The theory of evolution by natural selection is, perhaps, the crowningintellectual achievement of the biological sciences. There isconsiderable debate, though, about which entities are selected in anevolutionary process. This article aims to clarify these debates byidentifying four distinct, though often confused, theoretical andempirical research questions, as well as two schools of multilevelgenetics; the debates themselves are clarified by highlighting whichof these four research questions, or their combinations, are centralin each debate.

• 1. Introduction
• 2. Four Basic Research Questions
  • 2.1 The Interactor Research Question
  • 2.2 The Replicator/Reproducer Research Question
  • 2.3 The Manifestor-of-Adaptation Research Question
  • 2.4 The Beneficiary Research Question
  • 2.5 Summary
• 3. Two Traditions of Research into Multilevel Understanding of Units/Levels of Selection
  • 3.1 Evolutionary Change School of Multilevel Evolution
  • 3.2 Adaptationist/Kin Selection School of Multilevel Evolution
  • 3.3 Relationships Between Adaptationist/KS and Evolutionary Change Schools of Multilevel Evolution
• 4. The Anatomy of the Debates
  • 4.1 Group Selection
  • 4.2 Species Selection
  • 4.3 Genic Selectionism
  • 4.4 Genic Pluralism
  • 4.5 Units of Evolutionary Transition in Individuality
  • 4.6 Holobionts, and Demibionts
• 5. Conclusion
• Bibliography
• Academic Tools
• Other Internet Resources
• Related Entries

1. Introduction

When we think of evolutionary theory and natural selection, we usuallythink of, say, a herd of deer, in which some deer are faster thanothers escaping their predators. These swifter deer will, all thingsbeing equal, (on average) leave more offspring, and these offspringwill tend to be swifter than other deer. Thus, we get a change in theaverage swiftness of deer over evolutionary time. In a case like this,the “unit of selection”, often called the“target” of selection, is (at least) the singlemacro-organism, the individual deer, and the property being selected,swiftness, also lies at the organismic level; it is exhibited by thewhole deer, not by either parts of deer, such as cells, or groups ofdeer, such as herds. But there are other levels of biologicalorganization that have been proposed to be units or targets ofselection—levels at which evolution by selection may act toincrease the population’s distribution of a given property atthat level of biological organization.

But for over forty years, some participants in the “units ofselection” debates have argued that more than one issue is at stake.^[1] The notions of “replicator” and “vehicle”were introduced, to stand for different roles in the evolutionaryprocess (Dawkins 1976, 1982a,b). The individual deer, above, would becalled the “vehicles” and their genes that make them tendto be swifter would be called the “replicators”. The“genic selection” argument proceeded to assert that theunits of selection debates should not be about vehicles, as theyformerly had been, since Darwin, but about replicators. It was thenasserted that the “replicator” subsumes two distinctfunctional roles, which can be broken up into “replicator”and “interactor”:

Dawkins…has replicators interacting with their environment intwo ways—to produce copies of themselves and to influence theirown survival and the survival of their copies. (Hull 1980: 318)

The new view would call the individual deer, “interactors”(Hull 1980).^[2] However, one philosopher found that the two-partreplicator-interactor distinction was still inadequate for addressingkey controversies about units, specifically debates about speciesselection and group selection, (as well as holobionts, later), and sointroduced two additional distinct research questions that are alsopursued under the rubric of “units or levels ofselection”: In a given selection process, “what entityacts as themanifestor of adaptation?”, and “Whatentity is thebeneficiary?” (Lloyd 1988[1994]).

The purpose of this entry is to delineate further the four quitedistinct research questions pursued under the rubric of “theunits and levels of selection”.^[3] (see the entry onbiological individuals) Insection 2, four distinct research questions are introduced.Section 3 briefly describes a contrast in theoretical and modeling schools thathas affected these debates very profoundly, andsection 4 returns to the sites of several very confused and confusing,occasionally heated debates and claims about “the” unit ofselection. Several leading positions on the issues are analyzedutilizing the proposed taxonomy (or “anatomy”) ofdistinct questions (Lloyd 1992; 2001; Griesemer 2005) while brieflycomparing this analysis to others. The identification and status ofthese various units of selection, under these four meanings, is stilla very active area of disagreement and discussion in both biology andphilosophy (e.g., Okasha 2022; Suárez & Triviño2020; Suárez & Lloyd 2023).

When Damuth and Heisler introduced their statistical tool to helpanalyze units of selection in experiments, they wrote:

We think that failure to distinguish the various alternative goals ofinvestigations of multilevel selection has both generated needlesscontroversy biological subdisciplines and has impeded unification ofthe multilevel selection traditions.^[4] (1988: 409)

Our analysis in terms of the different “schools”introduced insection 3 is meant to aid such unification and mutual understanding. Thisentry’s analysis is not meant to resolve any of the conflictsabout which research questions are most worth pursuing; moreover,there is no attempt to decide which of the questions or combinationsof questions discussed ought to be considered “the” unitsof selection question in every context.

2. Four Basic Research Questions

Four basic research questions (represented as “RQs” often,throughout) are delineated as distinct and separable. These RQs may beapproached differently in the two schools or branches of multilevelselection evolutionary theory, introduced insection 3. As shown inSection 4, these questions are often used in combination to represent“the” units of selection problem. In this Section, webegin by clarifying terms. (See the entry onbiological individuals.)

The termreplicator, originally introduced in the 1970s as ageneralization of thegene, but since modified byphilosophers, is used to refer to any entity of which copies are made;it plays a functional role in the selection process (Dawkins 1976,1982a,b; Hull 1980; Brandon 1982; see§2.2).^[5]

The original “replicator” was introduced along with theterm “vehicle”, defined as

any relatively discrete entity…which houses replicators, andwhich can be regarded as a machine programmed to preserve andpropagate the replicators that ride inside it. (Dawkins 1982b:295)

On this view, most replicators’ phenotypic effects arerepresented in vehicles, which are themselves the proximate targets ofnatural selection (Dawkins 1982a: 62).

In the introduction of the terminteractor in place ofvehicle, it was observed that the previous theory hasreplicators interacting with their environments in two distinct ways:they produce copies of themselves, and they may influence their ownsurvival and the survival of their copies through the production ofsecondary products that ultimately may have phenotypic expression(Hull 1980). The term “interactor” was suggested for theentities that function in this second process. Aninteractordenotes that entity which interacts, as a cohesive whole, directlywith its environment in such a way that replication isdifferential—in other words, an entity on which selectionacts directly (Hull 1980: 318). The process of evolution by naturalselection is

a process in which the differential extinction and proliferation ofinteractors cause the differential perpetuation of the replicatorsthat produced them.^[6] (Hull 1980: 318; see Brandon 1982: 317–318)

In the late 1990s and early 2000s, however, theoreticians,developmental evolutionary biologists, and philosophers thinking aboutthe origins of life expanded the application of the notion ofreplication described by Dawkins and Hull, and introduced the conceptof “reproducer” as a generalization of the notion of“replicator” (Griesemer 1998, 2000a,b,c; see discussion inhis 2005, 2016^[7]).

In sum, “the units of selection question” is not a singleRQ at all, but rather, several distinct, functional, interrelatedRQs.

2.1 The Interactor Research Question

The interactor RQ, as introduced above, is: “what organizationallevel of units or entities are being actively, directly, selected in aprocess of natural selection?” Or, more traditionally:Aninteractor is an entity which interacts, as a cohesive whole, directlywith its environment in such a way that replication isdifferential (Hull 1980: 318). As such, this question appears inthe oldest forms of the units of selection debates (Darwin 1859;Haldane 1932a,b; S. Wright *1931, *1945). In an early, veryinfluential review on “units of selection”, the purpose ofthe article was stated: “to contrast the levels of selection,especially as regards their efficiency as causers of evolutionarychange” (Lewontin 1970: 7). Similarly, others from theEvolutionary Change school (see§3), assumed that a unit of selection is something that “responds toselective forces as a unit—whether or not this corresponds to aspatially localized deme, family, or population” (Slobodkin& Rapoport 1974: 184).

Questions about interactors focus on the description of the selectionprocess itself, that is, on the interaction between an entity, thatentity’s traits, and its environment; these interactions arestudied through components of the entity’s fitness and itstraits or properties, and on how this interaction may produceevolutionary change.They do not ordinarily focus on the adaptiveoutcome of this process. (Note that this applies especially tothe school of evolution calledEvolutionary Change,§3).^[8] The interaction between some interactor at a certain level and itsenvironment is assumed to be mediated by “traits” thataffect the interactor’s expected survival and reproductivesuccess, i.e., via some aspect of its fitness components of thosetraits.

Here, the possible and responsive answers to RQs about the interactorinclude entities possibly at any level of biological organization,including a lineage, a cluster of interacting species—such as asymbiont or holobiont (§4.6), a species, a group, a kin-group, an organism, a gamete, a chromosome,or an allele (Lewontin 1962, 1970; Brandon 1988; Lloyd 2017). Someportion (or ‘component’) of the variance in the expectedreproductive success of the interactor is commonly expressed in termsof the value of the trait and genotypic or genic (or otherreproducer-level) fitness parameters, that is, in terms of the fitnessof combinations of replicators or reproducers. Several technicalmethods are available for expressing the relevant formal relationshipsbetween interactor traits and (genotypic or reproducer) components oftrait fitness, including analysis of partial regression, as well asvariances, and covariances, discussed further in 4.1.

As explained in the notes, partial regression analysis, i.e.,contextual analysis, is most commonly applied today in studies ofinteractors (that is, in Evolutionary Change school investigations ofplain interactors, without further investigation of the unit’sengineering adaptations per se, which is standard).^[9] The early introducers of the contextual analysis tools remarked onthe valuable potential of contextual analysis for

…empirically resolving questions about the importanceof group selection in nature. The debate about group selection has[thus far in 1987] relied primarily ontheoretical arguments aboutthe causal efficacy of particular group or individual charactersthat appeared tobe correlated with individual fitness. Aftertwo decades of speculation, contextual analysis could provide a meansfor discussing the group-selection issue on a foundation of empiricalresearch. (Heisler & Damuth 1987: 597).

This was prescient: empirical research from both laboratory, field,and breeding experiments soon proved that group multilevel selectionof interactors was extremely effective at producing evolutionarychange (see summaries in: Wade & Breden 1981; Goodnight &Stevens 1997; Muir 2005; Wade, Bijma, et al. 2010; Wade 2016;Goodnight 2015).

Nevertheless, it has sometimes been objected in the philosophicalliterature debating units of selection that the notion of interactoristoo vague to be useful or applicable (both mistaken andsurprising, given the mathematical precision of the statisticalmethods used for definition and identification); or that theparticipating entity is not “cohesive” enough to play thisfunctional role in the selection process as assigned (e.g., Booth2014; Bourrat 2019; Okasha 2022; see§4.5,§4.6).^[10]

But note that the researchers using the multilevel models cited abovediscuss and model the functional role of interactor with verysignificant precision, as an application of Hull’s originalconception in terms ofcohesiveness in interaction with theirenvironments, represented as “emergent” fitnesses inrelation to multilevel traits and environments that are traceablethrough statistical tools such as a contextual analysis approach,depending on the selective context (Lloyd 1988 [1994]; Brandon 1990,contrast Sober & Wilson 1998; Okasha 2006; see work of bothAdaptationist/KS and Evolutionary Change schools, especiallyGoodnight, Bijma, Muir, Damuth and Heisler, and Wade; discussedfurther in§§4.1–4.6).

It is important to see that, in the midst of deciding among thevarious methods for representing selection processes, these choiceshave consequences for the empirical adequacy of the selection models.^[11]

It is also extremely significant that the “interactor researchquestion” does not involve attributing engineering adaptations,or “benefits” in the adaptive sense, sometimes accumulatedthrough selection over time, to the interactors, or indeed, to anycandidate unit of selection. Attributing interaction at a particularlevel involvesonly the presence of an entity’s traitat that level with a special relation to genic, genotypic, replicator,or reproducer expected success that is not reducible to interactionsat a lower level.

As seen inSections 4.1–4.6, the most common error made in interpreting many of theinteractor-based evolutionary models and model-types is that thepresence of an interactor at a level is taken to imply that theinteractor isalso a manifestor of an adaptation at thatlevel. We will consider a couple of approaches to “units ofselection” inSection 4.1 Group Selection, and4.2 Species Selection, wherein a combination of interactor and manifestorof adaptation functional evolutionary roles are merged from the verystart. This conflation of functional roles has also led to confusionin the interpretations of evolutionary transitions (§4.5), as well as holobionts (§4.6).

2.2 The Replicator/Reproducer Research Question

The “replicator Research Question” originally concernedwhich organic entities met Dawkins’ definition besides the gene.The subsequent, more widely used, meaning in philosophy of the termwas: “an entity that passes on its structure directly inreplication” (Hull 1980: 318). The termreplicator isused in this latter sense, henceforth.

The issue corresponds to a long-standing debate in genetics about howlarge a fragment of a genome ought to count as areplicating/reproducing unit—something which can be treatedseparately in evolutionary theory as inheriting traits of interactors(Lewontin 1970; Hull 1980). This debate revolved around linkagedisequilibrium and led some biologists to advocate the use ofparameters of the entire haploid genome rather than of allele andgenotypic frequencies in genetical models, a view far ahead of itstime (Lewontin 1974).^[12][13]

The basic point is that with much linkage disequilibrium, individualgenes cannot be considered as separately-acting replicators becausethey do not behave as separate units during reproduction. Althoughthis debate remains pertinent to the choice of state space ofgenetical model (§§4.3–4.4), it has been eclipsed by concerns about interactors in evolutionarygenetics.

Significantly, recent developments in debates about units of selectionshow that there has also been a profound reconception of theevolution-by-selection process, which has rejected the original roleof replicator as misconceived, and too narrow in many circumstances.In its place the role of “reproducer” is proposed, whichfocuses on thematerial transference of genetic and othermatter, and itsdevelopment from generation to generation(Griesemer 1998, 2000a,b, 2003, 2005, 2014, 2016; see Forsdyke 2010;§3,§4.1,§4.5,§4.6). On this approach, thinking in terms ofreproducers incorporates development into heredity and theevolutionary process. It also incorporates both epigenetic and geneticinheritance within the same framework, and includes the traditionalreplicators as a subset (Griesemer 1998, 2000c, 2018). A more recentand influential conflicting characterization of“reproducer” disagrees about retroviral reproduction, andwhat counts as a salient material bond between generations (GodfreySmith 2009, 2012; Griesemer 2016, 2018).^[14]

Griesemer’s functional role of evolutionaryreproducer—expanding the replicator—can play a centralrole, along with a hierarchy of interactors, in work similarlyexpanding the RQs framework into the units of evolutionary transition(seesection 4.5 and the entries onbiological individuals,philosophy of macroevolution, andevolutionary game theory).^[15]

2.3 The Manifestor-of-Adaptation Research Question

At what level do adaptations occur? Or, “When a populationevolves by natural selection, what, if anything, is the entity thatdoes the adapting?” (Sober 1984: 204).

The presence of adaptations at a given level of entity is sometimestaken to be a requirement for something to be aunit ofselection, in addition to a usually-silent accompanyingrequirement that it also be an evolutionary interactor.^[16] Moreover, an adaptation at a level is one important sort of“benefit” for entities—whether organismic,familial, or group—resulting from natural selection of thoseentities at that level (G. C. Williams 1966).

Significantly, though, multilevel,Evolutionary Change schoolgeneticists argued that group selection for “groupadvantage” should be distinguished from the process of groupselectionper se, that is, the sole process of selection ofinteractors at the group or higher level (S. Wright 1980;§3,§4.1). In fact, the combination and blurring of the interactorresearch question with the question of what entity has adaptations hascreated a great deal of confusion in the units of selection debates ingeneral (see§§4.1–4.6.)

(The distinction between interactor and manifestor-of-adaptationfunctional roles has recently been more or less reintroduced by aphilosopher from the Adaptationist school [see§3] in work on “Type-1 agents” in evolution by selection(Okasha 2018)^[17].)

Some of this confusion between interactor and manifestor of adaptationis a result of a very important but neglected duality in the meaningof evolutionary “adaptation” (in spite of discussions inBrandon 1978, 1990; Burian 1983; Krimbas 1984; Sober 1984; Lewontin1978; Munson 1971).

Sometimes “adaptation” is taken—by both philosophersand biologists—to signify any trait at all that is a directresult or outcome of a selection process involving entities at thatlevel in the population. In this view, any trait that arises directlyfrom a selection process is claimed to be, by definition, anadaptation, and I call this the “selection-product”version of adaptation.^[18]

Sometimes, on the other hand, the term “adaptation” isreserved for traits that provide a “better fit” with theenvironment through accumulated build-up of modifications in phenotypethat intuitively satisfy some notion of “good design” or“improved engineering” that goes beyond the original rangeof variation in the population.^[19] I call this second meaning theengineering definition ofadaptation, which is distinct, and sometimes, in tension with theselection-product meaning of the term.

Consider the famous peppered moth case: natural selection is acting onthe coloring of the moths, and the population evolves over time, butno “engineering” adaptation emerges; that is, there hasbeen no “accumulated change in” phenotype that goes beyondthe original range of variation of the phenotype in the population.Rather, the proportion of dark moths simply increases over time,relative to the industrial environmental conditions, a clear case ofevolution by natural selection, on which a good fit to the environmentis reinforced. Note that the dark moths liewithin the range ofvariation of the ancestral population; they are simply morefrequent now, due to their superior fit with the changedenvironment.

The dark moths are a “selection-product” adaptation; therehas been no accumulated “engineering” of the form, norlong-term change in the formation or function of any phenotypic trait.Contrast this to the cases of Darwin’s finches, in whichdifferent species evolved distinct beak shapes specially adapted totheir diets of particular seeds (B. R. Grant & P. R. Grant 1989;P. R. Grant 1986). Natural selection here occurred against constantlychanging genetic and phenotypic backgrounds in which accumulatedselection processes had changed the shapes of the beaks, thusproducing “engineering” adaptations due to selection. Thefinches now possess newly shaped beaks that are new mechanismsbeyond the original range of variation in the ancestralpopulation; see Suárez & Triviño 2020 for cases ofholobionts as manifestors of adaptation).

Some evolutionary biologists have strongly advocated an engineeringdefinition of adaptation (e.g., G. C. Williams 1966). The basicalternative idea is that it is possible to have evolutionary changeresult from direct selection favoring a traitwithout havingto consider that changed trait as an (engineering) adaptation, becauseno accumulated phenotypic “design” change occurred; i.e.,the entity is merely an interactor in the evolutionary process.^[20]

It is important to see how this biologically-based definition ofengineering adaptation differs from the philosophical definitions of“biological function” usually on offer (see entries onteleological notions in biology andadaptationism); I note that the accumulated,engineering adaptationdescribed here fits both prongs of the usual philosophicalchoices.

In his review of “function”, Colin Allen offers us the twomain philosophical accounts of the concept usually used when definingevolutionary adaptation:

Etiological approaches to function look to a causal-historical processof selection; functions are identified with those past effects thatexplain the current presence of a thing by means of a historicalselection process (typically natural selection in the case ofbiological function).

Systems-analysis approaches invoke an ahistorical, engineering styleof analysis of a complex system into its components. Functions ofcomponents are identified with their causal contributions to broadercapacities of the system.^[21] (C. Allen 2002: 375).

Significantly, both the engineering adaptations and selection-productadaptations, introduced above, fall under Allen’s first,etiological approach to function; they both invoke the historicalselection process to explain the current presence of a trait. But theengineering adaptationsalso appeal to design analyses oftenidentified with the systems-analysis notion of function, as well.Thus, engineering adaptations—as the concept is used by leadingevolutionary biologists—appeal tobothetiological and systems-analysis approaches/definitions of function inthe philosophical senses just presented (Lloyd 2021).^[22]

Thus, we distinguish between selection and evolution of entitiesthrough emergent fitnesses in relation to the phenotypic trait in thecontext/environment [an interactor], and through emergent engineeringphenotypes acting as interactors [the manifestor of adaptation]. Oneway to put the point, is that the interactor concept is adaptation-neutral.^[23]

In sum, when asking whether a given level of entity possessesadaptations, it is necessary to state not only the level of selectionin question but also which notion ofadaptation—selection-product orengineering—is being used. In fact, this question oftendiffers between the two schools of multilevel selection outlined inSection 3.

2.4 The Beneficiary Research Question

Who benefits from a process of evolution by selection? There are twocommon interpretations: Who benefits in the long term from such aprocess? And who gets the benefit of possessing (engineering)adaptations as a consequence of an evolution-by-selection process? Aswe have addressed the latter question in§2.3, take the first, the issue of the “ultimatebeneficiary”.

There are also two obvious answers to this question, i.e., ways ofcharacterizing the long-term beneficiaries of theevolution-by-selection process. One might respond that the species orlineages are the ultimate beneficiaries of the process.

Alternatively, one might say that the lineages characterized on thereproducer/replicator level, that is, the replicating alleles,genomes, or reproducers, are the relevant long-term, or ultimate,beneficiaries. I have not located any authors holding the first view,but, for Dawkins, the latter interpretation is theprimaryfact about evolution (1976; Ågren 2021a). To arrive at thisconclusion, he adds the requirement of agency to the notion ofbeneficiary (Hampe & Morgan 1988). A beneficiary, by definition,does not simply passively accrue credit in the long term; it mustfunction as the initiator of a causal pathway (Dawkins 1982a,b;revived in Okasha 2018,compare manifestor). Under thisdefinition, the replicator as agent is causally responsible for allthe various effects that arise down the biochemical or phenotypicpathway, irrespective of which entities might reap the long-termrewards (Sapienza 2010).

In some recent follow-up work on agency from a philosopher, mentionedin§2.3, the evolutionary phenomenon focused on is agency of

type 1 [which] is a legitimate expression of adaptationism, but itrelies on a crucial presupposition. It presupposes that the entitythat is treated as an agent exhibits a“unity-of-purpose”,… [i.e.,] its evolved traitscontribute to a single overall goal. (Okasha 2018: 5)

Note that both meanings of beneficiary distinguished above aremobilized here: adaptationist goals and features of both traits andgenes.

This second and quite distinct, “ultimate”, version of thebeneficiary question can be intertwined with the notion of adaptation.The evolution-by-selection process may be said to“benefit” a particular level of interactor, throughproducing engineering adaptations at that level (e.g., G. C. Williams1966; Maynard Smith 1976; Eldredge 1985; Vrba 1984).

It is crucial to distinguish the research question concerning thelevel at which engineering adaptations evolve and accumulate from aselection process, (§2.3), from the question about the identity of the ultimate beneficiaries ofthat selection process in the Dawkins sense, as here (§2.4). One can think that interactors have adaptations without thinking thatthose interactors are the “ultimate beneficiaries” of theselection process.^[24]

2.5 Summary

Four distinct research questions have been isolated and identified, inthis Section, that have appeared under the rubric of“the units of selection question” in evolutionarytheory and practice: What is theinteractor? What is thereproducer, or more narrowly constrainedreplicator?What entitymanifests accumulated, engineering,adaptations resulting from evolution by selection at a levelof biological organization? And what is the ultimatebeneficiary? There is an important ambiguity in the meaningofadaptation; which meaning is in play has had significantconsequences for both the group selection and species selectiondebates (see§4.1,§4.2).

Below, the anatomy and applications of this collection of RQs alongwith their possible and responsive answers, all included under a“units of selection” framework, are analyzed and reviewed.Identifying which specific meaning(s) of the level and unit ofselection is being used, alone or in combination, in evolutionarydiscussions and models, both formal and informal, is a useful initialapproach for thinking about both theoretical and experimental researchinto natural selection processes. When applying this “logic ofresearch questions”^[25] analytical method like this—that is, clarifying the variety ofRQs in a controversy, and carefully specifying their possible andresponsive answers within their wider scientific (as well as socialand cultural, although those are not emphasized here) frameworks andcontexts. The method is conceived as a tool that integrates knowledgeand theory in evolution with the empirical, experimental, practiceside of the science. It simultaneously also frames the adaptationistresearch approach in contrast with other research approaches inevolutionary theory. This “anatomy” framework combinedwith the “logic of research questions” tool or analysis,is a clear way to understand how the expression “unit ofselection” is used in the many different research contexts inwhich it appears.

