomes of selection in a population (the pool of replicators). Group selection means that groups are objects of selection, and thus interactors. For this to happen, variation within groups must be less than variation between groups.
Genetic group selection refers to variation in regard to genes and changes in the gene pool of a population. Cultural group selection refers to variation in regard to cultural replicators (habits, customs, and routines) and outcomes in regard to their distribution.
In human evolution, genetic intermixing between groups may have diminished the effects of genetic group selection, whereas processes such as conformism may have created sufficient group homogeneity for cultural group selection to work.
Up to this point we draw on existing literature. From about page 103 we make a new contribution. We ask what are the features that make a selected group (in viable group selection) relatively coherent, cohesive and long-lasting. Such groups must be more than mere aggregates of individuals.
To this end we modify David Hull’s classic definition of an interactor. Our definition appears on pages 106-7. Among other things, it differentiates an interactor from its environment, and shows the dependence of component replicators on their interactor. In turn, replication depends on interaction between the interactor and its environment.
Finally we discuss how business organizations can be interactors, and how their component routines can replicate. Much replication in the social world is through diffusion, but in the case of firm spin-offs replication involves offspring replicators.
KEY QUESTIONS
1. What are the conditions for group selection to occur?
2. In what respects can groups be more than the sum of their members?
3. What are interactors?
4. What are possible mechanisms of replication in the social world?
Darwin's Conjecture 
Dear Geoff,
I am confused by the terminology used in par. 7.4 and, in particular, I think you need not to introduce the term ‘institution’ and to state that “all organizations are institutions”.
One could say that
1. Organizations are social groups (but not all social groups are organizations), and that
2. The “social structure” (p. 163) of organizations involve the points (a), (b) and (c) of page 170
without leaving aside anything relevant.
Best wishes,
Gerardo
PS
I do not agree that organizations are institutions (and I feel this is not an “untenable claim”, n. 8, p. 170). In Hodgson (2006c) Douglas North agreed that organizations have an internal articulation (in your words: “Organizations themselves had internal players and systems of rules”). But from this consideration does not follow that organizations ARE institutions; they are institutions + something else (individuals and – usually, but not necessarily – artefacts).
In more general terms one could say that social groups are individuals + “internal” institutions (i.e. systems of rules that structure the interaction between the individuals within the group)
Hi Gerardo,
Terms like organization and institution are defined in different ways, but a near-consensus exists that institutions are SYSTEMS OF RULES. Organizations are also systems of rules, so all organizations must be institutions (by this definition), but with additional characteristics. Some instititions, such as language and table manners, are not organizations. So one set (organizations) is a subset of the other (institutions). In my 2006 paper (to which you refer) I document that Douglass North agrees that organizations are institutions, contrary to widespread misinterpretation. But his writings are sometimes unclear on this point.
Groups are not necessarily organizations or institutions.
Perhaps you have a different definition of organization or institution? If so, what?
If institutions are defined as systems of rules, then organizations are not institutions.
Organizations are not only systems of rules, but systems of rules + individuals (i.e., members of the organization).
(Organizations are a subset of social groups and their systems of rules are a subset of institutions.)
Forgive me Gerardo, but I think you misunderstand what a “definition” means.
A definition cannot list everything that is vital for an entity. When a mammal is defined as “a species that suckles its young” it says nothing about warm blood, a skeletal frame, etc., all of which are important but not part of the definition. A definition is not a description. A definition lists the minimum characteristics that differentiate one type from another. When institutions are “defined” as “systems of rules” this does NOT mean that institutions are ONLY systems of rules, and nothing else.
Then, following your definition of “definition”, I would rather say that “organizations are social groups”.
The definition of definition is not mine – it is found in philosophy, dating back to Aristotle.
Your statement that “organizations are social groups” is true, but it is not a DEFINITION of an organization. It does not list the minimum characteristics that differentiate organizations from non-organizations. While “all organizations are social groups” is true, some social groups are not organizations.
Similarly, my statement that “organizations are institutions” is also true, but it is not a DEFINITION of an organization. It does not list the minimum characteristics that differentiate organizations from non-organizations. While “all organizations are institutions” is true, likewise some institutions are not organizations.
Our definition of an organization is in the text and the glossary.
So we go back to my first comment where I stated:
“One could say that
1. Organizations are social groups (but not all social groups are organizations), and that
2. The “social structure” (p. 163) of organizations involve the points (a), (b) and (c) of page 170
without leaving aside anything relevant.”
Gerardo:
Your point (1) is entirely valid. But it is not an adequate definition of an organization.
Your point (2) is unclear to me. All organizations are social structures but not all social structures are organizations. Points (a), (b) and (c) on page 170 are necessary to define and distinguish organizations (from other social structures and other social institutions).
Points (1) and (2) articulate the definition of an organization.
That is: “An organization is a special type of social group whose social structure (p. 163) involves (a), (b), (c) (p. 170)”.
This is just:
A) To stress that an organization is a group, i.e. a set of individuals (more exactly, a cohesive set of interacting individuals that features emergent properties, among which a social structure)
B) To show that the concept of “institution” is not needed to define an organization (thus avoiding any possible confusion)
In response to Gerardo:
His points (1) and (2) involve more general concepts such as “group” and “social structure”, which results in (1) and (2) together being inadequate as a definition of an organization, in the required terms of pointing to what is distinctive about organizations.
Groups and organizations (like most (or all?) structures) have emergent properties. That does not point to what is distinctive about organizations.
The word “institution” can be substituted, as in the following definition of an organization: “An organization is a special type of SYSTEM OF SOCIAL RULES involving (a) criteria to establish its boundaries and to distinguish its members from its non-members, (b) principles of sovereignty concerning who is in charge and (c) a structure of command and responsibility delineating roles within the organization.”
So I agree that strictly the word “institution” is not needed. The organization/institution confusion was almost entirely caused by North. He has since clarified that organizations are special institutions. Many other leading scholars agree.
Trying to go to the heart of the matter: I consider an organization as (a specific type) of collective agent, this is why I find it hard to consider an organization just as a (specific type of) system of rules. Systems of rules do not learn and act; they do not have intentions, goals; they do not compete or cooperate with other systems of rules; …
Gerardo:
Defining a mammal as “an animal that suckles its young” does not mean that we define a mammal “JUST AS a (specific type of)” animal. Also, the fact that some animals do not suckle their young does not imply that a mammal is a non-animal.
Of course, systems of rules do not necessarily have intentions. But when they are part of an intentional or purposeful system, then they are essential mechanisms for forming and acting upon intentions.
Our definition of an organization includes other essential features, including “membership … boundaries … sovereignty … and … command”. Especially in the light of these, organizations can be – and often are – collective entities with intentions and goals. But sometimes organizations can be riven by internal dispute, and/or have inconsistent goals. Although some people include collective intentionality in their definition of an organization, for the reason just stated (and Occam’s razor) we chose otherwise. But this choice is open to discussion and dispute.
As I stated commenting ch. 6, the ability of social interactors to manipulate their replicators and the environment is relevant in social evolution.
This is why I consider intentional action an essential feature of an organization.
From your definition follows that “(…) an organization has the capacity to be a collective actor” (n. 9, p. 171).
From my definition follows that an organization IS a collective actor.
Thank you for the stimulating discussion.