Contribution
A conceptual map of the major-transitions research programme from the philosophy of science: what counts as a transition, whether a general theory of them is possible, how genic and hierarchical explanations relate, what the transitions do to the levels of selection debate, and whether the resulting hierarchy has a privileged level.
Key insight
The levels at which selection acts are not a fixed background. Every higher-level unit was once a collective of free-living units, and it holds together only if selection among collectives comes to dominate selection among its members. So the hierarchy itself is something evolution explains.
Assumptions
- A transition is best defined by Maynard Smith and Szathmáry’s second characterisation, the formation of a higher-level unit (p. 2).
- Explanations of evolutionary advantage are one kind among several; mechanism and the sequence of stages matter too (p. 4).
- Organisms are entities that exhibit adaptations, functional organization and a certain “autonomy” or “agency” (p. 7).
Key results
- Definition (p. 2) and the pared-down list of transitions it gives.
- Contingency (p. 3): transitions may be “rare singularities”, and an MET may be “qualitatively different” from the evolution of altruism.
- General principles (pp. 3–4): short-term advantage to form a collective; instability “owing to defectors pursuing their own interests”; kinship, policing and division of labour as aligning mechanisms; higher-level selection that must “trump” lower-level selection. Relevant, Okasha says, but possibly too coarse to explain any particular transition; “The jury is still out”.
- A caution (p. 4): after Rainey et al. (2014), “cheating” and “public good” have been misappropriated in studies of bacterial sociality, from an a priori commitment to a collective action problem in every transition.
- Genic vs hierarchical (pp. 4–5): compatible; but formal equivalence of multilevel and inclusive-fitness models rests on approximations and “is not the same as causal-explanatory equivalence”.
- Levels of selection (pp. 5–6): a transition creates the potential for selection at two levels (citing Okasha 2005); success needs mechanisms “such as policing, to regulate the selfish tendencies of the lower-level units”; hierarchical organization is “endogenized”.
- Privileged level (pp. 7–9): the highest occupied level is where organisms are; the part-whole question is left open between three defensible options.
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | A transition is best defined as the evolution of a higher-level unit from formerly free-living units | moderate | conceptual argument |
| C2 | For a transition to succeed, higher-level selection must trump lower-level selection, typically via policing or aligned interests | moderate | attributed to Buss and Michod and endorsed; no model here |
| C3 | General principles may explain transitions only coarsely | moderate | argued from the proto-cell example |
| C4 | Genic and hierarchical approaches are compatible, but formally equivalent models need not be causally equivalent | moderate | argued, citing model-equivalence results and their approximations |
| C5 | The transitions make the hierarchy an explanandum | strong | conceptual; follows from the definition |
| C6 | The highest occupied level is privileged as the level of organisms | weak | argued, with the alternatives left open |
Concepts
- major evolutionary transition (MET) — “the evolution of a higher-level biological unit out of formerly-free living units” (p. 2).
- fraternal and egalitarian transitions — after Queller: related lower-level units versus unrelated, possibly of different species (p. 3).
- endogenizing the hierarchy — making hierarchical organization part of what evolutionary theory explains, not part of its background (p. 6).
- organism — Queller and Strassmann’s “largest unit of near-unanimous design” (p. 8).
Connections
It cites Okasha (2005), LIT-721 (unread), for the two-level view and for the point about the hierarchy’s origin. It engages Buss, Michod, Queller, Bourke and West et al. The levels-of-selection entry the record holds (LIT-160) is the traditional debate this paper says the transitions reframe; the record’s note on it (NOTE-184) reports that entry as critical of Okasha’s earlier (2006) treatment of the levels. This paper does not use the MLS1/MLS2 terms that criticism concerns.
Aktipis et al. (LIT-740) is a worked case of the general principle at one transition, multicellularity: cell-level fitness suppressed for organism-level fitness, with cancer as the failure. Okasha’s caution applies to it. Aktipis et al. define cheating functionally rather than as a strategy, and that answers part of the caution. They do not establish a fitness conflict for each growth they report.
Bearing on the record
Source of THEORY-141. For the owner’s essay:
- §6.3 (capture from below). A collective “would not necessarily be stable owing to defectors pursuing their own interests”, and succeeds only if higher-level selection comes to trump lower-level selection. That is the biological form of the essay’s first mode of cross-scale conflict, with policing as one remedy.
- §6.2 (constitutive dependency). The record reads the endogenized hierarchy as a parallel: the level an organization’s members depend on is itself produced, not given. Okasha does not discuss organizations.
- A caution for the essay. Calling a member’s conduct “capture” or “cheating” presupposes a collective action problem; Okasha’s warning against assuming one applies to organizations too.
Limitations
A review with no model or data, and “inevitably provisional” by its own account. It is not the 2005 argument, and the record cannot say how far the 2005 paper differs.
Open questions
Whether a general theory of transitions can be specific enough to explain any one of them, which Okasha leaves open; how to treat organisms within organisms.