Ironically, these same three or four functional evolutionary roles,introduced, defined, and distinguished decades ago (see comprehensiveanalysis of leading approaches to units in Griesemer 2005^[26]), have since been reintroduced or resurrected by a number of differentauthors under different names and terminology, but in relation to thesame or similar debates about units (e.g., Bourrat 2016, 2019,2021a,b, 2022; Bourrat & Griffiths 2018; Clarke 2016; Stencel& Wloch-Salamon 2018; Godfrey-Smith 2009, 2016; Doolittle &Booth 2017; Doolittle & Inkpen 2018; Inkpen & Doolittle 2021;Lean et al. 2022; Birch 2020; Booth 2014; Okasha 2018, 2022; Gardner2015a,b,c). It is “ironic”, because these same neweranalyses either explicitly reject or neglect the full functional andformal role analysis—or some representation ofsuch—offered earlier.^[27] A couple of these re-introductions of needed distinctions arediscussed inSection 4 (e.g., Okasha 2022; Doolittle & Booth 2017 and Stencel &Wloch-Salamon 2018: Section 4.6; see Suárez & Lloyd 2023for further analyses).^[28]

Finally, the claim that two meanings of “unit ofselection” aredistinct does not imply that they cannotbe interrelated—that is not the claim here—or that oneliving being or part cannot simultaneously satisfy two or morefunctional evolutionary roles; in fact, we often see this (discussedfurther in§4).

In any case, commenting on the four-pronged “anatomy”analysis of the debates over units of selection, John Maynard Smithwrote inEvolution:

[Lloyd 2001] argues, correctly I believe, that much of the confusionhas arisen because the same terms have been used with differentmeanings by different authors … [but] I fear that theconfusions she mentions will not easily be ended. (Maynard Smith 2001:1497)

InSection 4, a dissection of this anatomy of research questions is used to clarifysome of the most visible positions in six debates: (4.1) group selection; (4.2) species selection; (§4.3) genic selection; (§4.4) genic pluralism; (4.5) units of evolutionary transitions in individuality (ETI) and (4.6) holobionts and demibionts. But first we must discuss an extremelyinfluential but usually unrecognized bifurcation in the theory andpractices of evolutionary multilevel theory inSection 3.

3. Two Traditions of Research into Multilevel Understanding of Units/Levels of Selection

There is a high-level analysis,^[29] that whereas kin selection research questions aim primarily atidentifying character states and genic/ kin selection processesthat maximize fitness, multilevel and group selection methods aimto understand theeffects of selection on interactor trait changesover time (Goodnight 2013, 2015; Wade 2016; Maynard Smith 1974,1976, 1987, 2001; Futuyma & Kirkpatrick 2017). Here, we introduceboth research approaches and priorities. But first, we need some basictools for our discussion.

As argued and illustrated by several philosophers many decades ago,the structure of evolutionary theory itself may be best understood asan interrelated family ofinformal and formal model-types ormodel outlines (Beatty 1980; Lloyd 1983 (proposing informalmodel-outlines or model-types of Darwinian theory;contraGodfrey Smith’s (2007: 731) rejection of the “semanticview” of scientific theories, explanation, and confirmation, onthe basis that it necessitated “mathematization”, which itdid not when Beatty and Lloyd used it to analyze Darwin’sinformal models), nor later in Lloyd’s 1988 informal models ofspecies selection, applied in Lloyd and Gould 1993 (see§4.2 below); 1984 (analysis of formal genetics models; Lloyd 1988[1994](arguing for seeing evolutionary biology in terms of families ofinformal models of organismic, species selection, and group selectionmodels, as well as formal models of population genetics andquantitative genetics of organismic and higher level selection, usingthe semantic view); Griesemer 1984, (informal and formal evolutionarymodel-types, using semantic view and their“presentations”, congruent with Godfrey Smith’slater model “construals”, 2007); Griesemer 1991, 2013;Thompson 1989; Ketcham 2018; and biologists Wade & Kalisz 1990;Wagner et al. 2014). Philosophers Downes 1992; and Weisberg 2012, alsoendorsed a formal and informal model-based approach to evolutionarybiology, under different names. One key agreement amongallof the above past and present participants in model-based analyses ofevolutionary theory, and many more, is that it can be understood as aninterrelated group of formal and informal evolutionary models andmodel-types, where a model-outline ormodel-type isdefined as a basic structure whose shared formal or informalparameters and variables may not all have been assigned values in agiven application.^[30]

This is all significant for our analysis of units of selection becauseit is needed to grasp that the basic model-types used by the twodifferent schools we are about to examine are fundamentally useddifferently in significant ways, and are usually used to answerdifferent research questions (RQs) in evolutionary research.

One more key distinction must be emphasized before proceeding. Asdiscussed in§2.1,§§4.1–4.6,interactors are characterized primarily bytheir cohesive responses/interactions with (i.e., when they“interact as a whole with”) their environments, bothbiotic and abiotic (see§2.1,§3,§§4.1–4.6).^[31] As we will see below repeatedly, they are very significantlydistinguished frommanifestors-of-adaptation (§2.3,§3,§§4.1–4.6; see McKenna et al. 2021; Lloyd 1988 [1994]; G.C. Williams 1992).Specifically, any “cohesiveness” in theactivities/traits the interactors may exhibit while in these holisticinteractions, doesnot necessarily yield“emergent” engineering adaptations, more widely understoodby philosophers as etiological and systems “functions”(see definitions in§2.3, also see§3,§4.1,§4.2). Also please note that Lewontin’s original 3-part“recipe” for a unit of selection to evolve, upon which somuch discussion is based, does not mention or include engineeringadaptation concepts or results as central. It involves onlyinteractors, as written, even while many, if not most of somesub-communities of,evolutionary researchers are motivated to seek outadaptations of distinct units of selection.

Instead, a standard for a manifestor of adaptation must also beestablished and explored separately, usually through game theory orother forms of exploring optimality, a notion from engineering anddesign. The genetics models are then produced separately, needed toproduce or maintain the optima, as we shall see below.

3.1 Evolutionary Change School of Multilevel Evolution

Multilevel selection methods for model-building and detection ofinteractors, such as contextual analysis (§2.1,§4.1,§4.2) and multilevel modeling,^[32] arising out of the quantitative genetics tradition, are used todescribe the evolutionary processes acting on a population in itscurrent state of traits or fitness; the focus is on evolutionaryprocesses,in contrast to searching for or modeling whichlevel entity might be a manifestor of adaptation.^[33] This is often called the “Evolutionary Change” or“Wrightian” school, because of its interest in SewallWright’s work on evolution in structured populations, and itsemphasis on the effects of epistasis and gene interaction (seebelow).

Wright found that when populations are subdivided into smaller groupsor demes, as many actual biological populations are, this populationstructure challenges the Fisherian modeling assumption of panmixia(freely mixed, random mating) in the metapopulation (collection of alldemes). For example, the evolutionary response to selection on a givencombination of geneswithin a deme depends on the overallmigration patterns (e.g., Slatkin 1981b; Uyenoyama & Feldman 1980)and rates within the metapopulation, as well as on theinteractions of genes within the demes.

Griffing (1967) highlighted the social or competitiveeffectsorganisms have on each other, because when these effects have agenetic component, they provide an additional source of heritablegenetic variation within the population.

These effects and processes are referred to as “indirect geneticeffects” (IGEs) (Moore, Brodie, & Wolf 1997; Wolf, Brodie etal. 1998; Wolf, Brodie, & Wade 2000) or “associativeeffects” (Griffing 1967; Muir 2005). They have the potential toaccelerate, slow, or even reverse an evolutionary response to selection.^[34]

Moreover, traits of an individual’s social environment, such asthe aggressiveness of an ant or bee colony when foraging, can directlyinfluence behavior (Sih & Watters 2005; Westneat 2012), and imposemultilevel selection.^[35]

Hence, modern Evolutionary Change theorists tend toemphasizethe importance of epistasis and IGEs, which areexplicitly,deliberately, excluded in the Adaptationist methods and models,although they are essential to higher levels ofbothselection and engineering adaptation (Bijma 2010a,b, 2011,2014; Bijma, Hulst, & de Jong 2022; Breden & Wade 1991; Wade2016; Goodnight 2014 (see Other Internet Resources), 2015; contrast,e.g., Gardner & Welch 2011.^[36]

3.2 Adaptationist/Kin Selection School of Multilevel Evolution

Kin selection (KS) and inclusive fitness genic model-types^[37] arose primarily out of game theory, evolutionary stable strategies(ESS), and classic Fisherian mass genetics, and are used to identifythe optimal solution[s] to particular adaptive questions. Thisresearch practice is thus usefully labeled the“Adaptationist” approach to multilevel evolution(Goodnight 2015), due to its focus on locating interactors that arealso engineering adaptations. The goals are to explain bothfunctional roles—interactor and manifestor ofadaptation—by representing them with KS/inclusive fitness genicmodels.

Any typical Adaptationist/kin selection/inclusive fitness (referred tooften below asAdaptationist/KS) school multilevel RQembodies, at best,two interdependent research questions:

First RQ: Identify an equilibrium or optimal trait at a given level oforganization which is an engineering adaptation (often done usingengineering-style optimality modeling, see Oster & Wilson 1978;Gardner & Welch 2011.^[38]) and which isunlikely to evolve by selection of thelower-level entity traits. The paradigms are “altruistic”or “self-sacrificing” traits at the lower level, loweringself-fitness, but which are “good for the group” orhigher-level fitness.

Second, co-ordinated RQ: Characterize a KS or inclusive fitness/genicselection process in which this engineering higher-level-optimaladaptationcould have evolved,against the pressuresof the lower-level entities’ selective processes.

For example, Gardner & Welch, operating within thisAdaptationist/KS school, claim that their simpler, genic-orientedmodels provide goodjustification for excluding thenon-additive genotypic effects we just discussed in the previoussection concerning the Evolutionary Change school—theverysame interactions that drive both upper-level selection, and anyadaptation that might occur, in the Evolutionary Change multilevelselection models—because such effects “show nocorrespondence with the optimization program” (2011: 1809),i.e., with the Adaptationist search for optimal gene’s-eyesolutions.

Thus, while they acknowledge that

neglect of these epistatic effects will lead to a less complete (andpotentiallyquite inaccurate) account of the evolutionaryprocess (…) the gene’s eye viewdoesfunction adequatelyas a theory of adaptation. (2011: 1809;emphasis added)

It is often this conflict of priorities—between searching foroptimal adaptations from the gene’s-eye view—vs.searchingfor accurate accounts of evolutionaryprocesses—that plays a central role in the tensions betweenthe two research schools. We shall discuss examples of thesecontrasting RQs in§4.1,§4.5 &§4.6.

3.3 Relationships Between Adaptationist/KS and Evolutionary Change Schools of Multilevel Evolution

As Wade and Goodnight summarize the biggest contrasts between theAdaptationist/KS (Fisherian) and Evolutionary Change (Wrightian)schools in evolution, an analysis using the “Logic andpragmatics of Research Questions” tells us that, not only do thetwo approaches address differentRQs, seeking differentpossible and responsive answers to those RQs (Lloyd 2015;see, e.g., Goodnight 2000), but they also make a variety of differentassumptions in building their model-types in addressing those RQs:

…the Wright-Fisher controversy involves the fundamental natureof evolutionary change…[including] its genetic basis (universalepistasis and pleiotropy vs. additive genetic effects), the ecologicalcontext in which it takes place (small, subdivided populations vs.large, panmictic populations), and the mechanisms by which it operates(local, mass selection, random genetic drift, and interdemic selectionvs. mutation and mass selection). (Wade & Goodnight 1998:1547)

Thus, the two research approaches to multilevel selection founded onthese two distinct schools’ model-types^[39] are not at all the same, because they simply do not ask the same RQs,nor use similar model components, values, or assumptions^[40] (Goodnight 2000, 2013, 2015; Linksvayer & Wade 2005; Wade 2016;Wade & Kalisz 1990; Rousset 2004; Ketcham 2018).^[41]

Rather than being competing approaches, as they have often beentreated since the 1960s, the two research schools may today beconsidered complementary approaches that, when used together, give aclearer picture of social evolution and multilevel theory and geneticsthan either one can when used in isolation, although there are soundreasons for preferring one approach over another in certain backgroundconditions and population structures (Goodnight 2013: 1547; 2015; Wade2016). In summing up their analysis, biologists Kramer and Meunierwrite:

The two theories thus provide different perspectives that might befruitfully combined to promote our understanding of the evolution ingroup-structured populations. (Kramer & Meunier 2016:abstract)

But many researchers from the Adaptationist/KS school hold lesstolerant attitudes than seeing this “fruitful combination”towards the multilevel selection models—concerned primarily withprocess—rather than their preferred objective of optimal fitnesstraits/engineering adaptations.^[42]

We shall see, insections 4.1–4.6, how approaching RQs using one rather than the other researchframework and model-types to multilevel selection makes a differencein outcomes.

4. The Anatomy of the Debates

4.1 Group Selection

The legendary near-deathblow in the 1960s to group panselectionism wasreally about groupbenefit, not group selection per se (G. C.Williams 1966). The interest was in cases of a genuine group selection(GS, often, henceforth) process among groups, where the groups also,as a whole,benefitedfrom organism-level traits(including behaviors) that seemed disadvantageous to the organism(Wynne Edwards 1962; G. C. Williams 1966; Maynard Smith 1964). Butfrequently, a group benefit was not necessarily a group engineeringadaptation (§2.3,§2.4; G. C. Williams 1966; Brandon 1981, 1985; Brandon &Burian 1984; Sober & Wilson 1998).^[43]

Implicit in this discussion of GS processes, is the assumption thatbeing a “unit of selection” at the group level requirestwo distinct, interrelated but separable evolutionaryfunctions or roles: (1) having the group as an interactor, and (2)having a group-level engineering-type adaptation accumulated from agroup selection process. Thus, the entire discussion confuses andcombines two distinct RQs, the interactor question and themanifestor-of-adaptation questions, calling this combined setthe “unit of selection” question (true for bothcritics, G. C. Williams 1966; Maynard Smith 1964; and defender, WynneEdwards 1962; see Borrello 2010).

Thus, the GS issue was understood to hinge on “whether entitiesmore inclusive than organisms exhibit adaptations”. (Hull 1980: 325;^[44] Maynard Smith 1976; Ramsey & Brandon 2011) Similarly,paleontological approaches required engineering adaptations combinedwith interactors, for species and lineages to count as units ofselection (§4.2).^[45]

Finally, in an argument meant to distinguish GS and kin selection (KShenceforth), it was argued that GS is favored by small group size, lowmigration rates, and rapid extinction of groups infected with aselfish allele;^[46] thus, the ultimate test of the group selection hypothesis will bewhether populations having these characteristics tend to show“self-sacrificing” or “prudent” behavior morecommonly than those which do not.^[47] (Maynard Smith 1976: 282; see Wade 1980a,b, 1985 on kinselection)

Contrary to Adaptationist/KS school theoreticians, Evolutionary Changebiologists have produced a wide variety of models, lab, and fieldexperiments that violateall of the Adaptationist/KSschool’s “necessary” conditions usually cited for GSto be effective (Matessi & Jayakar 1976: 384; Wade & McCauley1980: 811; Boorman 1978: 1909; Uyenoyama 1979; Uyenoyama & Feldman1980; Wade 1976, 1977, 1978, reviewed in 2016; contrast, e.g., MaynardSmith 1964, 1976).

That different researchers reach such disparate conclusions about thetheoretical efficacy of GS is largely because they areresponding to different RQs, using different model-types withdifferent parameters, and different assumptions from differentschools. That they make vastly different generalizations aboutbiologicalfacts has to do, rather, more with ignorance ofEvolutionary Change school experiments and findings (e.g., Gardner& Grafen 2009; Gardner 2015a; West, El Mouden, & Gardner 2011;compare Wade 1996, 2016; Goodnight 2015; see below).

For example, many Adaptationist/KS model-types use a specificmigration mechanism: they assume that migrating individuals mixcompletely, forming a “migrant pool” from which migrantsare assigned to the future generation randomly. Under this approach,small sample size is needed to get a large genetic variance betweenpopulations (Wade 1978: 110; 2016; Slatkin 1981b; Okasha 2006).

In fact, Evolutionary Change models and empirical studies allestablished thatthere are many different ways for a group toreproduce, giving rise tomany differententities—e.g., propagules, not just genicreproducers—playing the functional role of reproducers inmodels in which groups are interactors and manifestors of adaptation.^[48] This is very different from the gene’s eye view whereheritability is strongly tied to gene-level variation (Wade 2016).When comparing kin/genic selection with group/multilevel selectionmodel-types and their assumptions—e.g., the facts about spreadof phenotypes—through either migration or reproduction, thedevil is in the details.

This is evident in an early, very controversial case focusing on groupvs. organismic selection of a virus. Focusing only on theevolutionary interactors, using an Evolutionary Change approach, helpsclarify theoretical, methodological, and pragmatic issues that infuseGS debates.^[49]

TheMyxoma virus was introduced into Australia because it washighly fatal to the out-of-control rabbit population. After someyears, however, the rabbits had become resistant, and the virus hadbecome less virulent. The debate focused on whether the avirulence ofthe virus had evolved primarily from organismic or group selection. Itwas a challenging problem because both model-types produced the sameoutcome: decreased virulence (Fenner 1965). In the models,mixedgroups with more virulent viruses inoculated into the rabbitsspread fewer virulent viruses in the metapopulations of groups, andpure groups of a single type of very virulent virus particlesalso reproduced less well in the metapopulation, thus producing lowerlevels of virulence in either case. Both biologists and philosopherstook firm stances, claiming either that organismic selection (Futuyma1979; Alexander & Borgia 1978; Gilpin 1975), or group selection(Lewontin 1970; S. Levin & Pimentel 1981; Sober 1984; Sober &Wilson 1998) was responsible for the observed avirulence.

In a philosophical analysis, it was highlighted that we did not knowenough to actuallydecide this debate about GS (Lloyd 1986,1988[1994]). Researchers did not agree about the facts regarding thecomposition of the inoculations of viruses injected: were theyhomogeneous or heterogeneous—all virulent, benign, or mixed? Ifthe inoculated virus groups were observed to be heterogeneous, then itwould be possible to apply appropriate statistical examination tothem, as would be expected in the Evolutionary Change approach toGS.

An aspect of the debate about the efficacy of group-level interactorsthat has received a great deal of consideration in both biology andphilosophy, concerns the conceptual and statistical tools necessaryfor identifying an interactor, as mentioned inSection 2.1.^[50] It was proposed to look for genetic or reproducer interaction at thelower level of entities, and coordinated, “emergent”fitness—i.e., relations among the entities, traits (characters),environmental contexts, and fitnesses—at the upper level of entities.^[51] Today, such statistical and causal examination uses precisely suchcontextual analysis in actual multilevel studies (Arnold &Fristrup 1982; Griesemer & Wade 1988; Goodnight 2015; see Bijma2014 and Wade 2016 for theoretical context). Contextual analysisproduces a multilevel multiple partial regression analysis. Earnshawsummarizes its products nicely:

\(w \mathbin{\Delta} z = a\,\textrm{Var} (Z) +b\, \textrm{Var} (z)\),where \(a\) and \(b\) are partial regression coefficients: is theaverage linear effect of group character on individual fitness whenindividual character is held constant, and \(b\) is the average lineareffect of individual character on individual fitness holding groupcharacter constant. (2015: 306)

The higher and lower-level characters are treated as independentlyhaving their causal effects on fitness, and they are combined withvariances in lower and upper-level characters, to give a change in theorganismic (or group) characters. Once such an analysis isdone—and only then—can we state what entity-typesmight be acting as an interactor at a particular level, andexplore them causally, in the process. As Damuth and Heisler claimed,their method does not help answer questions about units of selectionexcept empirically.^[52]

The key issue highlighted in the Myxoma case was that at the time ofthe above discussion, no one knew what the composition of the groupswas, so no contextual or other statistical analysis could be performedto decide the case.^[53] Yet G. C. Williams had discouragedgathering suchinformation using his widely accepted and promulgated “principleof parsimony”:

In explaining adaptations, one should assume the adequacy of thesimplest form of natural selection, that of alternative alleles inMendelian populations,unless the evidence clearly shows that thetheory does not suffice. (G. C. Williams 1966: 55; emphasisadded).

The following crucial problem with Williams’ principle was madeclear: by highlighting the proposed emergent fitness/trait relationaldefinition of an interactor (e.g., applying the suggested additivitycriterion using contextual analysis), it

reveals the potential for dogmatic abuse of such sensible-lookingadvice. “The evidence”, as Williams puts it, is not found,it is created. Not performing analyses of group composition andcontribution to the global gene pool virtually guarantees…thatno evidence will be found that will reveal the inadequacy of thesimpler…. [genic or organismic level] models….Understanding the structure of selection models and the informationneeded to perform an adequate comparison reveals quite clearly thedogmatism of Williams’ maxim, as it is usually applied. (Lloyd1988 [1994: 95])

This critique of G. C. Williams’ parsimony principle and theresulting insistence to obtain the empirical information necessary toperform the needed contextual statistical and causal analyses toanswer the question about which entity was aninteractor inthis context, was, perhaps surprisingly, welcomed by G. C. Williams,citing the above page numbers, when he reviewed the quoted book in theQuarterly Review of Biology: he described Lloyd’sanalysis of group selection in this case as “mak[ing] more sensethan any other I have seen” (G. C. Williams 1990: 504).

The proposed Evolutionary Change school-style theoretical,methodological and pragmatic requirements for something to be arecognized instance of GS—where groups are interpretedasinteractors alone, andnot necessarily manifestors ofadaptation, as Williams had previously required—were thusacceptedby the original arch-critic ofgroup selection. The crucial distinction between the functionalevolutionary roles of interactor and manifestor was subsequently usedin Williams’ own work (1992) on hierarchical species and lineage selection.^[54]

These roles are subsequently conflated again by philosophers Okashaand Paternotte (2012), who give a persuasive analysis of how toidentify an evolutionary interactor, much like the above, usingcontextual analysis plus a linked causal analysis, as recommended byDamuth and Heisler (1988), but conclude—under anAdaptationist/KS school approach—that they have identified anadaptation of the original Williams-engineering variety.^[55]

Notably, phenotypic traits involved in group selection processes ofinteractors are often (usually)not optimal or engineeringadaptations of the entities at that upper level, but rather simplyselection-product adaptations following from their being interactorsat that level.^[56] The fact that this feature of Evolutionary Change models is alsocalled “adaptation” has sometimes proved confusing to many(Sober & Wilson 1998; see Lloyd 2008 correspondence with Sober andD. S. Wilson on this issue; Wade 2016; Gordon 2014).^[57] But contextual analysis is a good method for answering researchquestions aboutboth Adaptationist/KS and Evolutionary Changeinteractors at any given level.^[58]

The outcomes of group selection models can also be destructive (butstill, selected) traits. Consider, for example, very common phenomenathat Wade highlighted as almost completely neglected by theAdaptationist/KS school: these damaging population consequences ofartificial and commercial breeding have

received much less theoretical or empirical attention, even though itis the most common situation in the domestication of plants andanimals, (Wade 2016: 18)

He cites experimental and field results regarding selection andgroup-harmful traits, e.g.: cannibalism in flour beetles (Wade 1980b);field work on cannibalism in willow leaf beetle (Breden & Wade1989); leaf area in plants (Goodnight 1985); egg laying in hens (Muir1996); and eusocial mortality in hens (Ellen et al. 2008). Wadeconcludes:

The results of these studies leave little doubt that artificial groupselection can curb the evolution of traitsgood for the individualbut harmful to the group. (2016: 18–19; emphasis added)

It is easy to see, in this context, how the requirements for the rolesof multilevel interactors, and multilevel manifestors of engineeringadaptation, may be conflated and confused. In fact, Heisler and Damuththemselvesdid not make the relevant distinction, citingSober’s (1984) two-pronged, conflated,interactor-plus-manifestor, “common fate”^[59] definition of group selection as equivalent to their approach(Heisler & Damuth 1987: 588).^[60]

Let us take another look at Heisler and Damuth’s contextualanalysis, and the Price Eqs. Overall, while many of the suggestedtechniques to identify interactors have had strengths, no singleapproach to this aspect of the interactor question has been generallyaccepted across all researchers working fromboth schools ofevolution, and indeed it remains the subject of debate in biologicaland philosophical circles, perhaps partly because the Adaptationistschool, as well as many philosophers, largely prefer onemethod—using the Price equations—while the EvolutionaryChange school prefers a different one—using multiple regressionanalyses, in the form of contextual analyses for any and all modelingmethods and applications in the field and laboratory experiments(Gardner & Grafen 2009; Gardner 2015a,b,c; Grafen 2009; Queller1992; Goodnight 2015; see discussions in Wade 1995, Price 1970, 1972;Heisler & Damuth 1987; Damuth & Heisler 1988; Okasha 2006;Earnshaw 2015; Goodnight 2015. 2020; Wade 2016; Ketcham 2018; contrastBourrat 2021a,b).

Let us take a moment to consider the notion that perhaps thesedifferences in statistical tool use have consequences for thefoundational understanding, especially of the Adaptationist/KS school,affectingwhich of the units RQs are considered central tothe practice of evolutionary genetics, which has thus far beenreferred to as “conflation” and “confusion” ofRQs.

Lande, Arnold, and Wade, all at University of Chicago, found thatevolution by selection could be best understood if analyzeddistinguishing processes of selection from those of heritability. AsWade reviews in his Evolutionary Change-school book on multilevelselection theory and empirical applications, the strength of selectionhas been shown by breeders to be representable as the covariancebetween phenotype and relative fitness, )^[61] (Robertson 1966; Wade 2016; Lande & Arnold 1983; Ohta 1983). And,with regard to multilevel selection, it was later shown that thecovariance between phenotype and fitness

across an entire metapopulation, \(\textrm{CovTotal}(z, w[z])\), canbe partitioned into two components (Price 1972):

1. average individual selection within populations,\(\textrm{CovIndividual}(z,w[z]) = \Sigma j\textrm{Cov}j(z, w[z])/T\);and
2. group selection among populations, \(\textrm{CovGroup}(Z,W[Z])\).

This simple partitioning gave us a method for empirically comparingthe relative strengths of individual and group selection. (Wade 2016:88)

Recall that the Price EQ, in its single-level form,

represents the change in (the mean state of) a character, through theequation with this right-hand side, (where this first term on theright-hand side of the equation is usually called theselectionterm), which shows the covariance of the character, \(z\), andthe reproductive output of the entity manifesting the trait. Thesecond term, often called thetransmission-bias term,measures the extent to which the character of the offspring entities,on average,zbarprime, is different from theparent.

One usual way, a simple alternative to the one introduced above, towrite about the multilevel representation of the Price EQ is asfollows, where \(Z\) is a character-trait of the higher-levelcollective, \(k\) is an index individual of the collective, and the(relative) reproductive success of the collective is \(\Omega{:}\)

In this upper-level form, the left-hand term of the right-hand side ofthe equation is called “the between-higher-level (orbetween-collective) selection term”, while the farright-hand term is “thewithin-collective selectionterm”, conventionally (see, e.g., Sober & Wilson 1998).And whether or not the higher-level trait or fitnesses are aggregateor more complex relationships of the individual/particle levelfeatures, is a vexed issue (Okasha 2006; Lloyd 1988[1994]; Hamilton1975; Sober & Wilson 1998; Gould 2002).

But notice what has happened here, in moving back and forth betweenthe individual level and the multilevel representations of selectionusing the Price Eq. What is all aboutheredity—the“transmission-bias term”—right-hand term inthe single-layer Price Eq., turns into aselection right-handterm in the two-layer or multilevel Price Eq, and thus, all about aselection process: “thewithin-collective selectionterm” (Sober & Wilson 1998); (I thank J. R. Griesemerfor bringing this point to my attention; Price was aware of thisweakness of his formalism, which is one reason he did not think his“toy” model-type applicable in real biological situations).^[62]

Does this mean that we have confused an inheritanceprocess—reproduction—with lower-level selection, in theprocess of our formalizing and generalizing the theory!? Or…?We might understand the two causal evolutionary processes, ofinheritance of individual-level traits, vs. selection of lower-levelentities, on distinct levels of organization, as“entangled”.

As Griesemer put it, pointedly:

that’sconflation if you call the expectation term an“inheritance” term but “entanglement” if youtreat it as a representation of a hierarchical structure that slicesacross perspectives (sensu Wimsatt on “interactionalcomplexity” 2007)^[63]

Thus, this might be an explanation of why so many researchers from theAdaptationist/KS school domix togetherinto onething the requirements for the interactor functional role and therequirements for manifestor-of-adaptation, as documented throughoutthis entry.

As such, it might provide yet a further argument (see Earnshaw 2015)to abandon the Price EQ formalism as a good way to representany multilevel selection process, whether KS or GS orspecies/lineage selection; it predicts the entanglement of causalprocesses that most analysts are trying to keep distinct! In otherwords, the tool is inappropriate for the job. Why not use theavailable tools (contextual analysis and any other causal analyticaltools, such as graph theory (Otsuka 2016a, 2019a,b), as theEvolutionary Change biologists do? This remains an open questionconcerning best methods.

Today, the most urgent questions about the interrelations of genic andmultilevel selection models and model-types revolve aroundtwodistinct RQs (Research Questions) that are of concern to bothschools (Kramer & Meunier 2016). The first concerns thekey relationships between the models themselves: including genic,kin-, and inclusive fitness selection models and model-types from theAdaptationist school, and group-, and multilevel selection model-typesfrom the Evolutionary Change school. Let us put it like this:

(1)

Are kin selection (KS) as well as inclusive fitness, and groupselection (GS) model-types, equivalent, formally (i.e.,mathematically), empirically, conceptually, and/or explanatorily?

The kin selectionist or Adaptationist school, tends to say that thesetwo types of models, KS/genic/inclusive fitness and multilevel/GS, arefully formally (i.e., mathematically) equivalentalternatives, and the researchers usually conclude from thatostensible formal equivalence a kind of pragmatic equivalence; forexample, ‘use whichever model-type suits your project better’.^[64]

The basic prediction of KS theory is that social behavior, especiallysocial behavior that benefits others, should correlate with geneticrelatedness or similarity. This is commonly expressed throughHamilton’s rule, \(rb > c,\) where \(r\) is relatedness,\(b\) is the benefit that behavior offers the conspecific, and \(c\)is the cost to the actor.^[65] The critics following Hamilton, i.e., claiming that KS is a form ofGS, assert that Hamilton’s rule “almost never holds”(Nowak et al. 2010: 1059). Their opponents claim the opposite: that itis incorrect to claim that Hamilton’s rule requires restrictiveassumptions, or that it almost never holds. On the contrary, theyclaim, it holds a great deal of the time (Gardner, West, & Wild2011). On one philosophical analysis, there are three distinctversions of Hamilton’s rule, and three distinct versions of kinselection theory under discussion; thus the various parties aretalking past one another (Birch & Okasha 2015).

Various equivalences and non-equivalences among the key model-typesare discussed by researchers from both schools (see Adaptationist/KSworks by Foster 2004; Okasha 2006; 2018; and Evolutionary Change booksby Wade 1985, 1996, 2016; Simon 2014). Discussions by philosopherssuch as Earnshaw have highlighted some distinctive non-equivalences inmethods (2015), arguing that the methods favored by theAdaptationist/KS school (such as applying the Price Eqs) obscureimportant relationships (see examples of this in Bourrat 2021a,b;Queller 1992, 2011; discussion in Van Veelen et al. 2009a,b,2012).

Among the Adaptationist/KS school, the claims to complete“equivalence” of genic, kin selection, inclusive fitness,and multilevel or GS models, are usually taken to signify completemathematical, conceptual, and explanatory equivalence of the modeledevolutionary systems, a claim contested by Evolutionary Changetheorists.

As an illustration of this approach, West et al. describe theAdaptationist/KS school view in vivid detail: arguing that one

…source of confusion isthe incorrect ideathat inclusive fitness theory or kin selection are … justspecial cases of, new group selection….Perhaps most importantly,there is no biological model orempirical example that can be explained with the new group selectionapproach, that cannot also be understood in terms of kin selection andinclusive fitness. (West, Griffin, & Gardner 2007:425; emphasis added)^[66]

The introduction of this name, “new group selection”, bythe Adaptationist/KS school, to refer to the 1970s–2000s’stheoretical and empirical developments of Evolutionary Change theory,distinct from KS, was often accompanied bya complete erasure ofthe actual history of multilevel selection theory, as descendedfrom and continuous with the work of Sewall Wright, most specificallyhis work on evolution in structured populations, and especially theevolution of organismic adaptations within demes in such structuredpopulations (1931, 1938, 1945, 1960).

This ahistorical approach is complemented by a replacement, fictionalhistory that includes having multilevel selection theory developed,ab initio andex nihilo, by Hamilton in the early1960s. Some examples:

Historically, MLS [multi-level selection] and kin selection theorywere developed to address “altruism”: the question of whyan organism should engage in behaviours that seem to reduce its ownexpected reproductive success but which aids that of others (Hamilton1963, 1964). (Earnshaw 2015: 308)

Hamilton helped to found a field that shows how natural selection canact at any level of biological organization. (Foster 2011: 193; citingD.S. Wilson 1975 and Okasha 2006 for this claim.)

In any case, as well as being contested by some within theAdaptationist/KS school (e.g., Thies & Watson 2021),^[67] those who claim full equivalence tend to ignore the applications ofthe relevant GS models from Evolutionary Change genetic results (e.g.,Wade & Breden 1981), which produce easy maintenance of cooperationand altruism through GS and avoid the “cheater fallacy”;instead they argue that “only kin selection allows an easysolution” to evolutionary problems producing cooperation andaltruism (West, Griffin, & Gardner 2008; West, El Mouden, &Gardner 2011, see conflicting results below). Group selectionistclaims are also e.g., rebutted by Marshall 2011, who gives aparticularly succinct summary of this argument, using Price Eq andmodifications of Hamilton’s rule, see also Queller 1992,highlighting issues with Price Eq.^[68]).

In answer to the equivalence RQs, the Evolutionary Change school tendsto say, first, that the model-types are not conceptually orexplanatorily equivalent, and sometimes not even mathematically equivalent,^[69] or empirically equivalent. For example, mathematical geneticistFeldman et al. note:

[T]he comparison between those fertility models interpretable in termsof individual fitness effects, and the simple symmetric case thatcannot be so interpreted, underscores that the simplest interactionsbetween individuals in the process of selection can produceevolutionary conclusions not expected from standard individual fitness[genic] models. (Feldman, Christiansen, & Liberman 1983: 1009)

Moreover, the Evolutionary Change biologists sometimes add, theAdaptationist/KS practitioners may not truly understand the models andstructure of the model-types, and therefore think that they areequivalent in a number of ways, when they are not (Lloyd & Feldman2002; Lloyd, Lewontin, & Feldman 2008; Wade 2016; Goodnight 2015).^[70]

When these equivalence RQs about genic and multilevel selection modelsand model-types are discussed in the philosophical literature, muchhas depended on citing the work of Dugatkin and Reeve in establishingthis formal equivalence (Dugatkin & Reeve 1994; Sterelny 1996a,Sterelny 1996b: 577; Sober & Wilson 1998: 57, 98–99;Sterelny & Griffiths 1999: 168–169, 172; B. Kerr andGodfrey-Smith 2002: 479, 508; Waters 2005: 312; Okasha 2006). However,Dugatkin and Reeve’s very plain prediction of allele frequenciesis an extremely simplistic method for assessing formal model ormodel-type equivalence, which pays little mind to the details of themodel and model-types themselves or their dynamics.

Because allelic parameters and the changes in allelic frequenciesdepend on genotypic fitnesses, the genic model-types claimed to beequivalent to the hierarchical or multilevel model-types areneither parametrically nor dynamically sufficient (Dugatkin &Reeve 1994; compare Bijma & Wade 2008; Simon 2014; Lewontin 1974).Thus, the claimed equivalence seems to be based on conceptual andtheoretical errors or mis-identifications.

This citation practice continues in philosophy today, as part of aclaim to a metaphysical or epistemic (or methodological)conventionalism, without addressing the empirical and dynamical issueschallenging the claimed failures of equivalences in both biology andphilosophy (see§4.4).^[71]

As an example of theoretical misunderstanding of the genetics andevolutionary models, those supporting the Adaptationist/KS school inthe 1970s were taken to task by a founder of evolutionary genetics formistakenly jumping to the conclusion that because group engineeringadaptations might be rare, that “natural selection ispractically wholly genic” (S. Wright 1980: 841; Ågren 2021b).^[72] This is a fair criticism of the Adaptationist/KS school logic andrecord; for decades, critics of GS never considered groups asinteractors, only as interactors-cum-manifestors, and this mistakenpractice continues today in many animal behavior research groups.

Some leading Adaptationist/KS school genic selectionists did lateracknowledge the significance of theinteractor/manifestor-of-adaptation distinction RQs at the group level(G. C. Williams 1990: 504, above; 1992; Maynard Smith 1987: 123; 2001:1497), but their distinctions are not widely shared amongAdaptationist/KS school advocates. Despite the views of these leadingtheorists, the Adaptationist school maintains a bias towards using KSto modelall selection processes involving collectives,therefore leaving GS effectively replaced by KS.^[73] But see Hamilton (1975), the originator of KS theory, who argues theopposite (in contradiction to West, Griffin, & Gardner 2007, 2008,quoted above; also West, El Mouden, & Gardner 2011).

And when we consider the Wrightian, Evolutionary Change approach ofmultilevel selection model-types’ empirical accomplishments, itmay be difficult, at first, to understand the resistance of theAdaptationist/KS school.^[74] Overall, the Evolutionary Change school has demonstrated thatpopulations respond rapidly to experimentally imposed group selection,and that very commonly occurring indirect genetic effects (IGEs) areprimarily responsible for the strength and effectiveness of groupselection experiments, perhaps surprising, considering the many claimsof the “full mathematical equivalence” between multilevelselection models and kin/genic selection models of these same systemsthat do not show such phenomena.^[75]

Also significant are the field studies using contextual analysis thathave shown that effective and powerful multilevel selection is farmore common in nature than previously expected (Goodnight 2013; e.g.,Stevens, Goodnight, & Kalisz 1995; Tsuji 1995; Aspi et al. 2003;Eldakar et al. 2010; Wade 2016). Many of these traits are involved incooperative behaviors, functions integrated at the organismic level,and sometimes group benefits due to group selection (i.e., withgroup-level manifestors; Davidson et al. 2016; Suárez &Triviño 2020; Seeley 1997); see, for example, Muir’scooperative chicken evolution by group selection, discussed below (J.Craig & Muir 1996; Muir 2005; Wade, Bijma, et al. 2010; seeShaffer et al. 2016; Fewell & Page 2000). But these results havemuch bearing on the next RQ.

The other leading Research Question involving GS today,primarily significant for the Adaptationist/KS school,is:

(2)

Does or can Group Selection lead to Group Adaptation?^[76]

Let us reconsider the two interwoven RQs of a typical Adaptationist/KSschool multilevel research problem (section 3):

• First: Identify an equilibrium/optimal higher-level trait which isan engineering adaptation, i.e., a manifestor-of-adaptation, and whichis unlikely to evolve by selection of the lower-level entity traits,that is assumed to oppose it.
• Second: Characterize a KS or inclusive fitness model/process, inwhich this engineering upper-level-optimal adaptation could haveevolved against the pressures of the lower-level entities’selective processes.

We can see immediately now why kin selectionists/Adaptationist schoolpractitioners find a common Evolutionary Change approach tohigher-level selection unsatisfactory. Because Adaptationist schooltheorists are primarily interested in detecting and explaininghigher-levelmanifestors of engineering adaptation, thensimply modeling the changes of higher-levelinteractors, asthe Evolutionary Change school predominantly aims to do, will beinadequate; it won’t answer either of the RQs stated immediately above.^[77]

Many using an Adaptationist/KS school approach persist in conflatingGS itself with the combination of the two functional roles (e.g.,Ramsey & Brandon 2011). A vivid example of such confusion is therequirement, imposed nearly universally in the Adaptationist School,that selection favoring the higher-level trait of an interactor mustalways workopposing selection favoring the component orlower-level traits of the interactors making up those groups (MaynardSmith 1976; West, El Mouden, & Gardner 2011; contrast Wade 1978; 2016).^[78]

However, because GS model-types allow us to model a very wide varietyof processes and systems compared to an Adaptationist/KS approach,including many relevant to the evolution of social behavior,Evolutionary Change theorists have often argued that their theoreticaland empirical findings are important for the Adaptationist school tocomprehend and use (Goodnight 2015; Goodnight & Stevens 1997). Forexample, an Evolutionary Change biologist concludes, forcefully, thatif we follow Maynard Smith’s distinction between KS and GS:

… then this [experimental, empirical] study illustrates thatgroup selection is more favorable for the evolution of socialbehaviors than is kin selection. To the extent that morphology,physiology, and development affect the manifestation of socialbehaviors, then these traits will also be influenced in theirevolution by population structure. (Wade 1980b: 854; emphasis added)^[79]

This is an instance of the fact that KS usually relies upon differentmodel-assumptions about population structure than intergroup selection(Hamilton 1975; Grafen 1984, 2014; Wade 1980a, 1985); hence, it may bepossible to determine which explanation or model is more likelyaccurately describing the population structure (Wade 2016; Ketcham2018; for general criticism of pluralism in GS, see Shavit 2005; seealso Gordon 2023; Davidson et al. 2016 for an enactment of some ofthese GS dynamics in ants).^[80] In other words, the empirical evaluation of population structure,specifically, in models, can be a crucial part of model evaluation and confirmation.^[81]

Another Evolutionary Change school example demonstrates the power ofGS and its population structures on both organismic and group traitsin recent commercial breeding of the domestic chicken. Chickenbreeders had long selected for even tiny improvements (approx. 1%) inegg-laying rates of domesticated hens, which had not recentlyimproved, due to conflict among hens, with high mortality rates; thisled to the practice of burning their beaks off. However, when thesebreeders took a GS approach, selecting hengroups that gotalong comparatively better, this produced not only “kinder,gentler” hens, but also an improved egg-laying productivity upto 60% in only 6 generations (Muir 1996; Cheng & Muir 2005; Wade,Bijma, et al. 2010). This demonstrated both the group and organismicselective and adaptive power of GS regimes. Colony-level bee and antadaptations are other good examples of group- or higher-leveladaptations arising from higher-level selection processes (seediscussion in Seeley 1997; Davidson et al. 2016; Gordon 2014). Lammand Kolodny (2022) recently proposed a concept of “distributedadaptation”, which we may interpret as group-level engineeringadaptations obtained through processes of individual and higher-level selection.^[82] Shifting the focus to groups gives access to new sources of genetic,heritable variation available for breeders to manipulate.^[83]

Finally, for another crucial but underappreciated contribution fromthe Evolutionary Change theorists to the Adaptationist/KS schoolproblems, consider the so-calledfree rider problems with kinand GS, such as those seen or set up by evolutionists puzzling overthe evolution of altruism and the purported problems with cheaters tocooperation. Evolutionary Change theorists have repeatedlydemonstrated that thesefree rider problemsare pseudo-problems based on misconceptions (Breden & Wade 1991;Wade 2016; see also Bowles & Gintis 2011; Planer 2015; Sterelny2012; see Lloyd & Wade 2019 on mutualism selected as a runawayprocess, rather than selfishness).

Take a population mostly made up of altruists that have evolved by GS,e.g., as described by Hamilton 1975. Those from the Adaptationist/KSschool claim that such populations are constantly under threat ofdestruction by cheaters—individuals that enjoy the fitnessbenefits of altruistic behavior by others, but do not pay the costs byperforming it themselves, in Hamilton’s terms.

According to Wade and Breden (1981), a “selfish” mutationis no threat to the population of altruists.^[84] Take a population divided into groups with A, an altruistic gene withfitness costs and benefits following Hamilton’s rule. Mutationintroduces a mutant copy, a, into the population, and a heterozygousindividual carrying this mutation does not behave as altruistically asothers, but reaps the fitness benefits from them, and is thus at aselective advantage relative to altruists within its own group. Doesthis selfish gene spread through the metapopulation, destroying thesocial system, as claimed by the Adaptationist school? Wadeanswers:

No it does not. Felix [Breden] and I showed that it spreads if, andonly if, \((1/2k)b \textrm{Total} < c\) (Wade & Breden 1981;Breden & Wade 1981). This is the opposite of Hamilton’sRule, the condition necessary for the existence of the altruisticsystem in the first place…. Hamilton’s Rule gives theconditions for a single altruist to invade and replace a population of“original” cheaters. As long as it holds, mutant cheatersshare the fate of the original cheaters—they are replaced by altruists.^[85] (Wade 2016: 146–147)

There is thus an evolutionary equilibrium level of cheaters at akin selection-mutation balance, analogous to the usualmutation-selection balance, which depends on the strength ofgroup selection. Wade concludes very straightforwardly:

The perception that cheaters are a relentless threat to complex socialsystems is afallacy…If Hamilton’s Rule werecorrectly grasped as the condition necessary for group selectionto outweigh opposing individual selection (as Hamilton himself [1975]recognized), thecheater fallacy would have been laid to resta long time ago. (2016: 147; emphasis added)

As we shall see inSection 4.6, these very biases in the evolutionary dynamics ofinteraction—in that case, between multi-species consortia, inplace of multi-organismic populations—are an important featureof the former’s dynamics, and could contribute to explaining theabsence of cheaters from natural mutualisms—establishedrecently, see below—although cheaters are predicted asinevitable by gene-centered conflict theory tocause the strongevolutionary instability of such mutualisms (see, e.g., Foster etal. 2017).

As we have just seen in this section, according to the EvolutionaryChange school, the predicted problems with GS and KS model-types fromcheaters were based on a fallacy: It turns out that these expectedcheaters alsocould not be found in nature, when thoseresearchersmost committed to their existence went to findthem, using their best tools. Testing their own“best-studied” and routinely-taught exemplars of cheatersamong mutualisms in nature for fulfillment of their own besttheoretical requirements for cheaters, Jones and more than a dozenother Adaptationist/KS biologists conclude, in their own words:

We find … there is currently very little support fromfitness data that any of these meet our criteria to be consideredcheaters. (2015: 1270, emphasis added; see§4.6 for further discussion)

On the other hand, the Evolutionary Change school’s epigeneticscommunity genetics model-type easily explains the experimental resultsthis group found in this failed search for cheaters-in-nature experiment.^[86]

Additionally, it may help explain the more frequent origin ofmutualisms from parasitic than from free-living systems, anevolutionary trajectory opposite to that predicted by theAdaptationist/KS school’s genome conflict theory.

4.2 Species Selection

Ambiguities about the definition of a unit of selection have alsosnarled the debate about selection processes at the species level. Bythe 1970s several leading paleontologists had claimed that speciesmust have higher-level “emergent properties”—bywhich they meant species-level engineering adaptations—in orderto serve as genuine higher-level “units of selection” in amultilevel evolutionary process. Species selection thus succumbed fora time to the same confusions as GS had, conflating these two RQs: Dospecies function as interactors, playing an active and significantrole in evolution by selection? And does the evolution ofspecies-level interactors produce (and assume) species-levelengineering adaptations (and, if so, how often)?

See especially, for example, Vrba’s demand for “emergentproperties” at the species level, explicitly inspired by MaynardSmith’s demanding analysis of GS (Vrba 1984: 319; Maynard Smith1976). This amounts to assuming that there must be a group-levelengineering adaptation in order to say that group selection can occur,an early-Williams-style objection thatboth Maynard Smith andWilliams later abandoned, as illustrated in§4.1 (G. C. Williams 1990, 1992; Maynard Smith 1987; 2001), but this dualrequirement for both interactor and manifestor at a higher level wascarried over into early-mid species selection discussions, althoughnot in those terms.

For the early history of the species selection debate, then, askingwhether species could beunits of selection meant asking acombination of whether they fulfillboth the interactor andmanifestor-of-adaptation roles, based directly on that argument fromGS (Vrba 1983: 388; 1984; Vrba & Gould 1986; Vrba & Eldredge1984; Eldredge 1985; see also Sober 1984: 367–368; Cracraft1985: 225).

For example, certain cases are rejected as higher-levelselectionprocesses because

frequencies of the properties of lower-level individuals which arepart of a high-level individual simply do not make convincinghigher-level adaptations. (Eldredge 1985: 133)

Such an “emergent” character may be the result of aselection process at the species level, but it should not be treatedas a pre-condition of such a process. As Heisler and Damuth putit:

The issue is not whether some characters are emergent or not; theissue is the relationship between the characters and fitness (Heisler& Damuth 1987). (Damuth & Heisler 1988: 418)

Neither group effects on organismic fitness, nor emergent groupcharacters, are necessary for selection to occur at the level ofgroups in nature, which may involveany characters of thehigher-level group.^[87]

Consider the lineage-wide trait of variability. Treating species andlineages as interactors has a long tradition (Dobzhansky 1956; Thoday1953; Lewontin 1958; Lloyd 1988[1994]; from here on within thisdiscussion, I will just refer to “species” to signifyboth species and lineages). If species are conceived asinteractors (and not necessarily manifestors), then the notion ofspecies selection is not vulnerable to the original antigroup-selection objections from the early genic-selectionists (MaynardSmith 1976, 2001; G. C. Williams 1966; 1992), as they later agreed.^[88] Thoday’s old idea was that lineages with certain properties ofbeing able to respond to environmental stresses would be selected for,and thus that the trait of variability itself would be selected forand would spread in the population of populations or species. In otherwords, lineages were treated as interactors. The earlier researchersspoke loosely of “adaptations” where they seemingly meantproduct-of-selection adaptations.

These early researchers were explicitlynot concerned withthe effect of species selection on organismic level traits but withthe fitness effect on species- and lineage-level characters such asspeciation rates, lineage-level survival, and extinction rates ofspecies. Some argued that this sort of case represents a perfectlygood form of species selection, using so-called “emergentfitnesses” (Lloyd 1988[1994]),^[89] even though some balk at the thought that lineage-level variabilitywould then be considered, under a product-of-selection definition, aspecies-level adaptation; see discussion in Gould 2002; Ketcham 2018;Grantham 1995). Hautmann noted:

An important argument in favour of the “emergent fitnessconcept” is that species selection acting onaggregateorganismic traits can theoretically oppose selection at the organismiclevel and is therefore not reducible to this level (Grantham 1995).(2020: 3 of 11)

Paleontologists used this approach to species selection, e.g., intheir research on fossil gastropods, looking at relationships betweenlineage ranges and trait variabilities (Jablonski 1987, 2008;Jablonski & Hunt 2006; Erwin 2000; Sepkoski 2020; see recent studyof dynamics of these in Freedberg, Urban, & Cunniff 2021); theapproach has also been used in the leading text on speciation (Coyne& Orr 2004).^[90] The variabilities associated with increased or variable sized rangeshave been explored along these lines, as well.^[91]

While species geographic range is well-established as an interactortrait (e.g., Nanninga & Manica 2018), it has been noted that

A variety of other traits [beyond geographic range size] have beenexamined (reviewed in Coyne & Orr 2004; Jablonski 2008; Dynesius& Jansson 2014), many of which might covary with demographiccontrols. These results are consistent with links between demographyand speciation rates. (Harvey, Singhal, & Rabosky 2019: 83)

Ultimately, the current widely-accepted definition of speciesselection among the scientists is consistent with an interactorinterpretation of a unit of selection (in addition to those justdiscussed: Vrba 1989; Lieberman & Vrba 2005; Jablonski 2008;2017a,b; Erwin 2000; Polly et al. 2017; Anderegg et al. 2019; BenitoGarzón, Robson, & Hampe 2019).

4.3 Genic Selectionism

One may understandably think that the early genic selectionists wereinterested in the replicator version of the reproducerquestion—that is, what entity is the replicator?—becauseof their claims that the unit of selection ought to be the replicator.This would be a mistake. Rather, the primary interest is in a specificontological question about benefit (Dawkins 1976, 1982a,b), (§2.4), and the answer tothat question dictates the answers to theother three “units” questions delineated inSection 2.

Briefly, because replicators are the only entities that“survive” the evolutionary process, they must be thelong-term beneficiaries (Dawkins 1982a,b). What happens in the processof evolution by natural selection happensfor their sake, fortheir benefit. Hence, interactors interact for the replicators’benefit, and adaptations belong, ultimately, to the replicators.Replicators are the only entities with realagency asinitiators of causal chains that lead to the phenotypes; hence, theyaccrue the credit and are thereal units of selection.

This version of the answer to “the” units of selectionquestion amounts to a combination of the beneficiary question plus themanifestor-of-adaptation question—plus the interactor question,because a manifestor must also be an interactor in anevolution-by-selection process. Dawkins’ argument is that peoplewho focus on interactors in isolation—as so many do, who arestudying the dynamics of evolutionary change—or onorganismic-level adaptations, as many from the Adaptationist/KS andEvolutionary Change schools both do, are laboring under amisunderstanding of evolutionary theory (Dawkins 1976, 1982a,b).

In what follows, two aspects of Dawkins’ specific version of theunits of selection problem shall be characterized. I clarify the keyissues of interest to Dawkins and relate these to the issues ofinterest to others.

There are two mistakes that Dawkins isnot making. First, hestates explicitly that genes or other replicators/reproducers do not“literally face the cutting edge of natural selection. It istheir phenotypic effects that are the proximal subjects ofselection” (1982a: 47). Thus, he emphasizes that interactors(understood under Hull’s, Brandon’s, and Lloyd’sdefinition) are necessary for the evolutionary process; it is notnecessary to believe that replicators are directly“visible” to selection forces (1982b: 176;contraGould 1977; Istvan 2013). Moreover, he argues the debate about groupversus organismic selection is “a factual dispute about thelevel at which selection is most effective in nature”, whereashis own point is “about what we ought to mean when we talk abouta unit of selection” (1982a: 46).^[92] We shall return to this issue insection 4.4, Genic Pluralists.

Second, Dawkins does not specify how large a chunk of the genome hewill allow as a replicator; there is no commitment to the notion thatsingle exons are the only possible replicators/reproducers (see§2.2 Replicators/Reproducer Question).^[93]

On what basis, then, does Dawkins reject the importance of theresearch questions about interactors (§2.1,§3,§§4.1–4.6) and/or manifestors (§2.3,§3,§§4.1–4.6)? The answer lies in the particular question in which he is solelyinterested, namely, What is “the nature of the entityforwhose benefit adaptations may be said to exist?”^[94]

When Dawkins refers to “adaptations”, he is alwaysreferring to engineering adaptations. Moreover, he rejects the notionthat the individual organism or group that exhibits the adaptivephenotypic trait can be the “beneficiary” he seeks, (seeG. C. Williams 1966;§2.3,§2.4,§3,§§4.1–4.6),^[95] because a true unit ofselection must also be “the unit that actually survives or failsto survive” the evolution-by-selection process (Dawkins 1982a:60). Because organisms, groups, and even genomes do not actuallysurvive the evolution-by-selection process, the answer to the survivalquestion must be the replicator/reproducer, Dawkins reasons.^[96]

Strictly speaking, this is false, under Dawkins’ definition of areplicator; it iscopies of the replicators thatsurvive—or rather, Griesemer’s biologically-continuousreproducers,section 2.2; compare Griesemer (1998, 2000a,b,c, 2003, 2005, 2016) on reproductionand the evolution-by-selection process, which imposes the significantdistinction between “replicator” and“reproducer”.

The important point for Dawkins is the claim that replicators—insome formal sense, and reproducers, in the biological sense—arethe only long-term survivors of the evolution-by-selectionprocess—and more specifically, theevolution-of-engineering-adaptation process; this automaticallyanswers also the question of who manifests the adaptations.^[97]

But there is still a problem with putting the reproducer/replicator atthe heart of the process. Although this conclusion is:

there should be no controversy over replicators versus vehicles.Replicator survival and vehicle selection are two aspects of the sameprocess, (Dawkins 1982a: 60)

the genic selectionist does not just leave the vehicle selectiondebate alone. Instead, the argument is advanced thatwe do notneed the concept of discrete vehicles at all. We shallinvestigate this idea also insection 4.4 Genic Pluralisms.

It is crucial to notice that Dawkins’ argument against“the vehicle concept”, concerns exclusively an argumentagainst the desirability of seeing theindividual organism asthe vehicle. The narrowness of this point is lost on thegenic pluralists, to their detriment.

Dawkins’ target is explicitly those who hold what he calls the“Central Theorem”, i.e., thatindividual organismsshould be seen as maximizing their own inclusive fitness (1982b:5, 55). But it is inappropriatealways, Dawkinsargues, to ask how an organism’s behavior benefits thatorganism’s inclusive fitness.^[98] Dawkins warns, “theoretical dangers attend the assumptionthat adaptations are for the good of…the individualorganism” (1982b: 91). The precise view being criticized byearly genic selectionists thusassumes that the individualorganism is the interactor,and the beneficiary,andthe manifestor of adaptation (Lloyd 1988 [1994]).

These arguments are indeed damaging to the Central Theorem, but theyare ineffective against all other approaches that define thefunctional interactor role more generally, that is, with entities asinteractors up and down the biological hierarchy, whether or not thesearealso seen as engineering adaptations (see§3,§§4.1–4.6). These issues highlight why it is important not toconflate RQs about units of selection with questions about biological individuality^[99] (see the entry onbiological individuals).

In sum,^[100] Dawkins has identified or assumed four distinct meanings of a unit ofselection as necessary for something to be a unit ofselection—replicator (corrected to reproducer), long-term,ultimate beneficiary, ultimate manifestor of adaptation, and themaking (i.e., agency) of its own interactor or vehicle’s traits(including engineering adaptations). The issue of lower-level agencyand manifestor of adaptation in this context has been revived recentlyin philosophical discussion (Okasha 2018).

InSection 4.4, we will consider some relatively more recent work in which genicselectionism is defended through a pluralist approach to modeling.What matters in their final analysis, though, is exactly the same, andthat isthe search for the ultimate beneficiary of the evolutionby selection process. [[please add Links listed in NOTES 99 and100 for this page]]

4.4 Genic Pluralism

The original genic selectionists had objections (§4.3), to the interactor role in evolution. They admitted that such“vehicles” were necessary for the evolution-by-selectionprocesses, but accorded them no weight as “units ofselection”, because they were not the “ultimatebeneficiary”. Soon, there appeared to be a new angle available.^[101]

“Genic pluralism” attempts to bypass the challenges of therelationships among reproducer and interactor RQs by effectivelyturning genes into interactors (Waters 1986; Sterelny & Kitcher1988). There are thus two “images” of natural selection,presented as similar to a “Necker cube”: (1) the usualmultilevel one in which selection models are given in terms ofreproducer success via a hierarchy of interactors and possiblymanifestors, interfacing with their traits’ environments, up anddown the biological hierarchy (§3,§4.1). And (2)—here you imagine the flipped image of aNecker cube—is given in terms ofreplicators [genes orequivalent reproducers] having properties that affect their abilitiesto leave copies of themselves, while seeing only these “ultimatebeneficiaries”in the interactor roles, with geniccausal/interactive traits in various layered “environments”.^[102] When not done in a special, derivative, way, these latter are usuallycalledinteractors, but are now called “genicenvironments”.

For example, take the claim that

All selective episodes (or, perhaps, almost all) can be interpreted interms of genic selection. That is an important fact about naturalselection. (Kitcher, Sterelny, & Waters 1990: 160)

What exactly is meant by this? We are reminded strongly of many in theAdaptationist school’s claims about the mathematical,explanatory, and pragmatic equivalence of group selection models withkin selection/genic/inclusive fitness models, such as West, El Mouden,& Gardner’s 2011 false claims discussed in§4.1 (Wade 1996; 2016). And indeed, in the philosophical context, muchattention is paid to showing that the two model-types inquestion—genic and Evolutionary Change school’s multilevelones—can represent certain patterns of selection equally well,once underpinned with all the relevant theoretical, statistical, andempirical information. This is argued for using both specific examplesand schema for translating multilevel models into genic ones. Webriefly review how this is done, and then ask: But what exactly isaccomplished by demonstrating such a translation, theoretically,metaphysically, biologically, or philosophically?

In a classic account of the efficacy of multilevel selection, startingwith interdemic selection—the t-allele case that even G. C.Williams acknowledged was group selection—Lewontin & Dunn1960 and Lewontin 1962 found first, in this t-allele mouse example,that there was allelic selection; Second, they also found genotypicselection; Third, they also found a substantial effect of interdemicselection in the form of group extinction because female mice wouldoften find themselves in groups in which all males were sterile, andthe group itself would therefore go extinct. Establishing these factsand relations, then, is how a genuine and empirically robustmultilevel, interdemic, plus organismic, plus genic selection modelwas discovered, empirically explored, and developed (Lewontin &Dunn 1960 and Lewontin 1962).

The key to understanding the genic reinterpretation of this multilevelselection model type is to grasp that the pluralists use a concept ofgenic environment that their critics ignore. The pluralists explainhow to “construct”—actually, reconstruct—agenic model representing the evolutionary causes responsible for thefrequency of the t-allele. This is done by treating the informationabout interactors in the multilevel model (gained using theinteractor-identifying tools), as “genic environments” inthe new genic-level models.^[103]

Something philosophically significant is meant to follow from theclaims that multilevel Evolutionary Change or KS models’interactors can be reformulated or renamed in terms of thegenic/replicator level entities and their reconstituted hierarchicalproperties. These ostensible equivalences between multilevel andgenic model-types are actually very controversial, asreviewed inSection 4.1, and have been resisted on a variety of grounds.^[104]

But, assuming that the genic reformulated interpretations of themultilevel models’ structures have been accepted, the bigpayoff—both philosophical and biological—of the geniccausal point of view—and note that thecausallevel is simplyassumed in virtue of the allelic statespace—is claimed to be:

Once the possibility of many, equally adequate, representations ofevolutionary processes has been recognized, philosophers andbiologists can turn their attention to more serious projects than thatofquibbling about the real unit of selection. (Kitcher,Sterelny, & Waters 1990: 161; emphasis added)

By “quibbling about the real unit of selection”, here, theauthors seem to be referring to the large literature in whichevolutionists and philosophers have given concrete, empirical evidenceand/or proposed standards for something to serve in the significantlydistinct theoretical functional roles of an evolutionaryinteractor (and often also amanifestor ofengineering adaptation), or areproducer/replicator or abeneficiary—all instances of debates about “unitsof selection”—in an evolutionary causal selection process (§2,§3,§§4.1–4.6). This would, on the face of it, appear to be ordinary scientificpractice for evolutionary biologists, but these philosophers wouldlike to put a stop to it.

What interests the pluralist genic philosophers is aproposedequivalence between being able to tell the selection storyone way: in terms of the functional, theoretical, roles ofinteractors, manifestors, and reproducers; and telling the“same” story another way: purely in terms of theoretical“genic agency”.

Significantly, however, neither Dawkins himself, nor Williams, arguedtoeliminate the role, existence, or importance ofevolutionary interactors. Rather, both thought interactors(Dawkins’ vehicles) to beessential to understandingevolution-by-natural-selection processes. Thus, genic pluralists havefundamentally over-reached. The pluralists attack the view that

for any selection process, there is a uniquely correct identificationof the operative selective forces and the level at which eachimpinges. (Waters 1991: 553)

Rather, they claim, “We believe thatasking about thereal unit of selection is an exercise in muddledmetaphysics” (emphasis added; Kitcher, Sterelny, & Waters1990: 159), because “thecauses of one and the sameselection process can becorrectly described at differentlevels”, including especially the genic one (Waters 1991:555; emphasis added).

Moreover, these causal descriptions in purely genic terms, compared tomultilevel models, are onequal ontological and empiricalfooting. Equal, that is, except for when Sterelny and Kitcher make thedogmatic statement that biologists or philosophers make an error toclaim

that selection processes must be described in aparticularway, and their error involves them in positing entities,“targets of selection”, thatdo not exist.(Sterelny & Kitcher 1988: 359; emphasis added)

“Targets of selection” are also known asinteractors.

Thus, here, the genic pluralists are apparently denying theveryexistence of interactors in evolutionary selective processesaltogether, a position much more radical than Dawkins’or Williams’.

The pluralists thus are arguing against not only the utility, but alsothe existence, of any entity filling the functional, causal role ofthe evolutionary interactor in the evolution-by-selection process.They argue for this conclusion on the basis of their claim that all ofthese selection processes may berepresented using purelytheir translated models of genic-level causes. But they apparentlyfail to comprehend that the complete, genic-level story cannot be toldwithout taking the functional role of evolutionary interactors intoaccount; thus the pluralists cannot avoid ‘quibbling aboutinteractors’, as they desire (Dawkins 1982b; Brandon 1982;§4.1).

Recall what the interactor RQ amounts to: What levels of entitiesinteract as a whole with their environment through their traits (at agiven level of biological organization) in such a way that it makes adifference to reproducer/replicator success? (§2.1,§4.1). There has been much discussion about how to delineate,identify, and locate interactors among multilayered processes ofselection, see above,§2.1,§3,§4.1,§4.2, but each approach generally takes the notion of theinteractor as anecessary component to understanding anevolution by selection process, because it (Dawkins’ vehicle) isthe locus ofcausal interaction between the phenotypicdownstream product of a reproducer/replicator, and its environment(Dawkins 1976, 1982a,b; G. C. Williams 1966).

Or, put purely in Dawkins’ terms, most replicators’phenotypic effects are represented in “vehicles”(interactors), which are themselves the “proximatetargets of natural selection” (Dawkins 1982a: 62). For Dawkins,and this seems to also go for the genic pluralists, they simply do notwant to consider such objects ananswer totheessential units RQ, but rather prefer the answer isolating theultimate beneficiary of the evolution-by-selection process asthe sole “real unit of selection” (§2.4,§4.3).

Such a preference for answering the beneficiary question fails,however, to demonstrate that interactors (“targets ofselection”) do notexist, as Dawkins knew, and thusthey need to be explicitly and fully acknowledged in any completeaccount. As shown in§4.3, the only “vehicle”/interactor Dawkins was interested inarguingagainst was the animal behaviorist’sindividual organism per se, as “maximizing its inclusivefitness”, in the “Central Theorem”; this isdecidedlynot a rejection of the general evolutionaryfunctional role of interactors. Thus, this genic pluralists’misreading of Dawkins is both particularly widespread, andprofound.

Note that the pluralists use the identical methods for isolatingrelevant genic-level environments as others do for the isolating ofinteractors for multilevel genetic modeling. What, we may ask, is thereal difference between the resulting models in the end? The genicpluralists call them “causal alternatives”, but in whatsense are they trulydifferent alternatives?

What is different is that the genic pluralists want to tell the causalstory—discovered by investigating interactors andreproducers—in terms of genes alone (reproducer/replicators),andnever in terms of interactors, despite the fact thatthese were necessary for the causal analyses. Moreover, they proposedoing away with interactors altogether, byrenaming orreconfiguring them as the genic-perspective-environments. Are weto believe thatrenaming certain essential modelstructures actually changes the metaphysics or meanings of theevolutionary causal model-type?

This issue concerning renaming of model structures is especiallyconfusing in the genic pluralist presentations, because theyrepeatedly rely on an assumption or intuition that, by using anallelic state space, we aredealing with allelic causes. Thisassumption is easily traced back to the original genic selectionistviews on units of selection (§4.3) that alleles are theultimate beneficiaries of any long termselection process (G. C. Williams 1966; Dawkins 1976, 1982a,b).

Thus, the genic pluralist argument rests substantially, althoughalmost invisibly, on a commitment regarding the supreme importance ofthe ultimate beneficiary question (§2.4,§2.3,§2.1;§3;§4.1,§4.3). It is unclear what interest it has for eitherevolutionary theorizing, model-building, or research, (beyond itscorrective use for the “Central Theorem”,§4.3), and a further case for any philosophical or theoretical interest isnot clearly made by the genic pluralists. They make a faulty—tooradical—argument about the eliminability of interactors, andoffer nothing else that Dawkins has not already given.

Thus, while the genic pluralists attempt to dismiss the debates aboutresearch involving identifying andfinding interactors inselection as mere “quibbles”, and also claim such arewithout correct andincorrect answers, the genicpluralistsneed to know not only the research resultsinto the multilevel interactors, but also the correct answers tointeractor research questionsin order to get their desiredcorrect and accurate “allelic causes” in their genicmodels. Hence, how interactors are discerned/identified reallymatters, that is, it makes a metaphysical as well as practical,scientific, and epistemicdifference,to them (Lloyd 2005).^[105]).^[106]

In sum, genic pluralism’s impact has been largely philosophicalrather than biological. Within philosophy, critical responses to thegenic pluralist position fall into two main categories: pragmatic andcausal.

The pragmatic response to genic pluralism primarily notes that in anygiven selective scenario the genic perspective provides no informationthat is not available from the multilevel point of view. This state ofaffairs is taken by critics of this type as sufficient reason toprefer whichever perspective is most useful for solving the problemsfacing a particular researcher.^[107]

The other primary philosophical response to genic pluralism is basedon arguments concerning the problematic issues about the causalstructure of selective episodes (and these concerns are actuallyshared by many biologists). While genic pluralism may indeed get theshort-term “genetic book-keeping” correct, it does not,except through arbitrary renaming, or a kind of sleight-of-hand, andthrough blocking visibility of the essential interactor functionalrole, accurately reflect the evolutionary causal processes that bringabout the result in question (see discussion in§4.1; Wimsatt 1980a,b).^[108]

Let us summarize the consequences of the genic pluralists’celebrated genic equivalences and their actual derivativeness, interms of the science and metaphysics of the processes discussed.

First, the genic pluralists end up offering not, as they claim, avariety of genuinely diverse causal versions of the selection processat different levels. This is because the causes of the multilevelmodels, determined using ordinary statistical tools and definitions ofevolutionary interactors in their functional roles, are simply renamedin the lower-level models, but remain fully intact as relevantevolutionary causes at the full range of higher and lower levels,regardless. Perhaps more importantly, no new allelic causes areintroduced; they are simply renamed.

Second, while genic models may be derived from multilevel models, theyfail to sustain the necessary supporting methodology. Third, the lackof genuine alternative causal accounts damages the particular claimsof pluralism or, at least, of any interesting philosophical variety oftheoretical, empirical, or competitive pluralism separate from thisanatomy of different RQs. It is taken as given that there are avariety of ways to model many selection processes in populationgenetics; the key question here concerns specific informationstructured into those model-types and specific models. For example, abiologist and philosopher investigate various consequences of pursuingresearch using, alternatively, organismic and trait group models, andtheir paper is a useful modern application and explanation of IlanEschel’s 1972 neighbor selection work (B. Kerr &Godfrey-Smith 2002). But note that the authors help themselves at thestart of the paper to all the higher-level causal relationshipinformation they need.^[109]

Thus, there are no genuine causal alternatives being presented at thegenic level, unless you count renaming model structures asmetaphysically or explanatorily significant (Okasha 2011, 2018;Sterelny & Kitcher 1988). It should be noted that this result doesnot, of course, eliminate the possibility of the genic levelacting as interactors in any given case, such as meiotic drive (Lloyd1986). See discussion in§4.1 and§4.3 on Genic Selectionism.

4.5 Units of Evolutionary Transition in Individuality

The units under analysis so far—be they genes, organisms, orpopulations—were seen as more or less on the same level oforganization, before and after a given evolutionary process.“Evolutionary Transitions in Individuality” (ETI;alsoMET, “Major Transitions in Evolution”)present a unique set of complications to such assumptions within theunits and levels debates. This topic is explored under several entriesin the SEP:biological individuals,philosophy of macroevolution,evolutionary game theory. We emphasize its relations to the debates about units below.

New ETI Research Questions

Evolutionary transition is “the process that createsnew levels of biological organization” (Griesemer2000a: 69), such as the origins of chromosomes, multicellularity,eukaryotes, and social groups (Maynard-Smith & Szathmáry1995: 6–7). These transitions all share a common feature,namely, “entities that were capable of independent replicationbefore the transition can replicate only as part of a larger wholeafter it” (Maynard-Smith & Szathmáry 1995: 6). Theclaim is that these newly-reproducing, more-inclusive wholes mustinvolve new evolutionary dynamics collecting and reproducing them overgenerations of the more-inclusive entities.

ETIs can evolve such new potential levels and units of selection, bycreating new kinds of entities having the right sorts of relationsamong fitness and properties, orsuccessful reproduction.^[110] Such a task requires a diachronic perspective, one under which theproperties of our currently extant units of selection cannot bepresupposed or assumed. As Griesemer writes:

…[A]s long as evolutionary theory concerns [only] the functionofcontemporary units atfixed levels of thebiological hierarchy…, the functionalist approach, [usingevolutionary functional causal roles, e.g., of reproducers,interactors, beneficiaries, and manifestors], may be adequate to itsintended task. However, if a philosophy of units is to address… problems of evolutionarytransition, then adifferent approach is needed. (2003: 174)

Griesemer introduces and emphasizes the evolutionary functional roleof the “reproducer” concept (§2.2) to expand the “replicator” role, which incorporates theprocesses ofdevelopment into the units of selection; it is astep toward meeting the goal of addressing ETI, and

…the dependency of formerly independent replicators on the“replication” of the wholes—the basis for thedefinition of evolutionary transition … is adevelopmentaldependency that should be incorporated into the analysis ofunits. (Griesemer 2000a: 75)

Thus, thereproducer concept was introduced to encompass suchprocesses, expanding the notion of replicator (Griesemer 1998, 2000c).Griesemer emphasizes both development and materiality in hisdefinition:

The process of reproduction can be analyzed as multiplication ofmaterial overlapping propagules that confer the capacity to develop,specified in terms of the minimum notion of development as acquisitionof the capacity to reproduce. (2000c: 74)

Thinking in broader terms of reproducers avoids the presupposition ofevolved coding mechanisms implicit to the original concept ofreplicators. In the case of ETI, this allows us to separate the basicdevelopment involved in the origin of a new biological level, from thelater evolution of sophisticated developmental mechanisms for the“stabilization and maintenance of a new level ofreproduction” (Griesemer 2000a: 77; 1998; 2005; 2014; 2016;Grosberg & Strathmann 1998).

We can see that, in their picture of the ETI group of RQs, MaynardSmith and Szathmáry ask an overarching question about units ofselection from the Adaptationist/KS school perspective (1995):

Why did not natural selection, acting on entities at the lower level(replicating molecules, free-living prokaryotes, asexual protists,single cells, individual organisms), disrupt integration at thehigher-level (chromosomes, eukaryotic cells, sexual species,multicellular organisms, societies)?

In delineating the possible and responsive answers to this RQaccording to their preferred practices, Szathmáry and MaynardSmith state clearly their commitment to using a genic-centeredapproach—specifically, the cheater dynamics of anAdaptationist/KS genic approach (see§4.1 on “cheater fallacy”; also§§4.3–4.6):

The transitions must be explained in terms of immediate selectiveadvantage to individual replicators: we are committed to thegene-centred approach outlined by G. C. Williams (1966), and madestill more explicit by Dawkins (1976). (Maynard Smith &Szathmáry 1995: 8)

The notion is that higher levels of entities may beselected—i.e., be interactors, or perhaps even become attributedto genes. The fact that such Adaptationist/KS school researchers seemto be primarily focused on the unique reproductive capacities of theupper-level entity (expanded to, e.g., the multicellular organism, orfamily, or even superorganism, in ETI), explicitly to the detriment oflower-level entities (see Clarke 2013, 2016, who conflates multilevelselection and evolutionary transitions model-types and processesthroughout work on this topic,^[111] as do Sukhoverkhov & Gontier 2021;^[112] or Okasha 2006, 2022) should not surprise us (e.g., Gardner 2015a,b).What is more surprising is that some insist that such genicadaptation-forming processes are theonly major multilevelselection processes that produce or are involved in evolutionarychange (contra§4.1,§4.2,§4.6).

On one analysis, the different stages of an evolutionary transitioninvolve different conceptions and model-types of multi-level selection(Okasha 2006, 2015).^[113] Under the first stage of multilevel selection (MLS1), the lower-levelentities (usually organisms) are the interactors as well as thereplicators/reproducers, while in MLS2, both the upper-levelcollectives as well as the lower-level entities are independentinteractors. Thus, the issues surrounding evolutionary transitionsinvolve both the interactor question and the reproducer question, atleast. Unfortunately, as mentioned in§4.1, there is conflation of interactor with manifestor in both sources(Okasha 2006; Damuth & Heisler 1988), thus also tying bothfunctional roles in with the reproducer RQ, confusing theseissues.

On a different approach, ETI are seen as the appearance of a“new kind ofDarwinian populations…. that canenter into Darwinian processes in their own right”(Godfrey-Smith 2009: 122; emphasis added). Both the two-fold andfour-fold set of functional roles in evolutionary processes (isolatedby either Dawkins & Hull, or Lloyd) are denied or neglected, andit is directly denied that avariety of distinct units ofselection RQs exist (§§2.1–2.4,§3):

Questions about the “unit” of selectionare notambiguous; the units in a selection process are just the entitiesthat make up a Darwinian population at that level. (Godfrey-Smith2009: 111, emphasis added)

The characteristics of “the unit of selection”are then isolated (Godfrey-Smith 2009). In this view, as the essenceof the process of natural selection is the existence ofparent-offspring similarity between the elements that constitute aDarwinian population,the (“Unitary”^[114]) unit of selection must be any element of a population of causallyinteracting elements where the elements have the capacity of formingparent-offspring lineages with at least a weak degree ofparent-offspring similarity (i.e., individuals that have the“capacity to reproduce”, or “reproducers”,under a definition modified from Griesemer; see Godfrey-Smith 2016:10120). These elements are referred to asDarwinian individuals^[115], and to populations composed of Darwinian individuals asDarwinian populations.^[116]

Generally, in ETIs, the higher-level entities or populations mustengage in reproduction themselves, in which independent reproductionand evolution at the lower-level is suppressed, as Maynard Smith andSzathmáry emphasized. These processes all involve restrictionson the ability of the lower-level entities to function as interactorsand replicators/reproducers, and the emergence of upper-levelcollectivesas simultaneous interactors, manifestors, andreplicators/reproducers. But that is not how Godfrey-Smithanalyzes it.^[117]

Thus, there is a variety of philosophical and biological approaches toanalyzing ETI on offer, whether in terms of reproducers, multilevelselection, or Darwinian populations. The essential diachronic natureof the problem poses a challenge for modeling, and involves not justthe interactor and replicator/reproducer questions, but also thequestions of who is the beneficiary of the selection process, whetherreproduction at the higher level constitutes a group-level engineeringadaptation, and how that new level of coordinated trait emerges.

The clearest issues of interest emerging from these studies seem tobe:

1. that the introduction of the “reproducer” as anexpansive concept of replicator was a major innovation (see Griesemer1998, 2000c, 2003, 2005, with Godfrey Smith introducing a variation onthis in 2009, 2016);
2. the multilevel dynamics claimed to be appropriate in the“Evolutionary Transitions in Individuality” (ETI)tradition, (although the validity of these claims is under challenge),which focus on reproduction of the higher-level entity in coordinationwith suppression of its lower-level entity parts, seem not to bebroadly applicable to multilevel evolution in general; to treat themas such introduces confusion into the debates about the units ofselection, as we see in the following section on holobiontevolution.

4.6 Holobionts, and Demibionts

New ideas brought to notions of units of selection have grown out ofwork from Margulis, Buss, Bordenstein, Thies, O’Malley,Dupré, Griesemer, and evolutionary-developmental biologistssuch as Rudolf Raff and Scott Gilbert, part of it taken up inecological community genetics modeling from the Evolutionary Changeschool, and led to a recent focus onholobionts—aeukaryotic host and its diverse microbial symbionts and associates.^[118] An expansive definition of holobionts, going beyond their role assimple symbionts, is currently in wide use:

Microbial symbionts can be constant or inconstant, can be verticallyor horizontally transmitted, and can act in a context-dependent manneras harmful, harmless, or helpful. (Theis et al. 2016: 1;^[119] see Bruckner & Bordenstein 2012, 2013a,b; Bordenstein & Theis2015)

Each “individual” human being is accompanied by acommunity of organisms, some co-evolved for mutual benefit (Gilbert,Sapp, & Tauber 2012; Chiu & Gilbert 2015; Gilbert 2014).^[120] Our microbiota (bacteria, viruses, and fungi living in our gut,mouth, and skin) is necessary for our survival and development; often,our species is also needed, in turn, for their survival.^[121]

Zilber-Rosenberg and Rosenberg (2008) were the first^[122] to claim that holobionts were units of selection, but it was unclearwhat precisely this claim meant, among the various possible meaningsand their combinations discussed in this entry. The holobiont as aunit of selection was also accompanied by loosely-related, agitateddebates over whether and how it could be a biological individual^[123] (see entries onbiological individuals andevolution and development).

Within the decade, discussions of holobionts as “units ofselection” appeared in the literature, notably, oftenobjecting to holobionts as units of selection (Moran &Sloan 2015; Douglas & Werren 2016; Queller & Strassmann 2016;Skillings 2016; Burt & Trivers 2006). However, it was rarely clearwhichexact claim about holobionts as units was beingdenied—was it about well-defined interactors, manifestors,beneficiaries, reproducers, or some combination? Once specificdefinitions were introduced into the debates researchers began tooffer both claims of filling the functional roles of interactors andmanifestors of adaptation selection and evolution among holobionts,and demonstrating the empirical evidence supporting such claims(Griesemer 2005, 2014, 2016; Bouchard 2009, 2010; Lloyd 2017; Gilbert,Rosenberg, & Zilber-Rosenberg 2017; Griesemer 2018; Roughgarden etal. 2018; Suárez & Triviño 2020). But there is amore fundamental theoretical problem, arising from a specifictheoretical confusion with ETI (§4.5).

Explicit discussion and clarification of this central theoreticalquestion has been nearly completely neglected in the discussions ofholobionts as units:^[124] which evolutionary multilevel model-type(s) should we use torepresent holobiont evolution? A cross-time-scale emergenthierarchical model-type,^[125] like the ETI Maynard Smith/Szathmáry one discussed inSection 4.5, focusing on the emergence of new, higher-level-entity reproductioncombined with near-total suppression of the lower-level fitness/traitsignificance? Or, instead, an ordinary-contemporary-timescaleEvolutionary Change multilevel/GS selection model-type, which does notrequire such suppression, as we discussed in§3 and§4.1? Or both, perhaps?

As was highlighted in 2019 about this very issue in the context ofholobiont evolution:

The first confusion involves the attribution of “evolutionarytransition”-type, higher-order-entity selection as anecessary precondition for holobiont evolution to have acooperative, mutualistic dimension (Stencel and Wloch-Salamon 2018).In this view, a host and its loose association of microbes becomes aholobiont only if itfirst undergoes an evolutionarytransition, because the holobiont must bethe primary unit ofselection in ordertopermit the evolution ofan adaptive mutualism through the reciprocal coevolution ofits component species.

The Evolutionary Change school theorists characterizing and critiquingthis ETI view continue:

Without the primacy of holobiont-level selection, [it is believed]genomic conflict at lower-levels of selection sustainsdisharmony among the evolutionary interests of the componentspecies in a holobiont (Moran & Sloan 2015; Douglas & Werren2016). (Lloyd & Wade 2019: 3–4 of 20, emphasis added)

Under this ETI approach to units of selection (discussed§4.5), that we can now apply to holobiont selection, Maynard Smith andSzathmary’s ETI-style upper-level selection is required first,which success depends upon the total choking-off of lower-levelentities through selection at that lower level, or else, it isbelieved, as explained in§4.1, “cheater” mutations that benefit from, but do notcontribute to, the upper-level group—here, theholobiont—are thought to make mutualisms evolutionarily unstable(Stencel & Wloch-Salamon 2018).

In contrast, a multilevel Evolutionary Change GS or hierarchical model-type^[126] admitting the possibility of multiple, simultaneous levels ofselection at all stages of holobiont evolution, isnotdependent upon theprior occurrence of a major evolutionarytransition in individuality (ETI), and finds the introduction andnecessity of control of the “threat” of“cheater” mutations to be afallacy (§4.1,§4.5,§3).

This leaves us with at least two distinct theoretical RQs concerningholobionts: (1) how should we conceive and answer questions aboutwhether any parts or combinations of multilevel, multispeciesholobiont systems serve as a unit of selection? and (2) whichfunctional roles do such units play? We will gloss several preliminaryanswers for both RQs here, as well as views of some primary critics.^[127]

A holobiont has been claimed to sometimes be an “integratedcommunity of species, [which] becomes a unit of naturalselection” (Gilbert, Sapp, & Tauber 2012: 334; Gilbert,Rosenberg, Zilber-Rosenberg 2017). That is, theorists claim that theholobiont can sometimes function as an evolutionary interactor, sinceit has featuresthat bind it together as a whole in such away thatits traits interact with its biotic and abioticenvironments, affecting the relative success of its reproducers,in a selective process.

What ties or “binds” the different species together toproduce such an interactor in a selection process? According topioneering philosophical thought on holobionts and symbionts, it isthese species’ common evolutionaryfate, the holobiontacting as a “whole”, that characterizes it as anevolutionary interactor, “objects between which naturalselection selects” (Dupré 2012b: 160; 2021a,b; see alsoDupré & O’Malley 2013; Griesemer 1998, 2000a,b,c;Zilber-Rosenberg & Rosenberg 2008; Suárez 2020 emphasizesthe “stability of traits”; Schneider 2021; see Love 2018et al.).

The details of this biological description are filled out in variousinstantiations of the model-type. This community can also be describedas a “team” undergoing selection (Gilbert, Rosenberg,& Zilber-Rosenberg 2017). Others describe them as“collaborators” or “polygenomic consortia”,which has the advantage of encompassing both competition andcooperation within the holobiont (Dupré & O’Malley2013: 314; Dupré 2021a,b; Theis et al. 2016; see also Huttegger& Smead 2011, game-theoretic results regarding the range ofcollaboration). Note that all these descriptions are compatible withbeingsimple interactors at the combined level; and to callthemcooperators, and so on, here, does not necessarilyinvolve them in manifestor-style engineering higher-level adaptations,i.e., this does not engage themanifestor-of-engineering-adaptation functional role (§4.1), as Wright made clear, as did Hamilton, and Wade (see Koskella &Bergelson 2020 discussion).^[128]

And, anticipating many critics, Evolutionary Change multilevelmodel-types representing holobionts do not identify a bottleneck, suchas vertical transmission, as either a cause or a necessary conditionfor higher-level selection (e.g., Sober & Wilson 1998;Godfrey-Smith 2009, 2016, see below). Instead, multilevel selectioncan provide a pathway for the evolution of transmission modes andmating systems that enhance the efficacy of higher-level selectionwhile diminishing that of lower-level selection (§4.1). Evolutionary Change multilevel selection models demonstrate theemergence of holobiont dependencies whichprecede and attend,rather than follow, mutualistic processes.^[129]

In contrast, some philosophers describe cases of ordinary groupselection, correctly modeled using perhaps kin-group selection ormultilevel GS model-types; they nevertheless describe phenomena inthose models as, “key events in an evolutionarytransition”, which they are not (Clarke 2016: 41). Because thereis no first-time upper-level, multi-cellular or multi-entity eventemerging along with these groups of organisms, calling them an ETI,and using ETI-structured model-types to represent them, constitutes afundamental confusion (see also Herron 2021; Clarke 2013, 2016; Inkpen& Doolittle 2021).^[130]

In fact, consideration of the problems with generalizing ETI models toother evolutionary contexts such as holobionts re-invigorates thefour-fold taxonomy of the units of selection debates originallyintroduced as anepistemological and pragmatic anatomy of RQsbased on the different types of processes/functions^[131] that can affect different forms of biological organization(Dupré 2012a,b, 2021a,b).

For example, take the requirement that one single entity-typesimultaneously satisfiesthree of the key functionalevolutionary roles, grounded in an Adaptationist/KS approach (§3; Griesemer 2005; see meiotic drive in§4.4; Lloyd 1986, Hull 1988; Brandon 1988;§3,§§4.1–4.5), which reduces evolution solely to a process thatproduces engineering-typeadaptations. Some see this type ofreduction as disconnected from a substantial number of genetic modelsand biologicalprocesses, and one that matches badly withcontemporary experimental and theoretical biological practice (§3 &§4).^[132]

Given this latter, multilevel theoretical background, and theprominence of debates about it, it seems somewhat surprising that ithas often been simplyassumed, in the wide-ranging discussionof holobionts, that Maynard Smith and Szathmáry’sAdaptationist/KS, genic modeling approach was theonlyapproach available or appropriate (Moran & Sloan 2015; Douglas& Werren 2016; Queller & Strassmann 2016; Skillings 2016).Alternative Evolutionary Change model-types, especially from communitygenetics, have been available that make many fewerempirically-unsupported or questionable assumptions—e.g., about“cheaters”—regarding the biological systems inquestion (§4.1; Koskella & Bergelson 2020; Bijma 2014; Drown & Wade2014).

Some evolutionary biologists, for example, in applying Williams’“parsimonious” reductionist methods to holobionts(specifically, mammalian gut microbiota; see§4.1 discussion of Williams’ parsimony principle), are explicitabout using Adaptationist/KS school model-types and assumptions,claiming that “the study of [adaptive] function providesa universal logic to understand complexbiological systems”^[133] (Foster et al. 2017: 2, Box 2).

The holobionts are analyzed assimultaneously filling threeevolutionary functional roles: interactors, reproducers, andmanifestors of adaptation, in the active search foroptimathat satisfy their requirements for adaptative, optimal results ofcooperation, stability and longevity of multispecies systems that arealways potentially fundamentally undermined by cheaters (Foster et al.2017; they take an explicitly cheater-oriented approach to modelingsymbiotic evolutionary systems).^[134]

On this Adaptationist/KS approach, it is quite difficult to evolvecooperation or mutualisms, despite mutualisms’completepenetration and ubiquity in every system ever found in nature;the emphasis in their model-types is on the roles of cheaters andselfish traits.

One result appearing often in Adaptationist/KS school literature, isthat challenges to the “stability oftransmission”—partner choice of a symbiont—arecrucial to its long-term evolutionary success (Kaltenpoth et al. 2014;Douglas & Werren 2016; H. Williams & Lenton 2008).^[135]

This objection, that the symbiont or holobiont cannot evolve (as aninteractor, but especially as a both an interactor and a manifestor),because transmission is unstable or unreliable—and thus thatstrong or complete vertical transmission is required—is foundvery frequently in conference and informal one-on-one discussions ofthe possibilities of holobionts being units of selection, as well asin the literature on this topic (e.g., Skillings 2016; Godfrey-Smith2016; Douglas & Werren 2016; Foster et al. 2017; van Vliet &Doebeli 2019).

It has been established, however, through community geneticsmodel-types, that strong or complete verticality of transmission isnot, in fact, required for symbiont evolution, nor evolution ofmutualism, nor for holobiont evolution or stability, although it maylater emerge.^[136]

Thus, while we find some researchers fruitfully framing theirmodel-types in the context of ecological modeling practices to betterunderstand microbiome evolution within vertebrates and other complexhosts (e.g., Coyte, Schluter, & Foster 2015; Costello et al. 2012;Ley et al. 2008), others are stuck in unproductive model-types arisingfrom the ETI context, or fundamentals of the Adaptationist/KS schoolmodeling space that may make inappropriate and unproductiveassumptions.

Nevertheless, it should be recalled that even under the EvolutionaryChange and multilevel selection models, manifestors of adaptation maybe found at the holobiont level, as discussed in: Seeley (1997);Wagner et al. 2014; Suárez & Triviño (2020);Dupressoir, Lavialle, & Heidmann (2012); Lamm & Kolodny(2022).

Within the practice of philosophy of biology, on the other hand, whenbuilding a theoretical and analytical framework to examine whether ornot holobionts are units of selection, there is a tendency toresurrect the anatomy analysis given here by those previouslyfollowing the “Darwinian individuals” approach, whiledenying to do so; this can be found in a duo of philosopher &biologist:

We do not in general favor the replicator/interactor model (Hull 1980)as the best or only way to understand selection, but it is worthnoting that it can be coherently employed in this context, and in away that is substantially different from other proposals. Lloyd(201[8]), for example, conceives holobionts as interactors thatpromote the differential success of the lineages of cells that makethem up.We offer analternative in which holobiontsare seen as interactors that promote the differential success of theinteraction patterns they instantiate. The replicators in our modelareabstract functional relationships, not cell lineages.(Doolittle & Booth 2017: 18–19; emph. added)

Since all the roles in the analysis in question are already abstractand functional,^[137] the quoted material represents a misconception; the authors seem tobe in complete agreement with the definitions of the anatomyanalysis.

Later, with a different philosopher, the biologist promotes aprocess-centered approach to units of selection (Doolittle &Inkpen 2018; Suarez and Lloyd 2023). They again generalize thefunctional roles of interactors in selection processes, but seem tomisunderstand the function-based tradition in which that project isthe norm (Dupré 2012a, 2021a; Nicholson & Dupré2018; Griesemer 2005).

By 2021, these authors have adopted a particular units RQ as theirown, which appears to accept the beginnings of an anatomy analysis:“What are the referring replicators specifically responsible forthe differential extinction and proliferation of such higher-levelentities?” i.e., of “multi-species communities andecosystems”. (Inkpen & Doolittle 2021)

But referring to replicators as being causally responsible for thedifferential reproduction or extermination of any higher-level entityrequires their translation into interactors and their traits first, asDawkins himself recognized and claimed (see§§4.1–4.4). These authors seem to be making the same fundamental omission as thegenic pluralists, in thinking that reproducers as ultimatebeneficiaries—or as serving foundational roles—aredirectly exposed to selection forces themselves (§4.3,§4.4).

They thus leave out the names and descriptions of the fullydynamically and explanatorily essential functional parts of theanatomy analysis—beneficiaries, reproducers,interactors, andmanifestorsofadaptation—even though these authorsutilize allfour of these functional roles in their theoretical explorationand explanation (Inkpen & Doolittle 2021). Omitting naming them isnot an advance in clarity or understanding, it is a confusion echoingthose inSection 4.4.

Evolutionary Change modelers thus conclude that applying the ETImodel-type to holobionts as units of selection (regarding any of thefour distinct units RQs or their combos), as done by the opponents togranting holobionts status as units, isnot generallysuitable for modeling holobionts, demibionts or symbionts, for reasonsthat are familiar from the weaknesses of genic interpretations ofmultilevel model-types generally. For example, such model-types do notfully incorporate relevant gene interaction—e.g., IGE or sourcesof epistasis—in their models (see Koskella & Bergelson 2020;§4.1; Bijma 2014; S. Wright 1931, 1980; Wade 2016).

Leaving out the necessary gene- or entity-interactions amongparticipant species or higher-level entities, it is no wonder that theETI-oriented models do not get the empirically or theoreticallyrepresentative answers about upper-level properties.

There are thus crucial differences between the outcomes of the twoschools’ model-types. The community or multilevel geneticsmodel-types (incorporating IGE or epistasis) showhow theassumedconflict in the ETI model-type that shows betweengenomes (e.g., Douglas & Werren 2016; Moran & Sloan 2015), canbeself-limiting in the Evolutionary Change model-types,whilecooperation and mutualism tend to be self-accelerating,precisely contrary to the predictions of the ETI model-types, i.e.,those used by the opponents to holobionts as interactors, reproducers,or manifestors as units of selection (Drown & Wade 2014; Wade& Drown 2015; see theory in Wade 2016).

These very biases in the evolutionary dynamics of interaction betweenhosts and microbiota are an important feature of holobionts, and couldcontribute to explaining theabsence of cheaters from naturalmutualisms, although cheaters are predicted by gene-centered conflicttheory tocause the strong evolutionary instability ofmutualisms, as mentioned in Section 4.1 (see Foster et al. 2017).Recall fromSections 4.1–4.5 that, according to the Evolutionary Change school, problems with GSand KS model-types from cheaters were a fallacy. Recall also thatthese cheaters alsocannot be found in nature, when thoseresearchers most committed to their existence went to find them.^[138]

It is important to see that in addition to filling the functionalroles of interactors or manifestors, holobionts can also bereproducers, where the host usually reproduces vertically and themicrobiota reproduce vertically, horizontally, or both.^[139] This situation has provoked discussion especially among philosophers.^[140] Holobionts’ microbiota often reproduce outside the context ofthe original host organism, and some holobionts, e.g., the Hawaiianbobtail squid and its luminescent bacteria, are not seen as“Darwinian populations” (Godfrey-Smith 2009, 2016), andtherefore are judged not to be units of selection under that view (seeBooth 2014).

Others instead distinguish between holobionts and“demibionts”. In contrast with the adaptive-neutralconcept of the holobiont, ademibiont recognizes thepossibility of an asymmetric evolutionary history, in which thecombination evolves such that one or more lineages has adapted to thefirst, but the first has not adapted to the others (see Suárez& Stencel 2020).^[141]

This approach contrasts with that of the original reproducer and“Darwinian population” approaches, which would excludedemibionts given the specific requirements concerning reproduction(Griesemer 2000a, 2016; Godfrey-Smith 2016; Booth 2014). The Darwinianpopulation view excludes such holobionts and demibionts, because“it is a mistake to see things that do not reproduce as units ofselection” (Godfrey-Smith 2011: 509). Note that this conclusionrests on the merging of the interactor (and manifestor) with thereproducer requirements, and as such will not hold sway over those whodo not buy such a confounding of roles (e.g., Dupré &O’Malley 2013; Suárez & Lloyd 2023). This is yetanother case wherein distinguishing the interactor RQ from thereplicator/reproducer RQ can be “more illuminating”.^[142] (Sterelny 2011: 496)

In sum, both the ETI model-type, inappropriately, and theAdaptationist/KS gene-centric, species-centric approach—based ingenomic conflict and the threat of “cheaters”—havebeen predominant in critiques ofholobiontsas units ofselection under all four RQs discussed in this entry, and havebeen resisted by advocates using Evolutionary Change model-types,whose models fit the available empirical data concerning what isturning out to be the universal partially-coevolutionary systemscalledholobionts.

5. Conclusion

We should not treat different answers as competitors if they areanswering different questions. I offer a framework of four distinctResearch Questions to clarify the “units of selection”debates. We separate the classic question about the unit ofevolutionary selective interaction (interactor question), from theentity functioning as a replicating/reproducing unit (reproducerquestion), and also from which entity acquires“engineering” adaptations (manifestor-of-adaptationquestion). We note the ambiguity in “adaptation”: isadaptation a selection product versus an accumulated“engineering” design? Finally, we consider the ultimatebeneficiary, “in the long run”, from the evolution byselection process (beneficiary question). Manymisunderstandings—theoretical and empirical—may be avoidedby a precise characterization of which of the four units of selectionquestions, or their combinations, is being addressed, and by whichmultilevel school’s approach.

Bibliography

• Abbot, Patrick, Jun Abe, John Alcock, Samuel Alizon, Joao A. C.Alpedrinha, Malte Andersson, Jean-Baptiste Andre, Minus van Baalen,Francois Balloux, Sigal Balshine, et al., 2011, “InclusiveFitness Theory and Eusociality”,Nature, 471(7339):E1–E4. doi:10.1038/nature09831
• Ågren, J. Arvid, 2021a,The Gene’s-Eye View ofEvolution, Oxford: Oxford University Press.doi:10.1093/oso/9780198862260.001.0001
• –––, 2021b, “Sewall Wright’sCriticism of the Gene’s-eye View of Evolution”,Evolution, 75(10): 2326–2334.doi:10.1111/evo.14334
• Alexander, Richard D. and Gerald Bargia, 1978, “GroupSelection, Altruism, and the Levels of Organization of Life”,Annual Review of Ecology and Systematics, 9: 449–474.doi:10.1146/annurev.es.09.110178.002313
• Allen, Benjamin, Martin A. Nowak, and Edward O. Wilson, 2013,“Limitations of Inclusive Fitness”,Proceedings of theNational Academy of Sciences, 110(50): 20135–20139.doi:10.1073/pnas.1317588110
• Allen, Colin, 2002, “Real Traits, Real Functions?”, inFunctions: New Essays in the Philosophy of Psychology andBiology, André Ariew, Robert Cummins, and Mark Perlman(eds.), Oxford/New York: Oxford University Press, 373–389.
• Alves, Davi Mello Cunha Crescente, José Alexandre FelizolaDiniz-Filho, and Fabricio Villalobos, 2017, “IntegratingSelection, Niche, and Diversification into a Hierarchical ConceptualFramework”,Organisms Diversity & Evolution, 17(1):1–10. doi:10.1007/s13127-016-0299-x
• Anderegg, William R. L., Leander D. L. Anderegg, Kelly L. Kerr,and Anna T. Trugman, 2019, “Widespread Drought-induced TreeMortality at Dry Range Edges Indicates That Climate Stress ExceedsSpecies’ Compensating Mechanisms”,Global ChangeBiology, 25(11): 3793–3802. doi:10.1111/gcb.14771
• Aoki, Kenichi, 1982, “A Condition for Group Selection toPrevail Over Counteracting Individual Selection”,Evolution, 36(4): 832–842. doi:10.2307/2407896
• Arnold, Anthony J. and Kurt Fristrup, 1982, “The Theory ofEvolution by Natural Selection: A Hierarchical Expansion”,Paleobiology, 8(2): 113–129.doi:10.1017/S0094837300004462
• Aspi, Jouni, Anne Jäkäläniemi, Juha Tuomi, andPirkko Siikamäki, 2003, “Multilevel Phenotypic Selection onMorphological Characters in a Metapopulation of SileneTatarica”,Evolution, 57(3): 509–517.doi:10.1111/j.0014-3820.2003.tb01542.x
• Beatty, John, 1980, “What’s Wrong with the ReceivedView of Evolutionary Theory?”,PSA: Proceedings of theBiennial Meeting of the Philosophy of Science Association,1980(2): 397–426.doi:10.1086/psaprocbienmeetp.1980.2.192601
• Benito Garzón, Marta, T. Matthew Robson, and Arndt Hampe,2019, “ΔTrait SDMs: Species Distribution Models ThatAccount for Local Adaptation and Phenotypic Plasticity”,NewPhytologist, 222(4): 1757–1765. doi:10.1111/nph.15716
• Bijma, Piter, 2010a, “Estimating Indirect Genetic Effects:Precision of Estimates and Optimum Designs”,Genetics,186(3): 1013–1028. doi:10.1534/genetics.110.120493
• –––, 2010b, “Multilevel Selection 4:Modeling the Relationship of Indirect Genetic Effects and GroupSize”,Genetics, 186(3): 1029–1031.doi:10.1534/genetics.110.120485
• –––, 2011, “A General Definition of theHeritable Variation That Determines the Potential of a Population toRespond to Selection”,Genetics, 189(4):1347–1359. doi:10.1534/genetics.111.130617
• –––, 2014, “The Quantitative Genetics ofIndirect Genetic Effects: A Selective Review of ModellingIssues”,Heredity, 112(1): 61–69.doi:10.1038/hdy.2013.15
• Bijma, Piter, Andries D. Hulst, and Mart C. M. de Jong, 2022,“The Quantitative Genetics of the Prevalence of InfectiousDiseases: Hidden Genetic Variation Due to Indirect Genetic EffectsDominates Heritable Variation and Response to Selection”,Genetics, 220(1): iyab141. doi:10.1093/genetics/iyab141
• Bijma, Piter, William M. Muir, and Johan A. M. Van Arendonk,2007a, “Multilevel Selection 1: Quantitative Genetics ofInheritance and Response to Selection”,Genetics,175(1): 277–288. doi:10.1534/genetics.106.062711
• Bijma, Piter, William M. Muir, Esther D. Ellen, Jason B. Wolf, andJohan A. M. Van Arendonk, 2007b, “Multilevel Selection 2:Estimating the Genetic Parameters Determining Inheritance and Responseto Selection”,Genetics, 175(1): 289–299.doi:10.1534/genetics.106.062729
• Bijma, Piter and Michael J. Wade, 2008, “The Joint Effectsof Kin, Multilevel Selection and Indirect Genetic Effects on Responseto Genetic Selection”,Journal of Evolutionary Biology,21(5): 1175–1188. doi:10.1111/j.1420-9101.2008.01550.x
• Birch, Jonathan, 2014, “Hamilton’s Rule and ItsDiscontents”,The British Journal for the Philosophy ofScience, 65(2): 381–411. doi:10.1093/bjps/axt016
• –––, 2020, “Kin Selection, GroupSelection, and the Varieties of Population Structure”,TheBritish Journal for the Philosophy of Science, 71(1):259–286. doi:10.1093/bjps/axx028
• Birch, Jonathan and Samir Okasha, 2015, “Kin Selection andIts Critics”,BioScience, 65(1): 22–32.doi:10.1093/biosci/biu196
• Bock, Walter J., 1980, “The Definition and Recognition ofBiological Adaptation”,American Zoologist, 20(1):217–227. doi:10.1093/icb/20.1.217
• Bodmer, Walter F., 1965, “Differential Fertility inPopulation Genetics Models”,Genetics, 51(3):411–424. doi:10.1093/genetics/51.3.411
• Bolker, Jessica A., 2000, “Modularity in Development and WhyIt Matters to Evo-Devo”,American Zoologist, 40(5):770–776. doi:10.1093/icb/40.5.770
• Boorman, Scott A., 1978, “Mathematical Theory of GroupSelection: Structure of Group Selection in Founder PopulationsDetermined from Convexity of the Extinction Operator”,Proceedings of the National Academy of Sciences, 75(4):1909–1913. doi:10.1073/pnas.75.4.1909
• Boorman, Scott A. and Paul R. Levitt, 1973, “Group Selectionon the Boundary of a Stable Population”,TheoreticalPopulation Biology, 4(1): 85–128.doi:10.1016/0040-5809(73)90007-5
• Booth, Austin, 2014, “Symbiosis, Selection, andIndividuality”,Biology & Philosophy, 29(5):657–673. doi:10.1007/s10539-014-9449-8
• Bordenstein, Seth R. and Kevin R. Theis, 2015, “Host Biologyin Light of the Microbiome: Ten Principles of Holobionts andHologenomes”,PLOS Biology, 13(8): e1002226.doi:10.1371/journal.pbio.1002226
• Borregaard, Michael K., Nicholas J. Gotelli, and Carsten Rahbek,2012, “Are Range-Size Distributions Consistent withSpecies-Level Heritability? Range-Size Heritability and Srds”,Evolution, 66(7): 2216–2226.doi:10.1111/j.1558-5646.2012.01581.x
• Borrello, Mark E., 2010,Evolutionary Restraints: TheContentious History of Group Selection, Chicago, IL: Universityof Chicago Press.
• Bouchard, Frédéric, 2009, “UnderstandingColonial Traits Using Symbiosis Research and Ecosystem Ecology”,Biological Theory, 4(3): 240–246.doi:10.1162/biot.2009.4.3.240
• –––, 2010, “Symbiosis, Lateral FunctionTransfer and the (Many) Saplings of Life”,Biology &Philosophy, 25(4): 623–641.doi:10.1007/s10539-010-9209-3
• Bourrat, Pierrick, 2016, “Generalizing ContextualAnalysis”,Acta Biotheoretica, 64(2): 197–217.doi:10.1007/s10441-016-9280-5
• –––, 2019, “Evolutionary Transitions inHeritability and Individuality”,Theory in Biosciences,138(2): 305–323. doi:10.1007/s12064-019-00294-2
• –––, 2021a,Facts, Conventions, and theLevels of Selection, Cambridge/New York: Cambridge UniversityPress. doi:10.1017/9781108885812
• –––, 2021b, “Transitions in Evolution: AFormal Analysis”,Synthese, 198(4): 3699–3731.doi:10.1007/s11229-019-02307-5
• Bourrat, Pierrick and Paul E. Griffiths, 2018, “MultispeciesIndividuals”,History and Philosophy of the LifeSciences, 40(2): article 33. doi:10.1007/s40656-018-0194-1
• Bowles, Samuel and Herbert Gintis, 2011,A CooperativeSpecies: Human Reciprocity and Its Evolution, Princeton,NJ/Oxford: Princeton University Press.
• Boyd, Lawrence H. and Gudmund R. Iversen, 1979,ContextualAnalysis: Concepts and Statistical Techniques, Belmont, CA:Wadsworth.
• Brandon, Robert N., 1978, “Adaptation and EvolutionaryTheory”,Studies in History and Philosophy of Science PartA, 9(3): 181–206. doi:10.1016/0039-3681(78)90005-5
• –––, 1981, “Biological Teleology:Questions and Explanations”,Studies in History andPhilosophy of Science Part A, 12(2): 91–105.doi:10.1016/0039-3681(81)90015-7
• –––, 1982, “The Levels ofSelection”,PSA: Proceedings of the Biennial Meeting of thePhilosophy of Science Association, 1982(1): 315–323.doi:10.1086/psaprocbienmeetp.1982.1.192676
• –––, 1985, “Adaptation Explanations: AreAdaptations for the Good of Replicators or Interactors?”, inEvolution at a Crossroads: The New Biology and the New Philosophyof Science, David J. Depew & Bruce H. Weber (eds.),Cambridge, MA: MIT Press/Bradford, pp. 81–96.
• –––, 1988, “Levels of Selection: AHierarchy of Interactors”, inThe Role of Behavior inEvolution, H.C. Plotkin (ed.), Cambridge, MA: MIT Press, pp.51–71.
• –––, 1990,Adaptation and Environment,Princeton, NJ: Princeton University Press.
• –––, 1999, “The Units of SelectionRevisited: The Modules of Selection”,Biology &Philosophy, 14(2): 167–180.doi:10.1023/A:1006682200831
• Brandon, Robert N. and Richard M. Burian (eds.), 1984,Genes,Organisms, Populations: Controversies over the Units ofSelection, Cambridge, MA: MIT Press.
• Brandon, Robert N. and H. Frederik Nijhout, 2006, “TheEmpirical Nonequivalence of Genic and Genotypic Models of Selection: A(Decisive) Refutation of Genic Selectionism and Pluralistic GenicSelectionism”,Philosophy of Science, 73(3):277–297. doi:10.1086/515416
• Breden, Felix and Michael J. Wade, 1981, “Inbreeding andEvolution by Kin Selection”,Ethology and Sociobiology,2(1): 3–16. doi:10.1016/0162-3095(81)90018-2
• –––, 1989, “Selection Within and betweenKin Groups of the Imported Willow Leaf Beetle”,The AmericanNaturalist, 134(1): 35–50. doi:10.1086/284964
• –––, 1991, “‘Runaway’ SocialEvolution: Reinforcing Selection for Inbreeding and Altruism”,Journal of Theoretical Biology, 153(3): 323–337.doi:10.1016/S0022-5193(05)80573-9
• Brucker, Robert M. and Seth R. Bordenstein, 2012,“Speciation by Symbiosis”,Trends in Ecology &Evolution, 27(8): 443–451.doi:10.1016/j.tree.2012.03.011
• –––, 2013a, “The CapaciousHologenome”,Zoology, 116(5): 260–261.doi:10.1016/j.zool.2013.08.003
• –––, 2013b, “The Hologenomic Basis ofSpeciation: Gut Bacteria Cause Hybrid Lethality in the GenusNasonia”,Science, 341(6146): 667–669.doi:10.1126/science.1240659
• Bueno, Otávio, Ruey-lin Chen, and Melinda Bonnie Fagan(eds.), 2018,Individuation, Process, and ScientificPractices, New York: Oxford University Press.doi:10.1093/oso/9780190636814.001.0001
• Burian, Richard M., 1983, “‘Adaptation’”,inDimensions of Darwinism: Themes and Counterthemes in TwentiethCentury Evolutionary Theory, Marjorie Grene (ed.), Cambridge:Cambridge University Press, pp. 287–314.
• –––, 1992, “Adaptation: HistoricalPerspectives”, in Keller and Lloyd 1992: 7–12.
• Burt, Austin and Robert Trivers, 2006,Genes in Conflict: TheBiology of Selfish Genetic Elements, Cambridge, MA: Belknap Pressof Harvard University Press.
• Buss, Leo W., 1987,The Evolution of Individuality,Princeton, NJ: Princeton University Press.
• Calcott, Brett, 2008, “The Other Cooperation Problem:Generating Benefit”,Biology & Philosophy, 23(2):179–203. doi:10.1007/s10539-007-9095-5
• Carotenuto, Francesco, Carmela Barbera, and Pasquale Raia, 2010,“Occupancy, Range Size, and Phylogeny in Eurasian Pliocene toRecent Large Mammals”,Paleobiology, 36(3):399–414. doi:10.1666/09059.1
• Carotenuto, Francesco, Mirko Di Febbraro, Alessandro Mondanaro,Silvia Castiglione, Carmela Serio, Marina Melchionna, Lorenzo Rook,and Pasquale Raia, 2020, “MInOSSE: A New Method to ReconstructGeographic Ranges of Fossil Species”,Methods in Ecology andEvolution, 11(9): 1121–1132.doi:10.1111/2041-210X.13423
• Cassidy, John, 1978, “Philosophical Aspects of the GroupSelection Controversy”,Philosophy of Science, 45(4):575–594. doi:10.1086/288836
• Cheng, H. and W. M. Muir, 2005, “The Effects of GeneticSelection for Survivability and Productivity on Chicken PhysiologicalHomeostasis”,World’s Poultry Science Journal,61(3): 383–397.
• Chiu, Lynn and Scott F. Gilbert, 2015, “The Birth of theHolobiont: Multi-species Birthing Through Mutual Scaffolding and NicheConstruction”,Biosemiotics, 8: 191-210.
• –––, 2020, “Niche Construction and theTransition to Herbivory: Phenotype Switching and the Organization ofNew Nutritional Modes”, inPhenotypic Switching:Implications in Biology and Medicine, Herbert Levine, Mohit KumarJolly, Prakesh Kulkarni, and Vidyanand Nanjundiah (eds.), London:Academic Press, 459–482.doi:10.1016/B978-0-12-817996-3.00015-3
• Clarke, Ellen, 2013, “The Multiple Realizability ofBiological Individuals”,Journal of Philosophy, 110(8):413–435. doi:10.5840/jphil2013110817
• –––, 2016, “A Levels-of-Selection Approachto Evolutionary Individuality”,Biology &Philosophy, 31(6): 893–911.doi:10.1007/s10539-016-9540-4
• –––, 2018, “Adaptation, MultilevelSelection, and Organismality: A Clash of Perspectives”, inThe Routledge Handbook of Evolution and Philosophy, RichardJoyce (ed.), (Routledge Handbooks in Philosophy), New York: Routledge,Taylor & Francis Group, chap. 3.
• Colwell, Robert K., 1981, “Group Selection Is Implicated inthe Evolution of Female-Biased Sex Ratios”,Nature,290(5805): 401–404. doi:10.1038/290401a0
• Costello, Elizabeth K., Keaton Stagaman, Les Dethlefsen, BrendanJ. M. Bohannan, and David A. Relman, 2012, “The Application ofEcological Theory Toward an Understanding of the HumanMicrobiome”,Science, 336(6086): 1255–1262.doi:10.1126/science.1224203
• Coyne, Jerry A. and H. Allen Orr, 2004,Speciation,Sunderland, MA: Sinauer Associates.
• Coyte, Katharine Z., Jonas Schluter, and Kevin R. Foster, 2015,“The Ecology of the Microbiome: Networks, Competition, andStability”,Science, 350(6261): 663–666.doi:10.1126/science.aad2602
• Cracraft, Joel, 1985, “Species Selection, MacroevolutionaryAnalysis, and the ‘Hierarchical Theory’ ofEvolution”,Systematic Biology, 34(2): 222–229.doi:10.2307/sysbio/34.2.222
• Craig, David M., 1982, “Group Selection Versus IndividualSelection: An Experimental Analysis”,Evolution, 36(2):271–282. doi:10.2307/2408045
• Craig, J.V. and W.M. Muir, 1996, “Group Selection forAdaptation to Multiple-Hen Cages: Behavioral Responses”,Poultry Science, 75(10): 1145–1155.doi:10.3382/ps.0751145
• Crossa, José, Paulino Pérez-Rodríguez, JaimeCuevas, Osval Montesinos-López, Diego Jarquín, Gustavode los Campos, Juan Burgueño, Juan M. González-Camacho,Sergio Pérez-Elizalde, Yoseph Beyene, Susanne Dreisigacker,Ravi Singh, Xuecai Zhang, Manje Gowda, Manish Roorkiwal, JessicaRutkoski, and Rajeev K. Varshney, 2017, “Genomic Selection inPlant Breeding: Methods, Models, and Perspectives”,Trendsin Plant Science, 22(11): 961–975.doi:10.1016/j.tplants.2017.08.011
• Crow, J. F. and K. Aoki, 1982, “Group Selection for aPolygenic Behavioral Trait: A Differential Proliferation Model”,Proceedings of the National Academy of Sciences, 79(8):2628–2631. doi:10.1073/pnas.79.8.2628
• Cummins, Robert, 1975, “Functional Analysis”,TheJournal of Philosophy, 72(20): 741–764.doi:10.2307/2024640
• Damuth, John and I. Lorraine Heisler, 1988, “AlternativeFormulations of Multilevel Selection”,Biology &Philosophy, 3(4): 407–430. doi:10.1007/BF00647962
• Darwin, Charles, 1859,On the Origin of Species, London:John Murray.
• Davidson, Diane W., Alexey Kopchinskiy, Kamariah Abu Salim, MaricaGrujic, Linda Lim, Chan Chin Mei, Tappey H. Jones, Dale Casamatta, LeaAtanasova, and Irina S. Druzhinina, 2016, “Nutrition ofBorneo’s ‘exploding’ Ants (Hymenoptera: Formicidae:Colobopsis) A Preliminary Assessment”,Biotropica48(4): 518–527.
• Dawkins, Richard, 1976,The Selfish Gene, Oxford: OxfordUniversity Press.
• –––, 1982a, “Replicators andVehicles”, inCurrent Problems in Sociobiology,King’s College (University of Cambridge) (ed.), Cambridge/NewYork: Cambridge University Press, 45–64. [Dawkins 1982 available online]
• –––, 1982b,The Extended Phenotype: The Geneas the Unit of Selection, Oxford: Oxford University Press.
• De Mazancourt, Claire, Michel Loreau, and Ulf Dieckmann, 2005,“Understanding Mutualism When There Is Adaptation to thePartner”,Journal of Ecology, 93(2): 305–314.doi:10.1111/j.0022-0477.2004.00952.x
• De Lisle, Stephen P., Daniel I. Bolnick, Edmund D. Brodie, AllenJ. Moore, and Joel W. McGlothlin, 2022, “Interacting Phenotypesand the Coevolutionary Process: Interspecific Indirect Genetic EffectsAlter Coevolutionary Dynamics”,Evolution, 76(3):429–444. doi:10.1111/evo.14427
• Dobzhansky, Theodosius, 1956, “What Is an AdaptiveTrait?”,The American Naturalist, 90(855):337–347. doi:10.1086/281944
• –––, 1968, “Adaptedness andFitness”, inPopulation Biology and Evolution, RichardC. Lewontin (ed.), Syracuse, NY: Syracuse University Press, pp.109–121.
• Doolittle, W. Ford and Austin Booth, 2017, “It’s theSong, Not the Singer: An Exploration of Holobiosis and EvolutionaryTheory”,Biology & Philosophy, 32(1): 5–24.doi:10.1007/s10539-016-9542-2
• Doolittle, W. Ford and S. Andrew Inkpen, 2018, “Processesand Patterns of Interaction as Units of Selection: An Introduction toITSNTS Thinking”,Proceedings of the National Academy ofSciences, 115(16): 4006–4014.doi:10.1073/pnas.1722232115
• Douglas, Angela E. and John H. Werren, 2016, “Holes in theHologenome: Why Host-Microbe Symbioses Are Not Holobionts”,MBio, 7(2): e02099-15. doi:10.1128/mBio.02099-15
• Drown, Devin M. and Michael J. Wade, 2014, “RunawayCoevolution: Adaptation to Heritable and Nonheritable Environments:Brief Communication”,Evolution, 68(10):3039–3046. doi:10.1111/evo.12470
• Dugatkin, Lee Alan and Hudson Kern Reeve, 1994, “BehavioralEcology and Levels of Selection: Dissolving the Group SelectionControversy”, inAdvances in the Study of Behavior 23,Peter Slater, Jay S. Rosenblatt, Charles T. Snowdon, and ManfredMilinski (eds.), New York: Academic Press, 101–133.doi:10.1016/S0065-3454(08)60352-6
• Dunbar, R. I. M., 1982, “Adaptation, Fitness and theEvolutionary Tautology”, inCurrent Problems inSociobiology, King’s College (University of Cambridge)(ed.), Cambridge/New York: Cambridge University Press,9–28.
• Dupré, John, 2010, “The Polygenomic Organism”,The Sociological Review, 58(supplement 1): 19–31.doi:10.1111/j.1467-954X.2010.01909.x
• –––, 2012a,Processes of Life: Essays in thePhilosophy of Biology, Oxford: Oxford University Press.doi:10.1093/acprof:oso/9780199691982.001.0001
• –––, 2012b, “Postgenomic Darwinism”,in Dupré 2012a: 143–160 (ch. 9).
• –––, 2020, “Life as Process”,Epistemology & Philosophy of Science, 57(2):96–113. doi:10.5840/eps202057224
• –––, 2021a,The Metaphysics of Biology,Cambridge/New York: Cambridge University Press.doi:10.1017/9781009024297
• –––, 2021b, “Causally PowerfulProcesses”,Synthese, 199(3–4):10667–10683. doi:10.1007/s11229-021-03263-9
• Dupré, John and Maureen A. O’Malley, 2007,“Metagenomics and Biological Ontology”,Studies inHistory and Philosophy of Science Part C: Studies in History andPhilosophy of Biological and Biomedical Sciences, 38(4):834–846. doi:10.1016/j.shpsc.2007.09.001\
• –––, 2013, “Varieties of Living Things:Life at the Intersection of Lineage and Metabolism”, inVitalism and the Scientific Image in Post-Enlightenment LifeScience, 1800-2010, Sebastian Normandin and Charles T. Wolfe(eds.), (History, Philosophy and Theory of the Life Sciences 2),Dordrecht: Springer Netherlands, 311–343.doi:10.1007/978-94-007-2445-7_13
• Dupressoir, A., C. Lavialle, and T. Heidmann, 2012, “FromAncestral Infectious Retroviruses to Bona Fide Cellular Genes: Role ofthe CapturedSyncytins in Placentation”,Placenta, 33(9): 663–671.doi:10.1016/j.placenta.2012.05.005
• Dynesius, Mats and Roland Jansson, 2014, “Persistence ofWithin-Species Lineages: A Neglected Control of Speciation Rates:Perspective”,Evolution, 68(4): 923–934.doi:10.1111/evo.12316
• Earnshaw, Eugene, 2015, “Group Selection and ContextualAnalysis”,Synthese, 192(1): 302–316.doi:10.1007/s11229-014-0569-0
• Edenbrow, Mat, Bronwyn H. Bleakley, Safi K. Darden, Charles R.Tyler, Indar W. Ramnarine, and Darren P. Croft, 2017, “TheEvolution of Cooperation: Interacting Phenotypes among SocialPartners”,The American Naturalist, 189(6):630–643. doi:10.1086/691386
• Eldakar, Omar Tonsi, David Sloan Wilson, Michael J. Dlugos, andJohn W. Pepper, 2010, “The Role of Multilevel Selection in theEvolution of Sexual Conflict in the Water StriderAquariusremigis: Role of Multilevel Selection in Sexual Conflict”,Evolution, 64(11): 3183–3189.doi:10.1111/j.1558-5646.2010.01087.x
• Eldredge, Niles, 1985,Unfinished Synthesis: BiologicalHierarchies and Modern Evolutionary Thought, New York: OxfordUniversity Press.
• Ellen, Esther D., T. Bas Rodenburg, Gerard A. A. Albers, J.Elizabeth Bolhuis, Irene Camerlink, Naomi Duijvesteijn, Egbert F.Knol, William M. Muir, Katrijn Peeters, Inonge Reimert, EwaSell-Kubiak, Johan A. M. van Arendonk, Jeroen Visscher, and PiterBijma, 2014, “The Prospects of Selection for Social GeneticEffects to Improve Welfare and Productivity in Livestock”,Frontiers in Genetics, 5(November): article 377.doi:10.3389/fgene.2014.00377
• Ellen, Esther D., Jeroen Visscher, Johan A. M. van Arendonk, andPiter Bijma, 2008, “Survival of Laying Hens: Genetic Parametersfor Direct and Associative Effects in Three Purebred LayerLines”,Poultry Science, 87(2): 233–239.doi:10.3382/ps.2007-00374
• Ellis, Bruce J. and Timothy Ketelaar, 2002, “COMMENTARY:Clarifying the Foundations of Evolutionary Psychology: A Reply toLloyd and Feldman”,Psychological Inquiry, 13(2):157–164. doi:10.1207/S15327965PLI1302_05
• Ereshefsky, Marc and Makmiller Pedroso, 2013, “BiologicalIndividuality: The Case of Biofilms”,Biology &Philosophy, 28(2): 331–349.doi:10.1007/s10539-012-9340-4
• Erwin, Douglas H., 2000, “Macroevolution Is More thanRepeated Rounds of Microevolution”,Evolution andDevelopment, 2(2): 78–84.doi:10.1046/j.1525-142x.2000.00045.x
• Eschel, Ilan, 1972, “On the Neighbor Effect and theEvolution of Altruistic Traits”,Theoretical PopulationBiology 3(3): 258–277.doi:10.1016/0040-5809(72)90003-2
• Feldman, Marcus W., Freddy B. Christiansen, and Uri Liberman,1983, “On Some Models of Fertility Selection”,Genetics, 105(4): 1003–1010.doi:10.1093/genetics/105.4.1003
• Fenner, Frank, 1965, “Myxoma Virus andOryctolaguscuniculus”, inThe Genetics of Colonizing Species:Proceedings of the First International Union of Biological SciencesSymposia on General Biology, H. G. Baker and G. Ledyard Stebbins(eds.), New York: Academic Press, 485–501.
• Fenner, Frank and I. D. Marshall, 1957, “A Comparison of theVirulence for European Rabbits ( Oryctolagus Cuniculus ) of Strains ofMyxoma Virus Recovered in the Field in Australia, Europe andAmerica”,Journal of Hygiene, 55(2): 149–191.doi:10.1017/S0022172400037098
• Fenner, Frank and Gwendolyn M. Woodroofe, 1965, “Changes inthe Virulence and Antigenic Structure of Strains of Myxoma VirusRecovered from Australian Wild Rabbits between 1950 and 1964”,Australian Journal of Experimental Biology and MedicalScience, 43(S1): 359–370. doi:10.1038/icb.1965.69
• Fewell, J. H. and R.E. Page Jr, 2000, “Colony-LevelSelection Effects on Individual and Colony Foraging Task Performancein Honeybees, Apis Mellifera L”,Behavioral Ecology andSociobiology, 48(3): 173–181.doi:10.1007/s002650000183
• Fisher, D. N. and A. G. McAdam, 2017, “Social Traits, SocialNetworks and Evolutionary Biology”,Journal of EvolutionaryBiology, 30(12): 2088–2103. doi:10.1111/jeb.13195
• Fisher, Ronald Aylmer, 1930 [1958],The Genetical Theory ofNatural Selection, Oxford: The Clarendon Press. Revised edition1958, New York: Dover Publications.
• Forsdyke, Donald R., 2010, “The Selfish Gene Revisited:Reconciliation of Williams-Dawkins and ConventionalDefinitions”,Biological Theory, 5(3): 246–255.doi:10.1162/BIOT_a_00054
• Fortunato, Laura, and Marco Archetti, 2010, “Evolution ofmonogamous marriage by maximization of inclusive fitness”,Journal of Evolutionary Biology, 23(1):149–156.
• Foster, Kevin R., 2004, “Diminishing Returns in SocialEvolution: The not-so-tragic commons”,Journal ofEvolutionary Biology 17: 1058–1072.
• –––, 2011, “The Sociobiology of MolecularSystems”,Nature Reviews Genetics, 12(3):193–203. doi:10.1038/nrg2903
• Foster, Kevin R., Jonas Schluter, Katharine Z. Coyte, and SethRakoff-Nahoum, 2017, “The Evolution of the Host Microbiome as anEcosystem on a Leash”,Nature, 548(7665): 43–51.doi:10.1038/nature23292
• Foster, K. R. and T. Wenseleers, 2006, “A General Model forthe Evolution of Mutualisms”,Journal of EvolutionaryBiology, 19(4): 1283–1293.doi:10.1111/j.1420-9101.2005.01073.x
• Fortunato, Laura and Marco Archetti, 2010, “Evolution ofMonogamous Marriage by Maximization of Inclusive Fitness”,Journal of Evolutionary Biology, 23(1): 149–156.doi:10.1111/j.1420-9101.2009.01884.x
• Frank, S.A., 2013, “Natural Selection. VIII. History andInterpretation of Kin Selection Theory”,Journal ofEvolutionary Biology, 26(6): 1151–1184.doi:10.1111/jeb.12131
• Franklin, Ian and R. C. Lewontin, 1970, “Is the Gene theUnit of Selection?”,Genetics, 65(4): 707–734.doi:10.1093/genetics/65.4.707
• Freedberg, Steven, Caroline Urban, and Brianna M. Cunniff, 2021,“Dispersal Reduces Interspecific Competitiveness by SpreadingLocally Harmful Traits”,Journal of EvolutionaryBiology, 34(9): 1477–1487. doi:10.1111/jeb.13912
• Funkhouser, Lisa J. and Seth R. Bordenstein, 2013, “MomKnows Best: The Universality of Maternal MicrobialTransmission”,PLOS Biology, 11(8): e1001631.doi:10.1371/journal.pbio.1001631
• Fusco, Giuseppe (ed.), 2019,Perspectives on Evolutionary andDevelopmental Biology: Essays for Alessandro Minelli, Padua:Padova University Press.
• Futuyma, Douglas J., 1979,Evolutionary Biology,Sunderland, MA: Sinauer Associates.
• Futuyma, Douglas J. and Mark Kirkpatrick, 2017,Evolution, fourth edition, Sunderland, MA: SinauerAssociates.
• Gannett, Lisa, 1999, “What’s in a Cause? The PragmaticDimensions of Genetic Explanations”,Biology andPhilosophy 14(3): 349–374. doi:10.1023/A:1006583215835
• Gardner, Andy, 2015a, “The Genetical Theory of MultilevelSelection”,Journal of Evolutionary Biology, 28(2):305–319. doi:10.1111/jeb.12566
• –––, 2015b, “Group Selection versus GroupAdaptation”,Nature, 524(7566): E3-E4.
• –––, 2015c, “More on the Genetical Theoryof Multilevel Selection”,Journal of EvolutionaryBiology, 28(9): 1747–1751.
• Gardner, Andy and A. Grafen, 2009, “Capturing theSuperorganism: A Formal Theory of Group Adaptation”,Journalof Evolutionary Biology, 22(4): 659–671.doi:10.1111/j.1420-9101.2008.01681.x
• Gardner, Andy and J. J. Welch, 2011, “A Formal Theory of theSelfish Gene”,Journal of Evolutionary Biology, 24(8):1801–1813. doi:10.1111/j.1420-9101.2011.02310.x
• Gardner, Andy, S. A. West, and G. Wild, 2011, “The GeneticalTheory of Kin Selection”,Journal of EvolutionaryBiology, 24(5): 1020–1043.doi:10.1111/j.1420-9101.2011.02236.x
• Ghiselin, Michael T., 1974,The Economy of Nature and theEvolution of Sex, Berkeley, CA, University of CaliforniaPress.
• Gilbert, Scott F., 2011, “Symbionts as Genetic Sources ofHeritable Variation”, inTransformations of Lamarckism: FromSubtle Fluids to Molecular Biology (Vienna Series in TheoreticalBiology), Snait B. Gissis & Eva Jablonka (eds.), Cambridge, MA:MIT Press, pp. 283–293.doi:10.7551/mitpress/9780262015141.003.0027
• –––, 2014, “A Holobiont Birth Narrative:The Epigenetic Transmission of the Human Microbiome”,Frontiers in Genetics, 5(August): article 282.doi:10.3389/fgene.2014.00282
• –––, 2018, “Health Fetishism among theNacirema: A Fugue on Jenny Reardon’s The Postgenomic Condition:Ethics, Justice, and Knowledge after the Genome (Chicago UniversityPress, 2017) and Isabelle Stengers’ Another Science Is Possible:A Manifesto for Slow Science (Polity Press, 2018)”,Organisms. Journal of Biological Sciences, 2(1): 43–54.doi:10.13133/2532-5876_3.11
• –––, 2019a, “Evolutionary TransitionsRevisited: Holobiont Evo-devo”,Journal of ExperimentalZoology Part B: Molecular and Developmental Evolution, 332(8):307–314. doi:10.1002/jez.b.22903
• –––, 2019b, “Towards A DevelopmentalBiology Of Holobionts”, in Fusco 2019: 13–22.
• –––, 2020a, “Developmental SymbiosisFacilitates the Multiple Origins of Herbivory”,Evolutionand Development, 22(1–2): 154–164.doi:10.1111/ede.12291
• –––, 2020b, “Holobionts Can Evolve ByChanging Their Symbionts And Hosts”, inFeral Atlas: TheMore-Than-Human Anthropocene, Anna L. Tsing, Jennifer Deger,Alder Keleman Saxena, and Feifei Zhou (eds.), Stanford, CA: StanfordUniversity Press.
• Gilbert, Scott F., Thomas C. G. Bosch, and CristinaLedón-Rettig, 2015, “Eco-Evo-Devo: DevelopmentalSymbiosis and Developmental Plasticity as Evolutionary Agents”,Nature Reviews Genetics, 16(10): 611–622.doi:10.1038/nrg3982
• Gilbert, Scott F., Eugene Rosenberg, and Ilana Zilber-Rosenberg,2017, “The Holobiont with Its Hologenome Is a Level of Selectionin Evolution”, in Gissis, Lamm, and Shavit 2017:305–324.
• Gilbert, Scott F., Jan Sapp, and Alfred I. Tauber, 2012, “ASymbiotic View of Life: We Have Never Been Individuals”,TheQuarterly Review of Biology, 87(4): 325–341.doi:10.1086/668166
• Gilbert, Scott F. and Alfred I. Tauber, 2016, “RethinkingIndividuality: The Dialectics of the Holobiont”,Biology& Philosophy, 31(6): 839–853.doi:10.1007/s10539-016-9541-3
• Gilpin, Michael E., 1975,Group Selection in Predator-PreyCommunities, (Monographs in Population Biology 9), Princeton, NJ:Princeton University Press.
• Gissis, Snait, Ehud Lamm, and Ayelet Shavit (eds.), 2017,Landscapes of Collectivity in the Life Sciences, (ViennaSeries in Theoretical Biology), Cambridge, MA: The MIT Press.
• Glennan, Stuart, 2002, “Contextual Unanimity and the Unitsof Selection Problem”,Philosophy of Science, 69(1):118–137. doi:10.1086/338944
• Glymour, Bruce, 1999, “Population Level Causation and aUnified Theory of Natural Selection”,Biology andPhilosophy, 14(4): 521–536.doi:10.1023/A:1006516232674
• Godfrey-Smith, Peter, 1992, “Additivity and the Units ofSelection”,PSA: Proceedings of the Biennial Meeting of thePhilosophy of Science Association, 1992(1): 315–328.doi:10.1086/psaprocbienmeetp.1992.1.192764
• –––, 2007, “The Strategy of Model-BasedScience”,Biology & Philosophy, 21(5):725–740. doi:10.1007/s10539-006-9054-6
• –––, 2009,Darwinian Populations and NaturalSelection, New York: Oxford University Press.doi:10.1093/acprof:osobl/9780199552047.001.0001
• –––, 2011, “Agents and Acacias: Replies toDennett, Sterelny, and Queller”,Biology andPhilosophy, 26(4): 501–515.doi:10.1007/s10539-011-9246-6
• –––, 2013, “Darwinian Individuals”,inFrom Groups to Individuals: Evolution and EmergingIndividuality, Frédéric Bouchard and PhilippeHuneman (eds.), Cambridge, MA: The MIT Press, 17–36 (ch. 1).doi:10.7551/mitpress/8921.003.0005
• –––, 2016, “Complex Life Cycles and theEvolutionary Process”,Philosophy of Science, 83(5):816–827. doi:10.1086/687866
• Godfrey-Smith, Peter and Benjamin Kerr, 2002, “Group Fitnessand Multi-Level Selection: Replies to Commentaries”,Biology& Philosophy, 17(4): 539–549.doi:10.1023/A:1020583723772
• Godfrey-Smith, Peter and Richard Lewontin, 1993, “TheDimensions of Selection”,Philosophy of Science, 60(3):373–395. doi:10.1086/289742
• Goodnight, Charles J., 1985, “The Influence of EnvironmentalVariation on Group and Individual Selection in a Cress”,Evolution, 39(3): 545–558.doi:10.1111/j.1558-5646.1985.tb00394.x
• –––, 2000, “Heritability at the EcosystemLevel”,Proceedings of the National Academy ofSciences, 97(17): 9365–9366.doi:10.1073/pnas.97.17.9365
• –––, 2012, “Wright’s ShiftingBalance Theory and Factors Affecting the Probability of PeakShifts”, inThe Adaptive Landscape in EvolutionaryBiology, Erik Svensson and Ryan Calsbeek (eds.), Oxford: OxfordUniversity Press, 74–86.doi:10.1093/acprof:oso/9780199595372.003.0006
• –––, 2013, “On Multilevel Selection andKin Selection: Contextual Analysis Meets Direct Fitness”,Evolution, 67(6): 1539–1548.doi:10.1111/j.1558-5646.2012.01821.x
• –––, 2015, “Multilevel Selection Theoryand Evidence: A Critique of Gardner, 2015”,Journal ofEvolutionary Biology, 28(9): 1734–1746.doi:10.1111/jeb.12685
• –––, 2016, “On the Effectiveness ofMultilevel Selection”,Behavioral and Brain Sciences,39: e99. doi:10.1017/S0140525X15001053
• –––, 2020, “The Theory of MultilevelSelection”, inThe Theory of Evolution: Principles,Concepts, and Assumptions, Samuel M. Scheiner and David P.Mindell (eds.), Chicago/London: The University of Chicago Press,194–210.
• Goodnight, Charles J., James M. Schwartz, and Lori Stevens, 1992,“Contextual Analysis of Models of Group Selection, SoftSelection, Hard Selection, and the Evolution of Altruism”,The American Naturalist, 140(5): 743–761.doi:10.1086/285438
• Goodnight, Charles J. and Lori Stevens, 1997, “ExperimentalStudies of Group Selection: What Do They Tell US About Group Selectionin Nature?”,The American Naturalist, 150(S1):S59–S79. doi:10.1086/286050
• Gordon, Deborah M., 2014, “The Ecology of CollectiveBehavior”,PLoSBiology, 12(3): e1001805.doi:10.1371/journal.pbio.1001805
• Gould, Stephen Jay, 1977, “Caring Groups and SelfishGenes”,Natural History, 86(10): 20–24. Reprintedin hisThe Panda’s Thumb, New York: Norton, 1980,85–92.
• –––, 2002,The Structure of EvolutionaryTheory, Cambridge, MA: Belknap Press of Harvard UniversityPress.
• Gould, Stephen J. and Richard C. Lewontin, 1979, “TheSpandrels of San Marco and the Panglossian Paradigm: A Critique of theAdaptationist Programme”,Proceedings of the Royal Societyof London. Series B. Biological Sciences, 205(1161):581–598. doi:10.1098/rspb.1979.0086
• Gould, Stephen Jay and Elisabeth A. Lloyd, 1999,“Individuality and Adaptation across Levels of Selection: HowShall We Name and Generalize the Unit of Darwinism?”,Proceedings of the National Academy of Sciences, 96(21):11904–11909. doi:10.1073/pnas.96.21.11904
• Gould, Stephen Jay and Elisabeth S. Vrba, 1982,“Exaptation—a Missing Term in the Science of Form”,Paleobiology, 8(1): 4–15.doi:10.1017/S0094837300004310
• Grafen, Alan,
• –––, 1984, “Natural Selection, KinSelection, and Group Selection”,Behavioural Ecology: AnEvolutionary Approach, 2nd ed., JR Krebs and NB Davies (eds),Oxford/New York: Oxford University Press, 62–84.
• –––, 1985a, “A Geometric View ofRelatedness”,Oxford Surveys in Evolutionary Biology, Volume2, Richard Dawkins and Mark Ridley (eds), Oxford/New York: OxfordUniversity Press, 28–89.
• –––, 1985b, “Evolutionary Theory:Hamilton’s Rule OK”,Nature, 318(6044):310–311. doi:10.1038/318310a0
• –––, 2006, “Optimization of InclusiveFitness”,Journal of Theoretical Biology, 238(3):541–563. doi:10.1016/j.jtbi.2005.06.009
• –––, 2009, “Formalizing Darwinism andInclusive Fitness Theory”,Philosophical Transactions of theRoyal Society B: Biological Sciences364 (1533): 3135–3141.
• Grant, B. Rosemary and Peter R. Grant, 1989,EvolutionaryDynamics of a Natural Population: The Large Cactus Finch of theGalápagos, Chicago, IL: University of Chicago Press.
• Grant, Peter R., 1986,Ecology and Evolution of Darwin’sFinches, Princeton, NJ: Princeton University Press. Reprinted1999.
• Grantham, Todd A., 1994, “Putting the Cart Back behind theHorse: Group Selection Does Not Require That Groups Be‘Organisms’”,Behavioral and BrainSciences, 17(4): 622–623.doi:10.1017/S0140525X0003627X
• –––, 1995, “Hierarchical Approaches toMacroevolution: Recent Work on Species Selection and the ‘EffectHypothesis’”,Annual Review of Ecology andSystematics, 26: 301–321.doi:10.1146/annurev.es.26.110195.001505
• Griesemer, James R., 1984, “Presentations and the Status ofTheories”,PSA: Proceedings of the Biennial Meeting of thePhilosophy of Science Association, 1984(1): 102–114.doi:10.1086/psaprocbienmeetp.1984.1.192331 [Griesemer 1984 available online]
• –––, 1998, “The Case for EpigeneticInheritance in Evolution”,Journal of EvolutionaryBiology, 11(2): 193–200.doi:10.1046/j.1420-9101.1998.11020193.x
• –––, 2000a, “The Units of EvolutionaryTransition”,Selection, 1(1–3): 67–80.doi:10.1556/Select.1.2000.1-3.7 [Griesemer 2000a available online]
• –––, 2000b, “Development, Culture, and theUnits of Inheritance”,Philosophy of Science, 67(S3):S348–S368. doi:10.1086/392831 [Griesemer 2000b available online]
• –––, 2000c, “Reproduction and theReduction of Genetics”, inThe Concept of the Gene inDevelopment and Evolution, Peter J. Beurton, Raphael Falk, andHans-Jörg Rheinberger (eds.), Cambridge: Cambridge UniversityPress, 240–285. doi:10.1017/CBO9780511527296.013 [Griesemer 2000c available online]
• –––, 2003, “The Philosophical Significanceof Gánti′s Work”, inThe Principles ofLife, by Tibor Ganti, Eors Szathmáry and James Griesemer(eds.), Oxford/New York: Oxford University Press, 169–186.doi:10.1093/acprof:oso/9780198507260.003.0005
• –––, 2005, “The Informational Gene and theSubstantial Body: On the Generalization of Evolutionary Theory byAbstraction”, inIdealization XII: Correcting the Model.Idealization and Abstraction in the Sciences, Nancy Cartwrightand Martin R. Jones (eds.), (Poznan Studies), Amsterdam: RodopiPublishers, 59–115. [Griesemer 2005 available online]
• –––, 2013, “Formalization and the Meaningof ‘Theory’ in the Inexact Biological Sciences”,Biological Theory, 7(4): 298–310.doi:10.1007/s13752-012-0065-z
• –––, 2014, “Reproduction and ScaffoldedDevelopmental Processes: An Integrated EvolutionaryPerspective”, inTowards a Theory of Development,Alessandro Minelli and Thomas Pradeu (eds), Oxford University Press,pp. 183–202 (ch. 12).doi:10.1093/acprof:oso/9780199671427.003.0012 [Griesemer 2014 available online]
• –––, 2016, “Reproduction in Complex LifeCycles: Toward a Developmental Reason Norms Perspective”,Philosophy of Science, 83(5): 803–815.doi:10.1086/687865
• –––, 2018, “Individuation of DevelopmentalSystems: A Reproducer Perspective”, in Bueno, Chen, and Fagan2018: 137–164 (ch. 7). [Griesemer 2018 available online]
• –––, 2019, “Towards a Theory of ExtendedDevelopment”, in Fusco 2019: 319–334. [Griesemer 2019 available online]
• Griesemer, James R. and Michael J. Wade, 1988, “LaboratoryModels, Causal Explanation and Group Selection”,Biology& Philosophy, 3(1): 67–96. doi:10.1007/BF00127629 [Griesemer and Wade 1988 available online]
• –––, 2000, “Populational Heritability:Extending Punnett Square Concepts to Evolution at the MetapopulationLevel”,Biology & Philosophy, 15(1): 1–17.doi:10.1023/A:1006501313374 [Griesemer and Wade 2000 available online]
• Griesemer, James R. and William C. Wimsatt, 1989, “PicturingWeismannism: A Case Study of Conceptual Evolution”, inWhatthe Philosophy of Biology Is: Essays for David Hull, Michael Ruse(ed.), (Nijhoff International Philosophy Series 32), Dordrecht:Kluwer, 75–137. doi:10.1007/978-94-009-1169-7_6
• Griffing, B, 1967, “Selection in Reference to BiologicalGroups I. Individual and Group Selection Applied to Populations ofUnordered Groups”,Australian Journal of BiologicalSciences, 20(1): 127–139. doi:10.1071/BI9670127
• Griffiths, Paul E. and Russell D. Gray, 1994, “DevelopmentalSystems and Evolutionary Explanation”,Journal ofPhilosophy, 91(6): 277–304. doi:10.2307/2940982
• –––, 1997, “Replicator II –Judgement Day”,Biology & Philosophy, 12(4):471–492. doi:10.1023/A:1006551516090
• Grosberg, R.K. and R.R. Strathmann, 1998, “One Cell, TwoCell, Red Cell, Blue Cell: The Persistence of a Unicellular Stage inMulticellular Life Histories”,Trends in Ecology &Evolution, 13(3): 112–116.doi:10.1016/S0169-5347(97)01313-X
• Guerrero, Ricardo, Lynn Margulis, and Mercedes Berlanga, 2013,“Symbiogenesis: The Holobiont as a Unit of Evolution”,International Microbiology, 16: 133–143.doi:10.2436/20.1501.01.188
• Hadfield, Jarrod D. and Alastair J. Wilson, 2007,“Multilevel Selection 3: Modeling the Effects of InteractingIndividuals as a Function of Group Size”,Genetics,177(1): 667–668. doi:10.1534/genetics.107.075622
• Haldane, J. B. S., 1932a,The Causes of Evolution,London: Longmans, Green and Co..
• –––, 1932b, “A Mathematical Theory ofNatural and Artificial Selection. Part IX. Rapid Selection”,Mathematical Proceedings of the Cambridge PhilosophicalSociety, 28(2): 244–248. doi:10.1017/S0305004100010914
• Hall, Brian K. (ed.), 1994,Homology: The Hierarchical Basisof Comparative Biology, San Diego, CA: Academic Press.
• Hamilton, W. D., 1963, “The Evolution of AltruisticBehavior”,The American Naturalist, 97(896):354–356. doi:10.1086/497114
• –––, 1964, “The Genetical Evolution ofSocial Behaviour. II”,Journal of Theoretical Biology,7(1): 17–52. doi:10.1016/0022-5193(64)90039-6
• –––, 1975, “Innate Social Aptitudes inMan: An Approach from Evolutionary Genetics”, inBiosocialAnthropology, Robin Fox (ed.), (ASA Studies 1), London: MalabyPress, 133–155.
• Hampe, M. and S.R. Morgan, 1988, “Two Consequences ofRichard Dawkins’ View of Genes and Organisms”,Studiesin History and Philosophy of Science Part A, 19(1):119–138. doi:10.1016/0039-3681(88)90022-2
• Harvey, Michael G., Sonal Singhal, and Daniel L. Rabosky, 2019,“Beyond Reproductive Isolation: Demographic Controls on theSpeciation Process”,Annual Review of Ecology, Evolution,and Systematics, 50: 75–95.doi:10.1146/annurev-ecolsys-110218-024701
• Hautmann, Michael, 2020, “What Is Macroevolution?”,Palaeontology, 63(1): 1–11. doi:10.1111/pala.12465
• Heisler, I. Lorraine and John Damuth, 1987, “A Method forAnalyzing Selection in Hierarchically Structured Populations”,The American Naturalist, 130(4): 582–602.doi:10.1086/284732
• Herron, Matthew D., 2017, “Cells, Colonies, and Clones:Individuality in the Volvocine Algae”, in Lidgard and Nyhart2017: 63–83 (ch. 2).
• –––, 2021, “What Are the MajorTransitions?”,Biology & Philosophy, 36(1): article2. doi:10.1007/s10539-020-09773-z
• Herron, Matthew D. and Richard E. Michod, 2008, “Evolutionof Complexity in the Volvocine Algae: Transitions in Individualitythrough Darwin’s Eye”,Evolution, 62(2):436–451. doi:10.1111/j.1558-5646.2007.00304.x
• Heslot, Nicolas, Hsiao-Pei Yang, Mark E. Sorrells, and Jean-LucJannink, 2012, “Genomic Selection in Plant Breeding: AComparison of Models”,Crop Science, 52(1):146–160. doi:10.2135/cropsci2011.06.0297
• Hölldobler, Bert and Edward O. Wilson, 2009,TheSuperorganism: The Beauty, Elegance, and Strangeness of InsectSocieties, New York: W.W. Norton & Company.
• Hull, David L., 1980, “Individuality and Selection”,Annual Review of Ecology and Systematics, 11: 311–332.doi:10.1146/annurev.es.11.110180.001523
• –––, 1988,Science as a Process: AnEvolutionary Account of the Social and Conceptual Development ofScience, (Science and Its Conceptual Foundations), Chicago, IL:University of Chicago Press.
• Hull, David L. and Michael Ruse (eds.), 2007,The CambridgeCompanion to the Philosophy of Biology, Cambridge/New York:Cambridge University Press. doi:10.1017/CCOL9780521851282
• Hunt, Gene, Kaustuv Roy, and David Jablonski, 2005,“Species-Level Heritability Reaffirmed: A Comment on ‘Onthe Heritability of Geographic Range Sizes’”,TheAmerican Naturalist, 166(1): 129–135.doi:10.1086/430722
• Hurst, Gregory D. D., 2017, “Extended Genomes: Symbiosis andEvolution”,Interface Focus, 7(5): 20170001.doi:10.1098/rsfs.2017.0001
• Huttegger, Simon M. and Rory Smead, 2011, “Efficient SocialContracts and Group Selection”,Biology &Philosophy, 26(4): 517–531.doi:10.1007/s10539-011-9265-3
• Huxley, Julian, 1942,Evolution: The Modern Synthesis,New York/London: Harper & brothers.
• Inkpen, S. Andrew and W. Ford Doolittle, 2021, “AdaptiveRegeneration Across Scales: Replicators and Interactors from Limbs toForests”,Philosophy, Theory, and Practice in Biology,13: article 1. [Inkpen and Doolittle 2021 available online]
• Istvan, M.A., Jr, 2013, “Gould Talking Past Dawkins on theUnit of Selection Issue”,Studies in History and Philosophyof Science Part C: Studies in History and Philosophy of Biological andBiomedical Sciences, 44(3): 327–335.doi:10.1016/j.shpsc.2013.05.020
• Jablonski, David, 1987, “Heritability at the Species Level:Analysis of Geographic Ranges of Cretaceous Mollusks”,Science, 238(4825): 360–363.doi:10.1126/science.238.4825.360
• –––, 2008, “Species Selection: Theory andData”,Annual Review of Ecology, Evolution, andSystematics, 39: 501–524.doi:10.1146/annurev.ecolsys.39.110707.173510
• –––, 2017a, “Approaches to Macroevolution:1. General Concepts and Origin of Variation”,EvolutionaryBiology, 44(4): 427–450. doi:10.1007/s11692-017-9420-0
• –––, 2017b, “Approaches to Macroevolution:2. Sorting of Variation, Some Overarching Issues, and GeneralConclusions”,Evolutionary Biology, 44(4):451–475. doi:10.1007/s11692-017-9434-7
• Jablonski, David and Gene Hunt, 2006, “Larval Ecology,Geographic Range, and Species Survivorship in Cretaceous Mollusks:Organismic versus Species-Level Explanations”,The AmericanNaturalist, 168(4): 556–564. doi:10.1086/507994
• Jones, Emily I., Michelle E. Afkhami, Erol Akçay, Judith L.Bronstein, Redouan Bshary, Megan E. Frederickson, Katy D. Heath, JasonD. Hoeksema, Joshua H. Ness, M. Sabrina Pankey, Stephanie S. Porter,Joel L. Sachs, Klara Scharnagl, and Maren L. Friesen, 2015,“Cheaters Must Prosper: Reconciling Theoretical and EmpiricalPerspectives on Cheating in Mutualism”,EcologyLetters, 18(11): 1270–1284. doi:10.1111/ele.12507
• Kaltenpoth, Martin, Kerstin Roeser-Mueller, Sabrina Koehler,Ashley Peterson, Taras Y. Nechitaylo, J. William Stubblefield, GudrunHerzner, Jon Seger, and Erhard Strohm, 2014, “Partner Choice andFidelity Stabilize Coevolution in a Cretaceous-Age DefensiveSymbiosis”,Proceedings of the National Academy ofSciences, 111(17): 6359–6364.doi:10.1073/pnas.1400457111
• Keller, Evelyn Fox and Elisabeth Anne Lloyd (eds.), 1992,Keywords in Evolutionary Biology, Cambridge, MA: HarvardUniversity Press.
• Kellert, Stephen H., Helen E. Longino, and C. Kenneth Waters(eds.), 2006,Scientific Pluralism, (Minnesota Studies in thePhilosophy of Science 19), Minneapolis, MN: University of MinnesotaPress.
• Kerr, Benjamin and Peter Godfrey-Smith, 2002, “Individualistand Multi-Level Perspectives on Selection in StructuredPopulations”,Biology & Philosophy, 17(4):477–517. doi:10.1023/A:1020504900646
• Kerr, Peter, June Liu, Isabella Cattadori, Elodie Ghedin, AndrewRead, and Edward Holmes, 2015, “Myxoma Virus and theLeporipoxviruses: An Evolutionary Paradigm”,Viruses,7(3): 1020–1061. doi:10.3390/v7031020
• Ketcham, Ryan, 2018, “Equivalence, Interactors, andLloyd’s Challenge to Genic Pluralism”, inMultilevelSelection and the Theory of Evolution, Ciprian Jeler (ed.), Cham:Springer International Publishing, 71–98.doi:10.1007/978-3-319-78677-3_4
• –––, 2019, “Task Allocation and the Logicof Research Questions: How Ants Challenge Human Sociobiology”,Biological Theory, 14(1): 52–68.doi:10.1007/s13752-018-0308-8
• –––, 2024, “What is the AdditivityCriterion for?”, Michael R. Dietrich and Samuel R. Ketcham(eds.),Philosophy, Feminism, and Science: Critical Perspectiveson the Work of Elisabeth A. Lloyd, Pittsburgh, PittsburghUniversity Press.
• Kitcher, Philip, Kim Sterelny, & C. Kenneth Waters, 1990,“The Illusory Riches of Sober’s Monism”,TheJournal of Philosophy, 87(3): 158–161.doi:10.2307/2026634
• Koskella, Britt and Joy Bergelson, 2020, “The Study ofHost–Microbiome (Co)Evolution across Levels of Selection”,Philosophical Transactions of the Royal Society B: BiologicalSciences, 375(1808): 20190604 (8 pages).doi:10.1098/rstb.2019.0604
• Kramer, Jos and Joël Meunier, 2016, “Kin and MultilevelSelection in Social Evolution: A Never-Ending Controversy?”,F1000Research, 5(April): 776.doi:10.12688/f1000research.8018.1
• Krimbas, Costas B., 1984, “On Adaptation, Neo-DarwinianTautology, and Population Fitness”, inEvolutionary Biology,Volume 17, Max K. Hecht, Bruce Wallace, and Ghillean T. Prance(eds.), Boston, MA: Springer US, 1–57.doi:10.1007/978-1-4615-6974-9_1
• Laiolo, Paola and José Ramón Obeso, 2012,“Multilevel Selection and Neighbourhood Effects from Individualto Metapopulation in a Wild Passerine”,PLoSONE, 7(6): e38526. doi:10.1371/journal.pone.0038526
• Laland, Kevin N., Tobias Uller, Marcus W. Feldman, Kim Sterelny,Gerd B. Müller, Armin Moczek, Eva Jablonka, and John Odling-Smee,2015, “The Extended Evolutionary Synthesis: Its Structure,Assumptions and Predictions”,Proceedings of the RoyalSociety B: Biological Sciences, 282(1813): 20151019.doi:10.1098/rspb.2015.1019
• Lamm, Ehud and Oren Kolodny, 2022, “Distributed Adaptations:Can a Species Be Adapted While No Single Individual Carries theAdaptation?”,Frontiers in Ecology and Evolution,10(February): 791104 (13 pages). doi:10.3389/fevo.2022.791104
• Lancichinetti, Andrea and Santo Fortunato, 2009, “CommunityDetection Algorithms: A Comparative Analysis”,PhysicalReview E, 80(5): 056117. doi:10.1103/PhysRevE.80.056117
• Lande, Russell and Stevan J. Arnold, 1983, “The Measurementof Selection on Correlated Characters”,Evolution,37(6): 1210–1226. doi:10.1111/j.1558-5646.1983.tb00236.x
• Lee, Yun Kyung and Sarkis K. Mazmanian, 2010, “Has theMicrobiota Played a Critical Role in the Evolution of the AdaptiveImmune System?”,Science, 330(6012): 1768–1773.doi:10.1126/science.1195568
• Lehmann, Laurent, Laurent Keller, Stuart West, and Denis Roze,2007, “Group Selection and Kin Selection: Two Concepts but OneProcess”,Proceedings of the National Academy ofSciences, 104(16): 6736–6739.doi:10.1073/pnas.0700662104
• Leigh, Egbert G., Jr, 1977, “How Does Selection ReconcileIndividual Advantage with the Good of the Group?”,Proceedings of the National Academy of Sciences, 74(10):4542–4546. doi:10.1073/pnas.74.10.4542
• Levin, Bruce R. and William L. Kilmer, 1974, “InterdemicSelection and the Evolution of Altruism: A Computer SimulationStudy”,Evolution, 28(4): 527–545.doi:10.1111/j.1558-5646.1974.tb00787.x
• Levin, Simon and David Pimentel, 1981, “Selection ofIntermediate Rates of Increase in Parasite-Host Systems”,The American Naturalist, 117(3): 308–315.doi:10.1086/283708
• Lewontin, Richard C., 1958, “A General Method forInvestigating the Equilibrium of Gene Frequency in aPopulation”,Genetics, 43(3): 419–434.doi:10.1093/genetics/43.3.419
• –––, 1962, “Interdeme SelectionControlling a Polymorphism in the House Mouse”,The AmericanNaturalist, 96(887): 65–78. doi:10.1086/282208
• –––, 1970, “The Units of Selection”,Annual Review of Ecology and Systematics, 1: 1–18.doi:10.1146/annurev.es.01.110170.000245
• –––, 1974,The Genetic Basis of EvolutionaryChange, (Columbia Biological Series, no. 25), New York: ColumbiaUniversity Press.
• –––, 1978, “Adaptation”,Scientific American, 239(3): 212–230.doi:10.1038/scientificamerican0978-212
• –––, 1979, “Fitness, Survival andOptimality”, inAnalysis of Ecological Systems, DavidJ. Horn, Gordon R. Stains, and Rodger D. Mitchell (eds.), Cambridge,MA: Harvard University Press, pp. 58–68.
• Lewontin, Richard C. and L. C. Dunn, 1960, “The EvolutionaryDynamics of a Polymorphism in the House Mouse”,Genetics, 45(6): 705–722.doi:10.1093/genetics/45.6.705
• Ley, Ruth E., Catherine A. Lozupone, Micah Hamady, Rob Knight, andJeffrey I. Gordon, 2008, “Worlds within Worlds: Evolution of theVertebrate Gut Microbiota”,Nature ReviewsMicrobiology, 6(10): 776–788. doi:10.1038/nrmicro1978
• Li, C. C., 1967, “Fundamental Theorem of NaturalSelection”,Nature, 214(5087): 505–506.doi:10.1038/214505a0
• Lidgard, Scott and Lynn K. Nyhart (eds.), 2017,BiologicalIndividuality: Integrating Scientific, Philosophical, and HistoricalPerspectives, Chicago, IL: The University of Chicago Press.
• Lieberman, Bruce S. and Elisabeth S. Vrba, 2005, “StephenJay Gould on Species Selection: 30 Years of Insight”,Paleobiology, 31(s2): 113–121.doi:10.1666/0094-8373(2005)031[0113:SJGOSS]2.0.CO;2
• Linksvayer, Timothy A. and Michael J. Wade, 2005, “TheEvolutionary Origin And Elaboration Of Sociality In The AculeateHymenoptera: Maternal Effects, Sib-social Effects, AndHeterochrony”,The Quarterly Review of Biology, 80(3):317–336. doi:10.1086/432266
• –––, 2009, “Genes with Social Effects AreExpected to Harbor More Sequence Variation within and betweenSpecies”,Evolution, 63(7): 1685–1696.doi:10.1111/j.1558-5646.2009.00670.x
• Lloyd, Elisabeth A., 1983, “The Nature of Darwin’sSupport for the Theory of Natural Selection”,Philosophy ofScience, 50(1): 112–129. doi:10.1086/289093
• –––, 1984, “A Semantic Approach to theStructure of Population Genetics”,Philosophy ofScience, 51(2): 242–264. doi:10.1086/289179
• –––, 1986, “Evaluation of Evidence inGroup Selection Debates”,PSA: Proceedings of the BiennialMeeting of the Philosophy of Science Association, 1986(1):483–493. doi:10.1086/psaprocbienmeetp.1986.1.193148
• –––, 1987, “Confirmation of Ecological andEvolutionary Models”,Biology & Philosophy, 2(3):277–293. doi:10.1007/BF00128834
• –––, 1988 [1994],The Structure andConfirmation of Evolutionary Theory (Contributions in Philosophy,37), New York: Greenwood Press. Paperback edition with new preface,Princeton, NJ: Princeton University Press, 1994.
• –––, 1992, “Unit of Selection”, inKeller and Lloyd 1992: 334–340.
• –––, 2001, “Units and Levels of Selection:An Anatomy of the Units of Selection Debates”, inThinkingAbout Evolution (Historical, Philosophical, and PoliticalPerspectives, vol. 2), Rama S. Singh, Costas B. Krimbas, Diane B.Paul, and John Beatty (eds.), Cambridge: Cambridge University Press,267–291 (ch. 13).
• –––, 2005, “Why the Gene Will NotReturn”,Philosophy of Science, 72(2): 287–310.doi:10.1086/432425
• –––, 2007, Reprint of edited version of 2001, inA Cambridge Companion to Philosophy of Biology, Eds David LHull and Michael Ruse, Cambridge UP.
• –––, 2008, “An open letter to ElliottSober and David Sloan Wilson regarding their bookUnto Others: theEvolution and psychology of unselfish behavior”, in EALloyd,Science, Politics, and Evolution, New York/Cambridge:Cambridge University Press, 95–105.
• –––, 2015, “Adaptationism and the Logic ofResearch Questions: How to Think Clearly About EvolutionaryCauses”,Biological Theory, 10(4): 343–362.doi:10.1007/s13752-015-0214-2
• –––, 2017, “Holobionts as Units ofSelection: Holobionts as Interactors, Reproducers, and Manifestors ofAdaptation”, in Gissis, Lamm, and Shavit 2017:351–367.
• –––, 2021,Adaptation, (Elements in thePhilosophy of Biology), Cambridge: Cambridge University Press.doi:10.1017/9781108634953
• Lloyd, Elisabeth A. and Marcus W. Feldman, 2002,“COMMENTARY: Evolutionary Psychology: A View From EvolutionaryBiology”,Psychological Inquiry, 13(2): 150–156.Published with a rebuttal: Ellis and Ketelaar 2002.doi:10.1207/S15327965PLI1302_04
• Lloyd, Elisabeth A. and Stephen Jay Gould, 1993, “SpeciesSelection on Variability”,Proceedings of the NationalAcademy of Sciences, 90(2): 595–599.doi:10.1073/pnas.90.2.595
• Lloyd, Elisabeth A., Richard C. Lewontin, and Marcus W. Feldman,2008, “The Generational Cycle of State Spaces and AdequateGenetical Representation”,Philosophy of Science,75(2): 140–156. doi:10.1086/590196
• Lloyd, Elisabeth A. and Michael J. Wade, 2019, “Criteria forHolobionts from Community Genetics”,Biological Theory,14(3): 151–170. doi:10.1007/s13752-019-00322-w
• Love, Alan C., 2016, “Explaining the Origins ofMulticellularity: Between Evolutionary Dynamics and DevelopmentalMechanisms”, inMulticellularity: Origins andEvolution, Karl J. Niklas and Stuart A. Newman (eds.), Cambridge,MA: MIT press, 279–295.
• Love, Alan C., Otavio Bueno, Ruey-Lin Chen, & Melinda BonnieFagan, 2018, “Individuation, Individuality, and ExperimentalPractice in Developmental Biology”,Individuation, process,and scientific practices, Otavio Bueno, Ruey-Lin Chen, andMelinda Bonnie Fagan (eds.) 2018: 165–191 (ch. 8).
• Love, Alan C. and Ingo Brigandt, 2017, “PhilosophicalDimensions of Individuality”, in Lidgard and Nyhart 2017:318–348 (ch. 13).
• Margulis, Lynn, 1991, “Symbiogenesis andSymbionticism”, in Margulis and Fester 1991: 1–14 (ch.1).
• Margulis, Lynn and René Fester (eds.), 1991,Symbiosisas a Source of Evolutionary Innovation: Speciation andMorphogenesis, Cambridge, MA: MIT Press.
• Marín, César, 2021, “Spatial andDensity-Dependent Multilevel Selection on Weed-Infested Maize”,Genetic Resources and Crop Evolution, 68(3): 885–897.doi:10.1007/s10722-020-01031-1
• Marshall, James A.R., 2011, “Group Selection and KinSelection: Formally Equivalent Approaches”,Trends inEcology & Evolution, 26(7): 325–332.doi:10.1016/j.tree.2011.04.008
• Matessi, C. and S.D. Jayakar, 1976, “Conditions for theEvolution of Altruism under Darwinian Selection”,Theoretical Population Biology, 9(3): 360–387.doi:10.1016/0040-5809(76)90053-8
• Maynard Smith, John, 1964, “Group Selection and KinSelection”,Nature, 201(4924): 1145–1147.doi:10.1038/2011145a0
• –––, 1974, “The Theory of Games and theEvolution of Animal Conflicts”,Journal of TheoreticalBiology, 47(1): 209–221.doi:10.1016/0022-5193(74)90110-6
• –––, 1976, “Group Selection”,Quarterly Review of Biology, 51(2): 277–283.doi:10.1086/409311
• –––, 1978, “Optimization Theory inEvolution”,Annual Review of Ecology and Systematics,9: 31–56. doi:10.1146/annurev.es.09.110178.000335
• –––, 1980, “Models of the Evolution ofAltruism”,Theoretical Population Biology, 18(2):151–159. doi:10.1016/0040-5809(80)90046-5
• –––, 1984, “The Population as a Unit ofSelection”, inEvolutionary Ecology, 23rd BritishEcological Society Symposium, Bryan Shorrocks (ed.), Oxford, UK:Blackwell, pp. 195–202.
• –––, 1987, “Evolutionary Progress andLevels of Selection”, inThe Latest on the Best: Essays onEvolution and Optimality, John Dupré (ed.), Cambridge, MA:MIT Press, 219–230.
• –––, 1991, “A Darwinian View ofSymbiosis”, in Margulis and Fester 2019: 26–37 (ch.3).
• –––, 2001, “Reconciling Marx andDarwin”,Evolution, 55(7): 1496–1498.doi:10.1111/j.0014-3820.2001.tb00670.x
• Maynard Smith, John and Eörs Szathmáry, 1995,TheMajor Transitions in Evolution, Oxford/New York: W.H. FreemanSpektrum.
• Mayr, Ernst, 1976,Evolution and the Diversity of Life,Cambridge, MA: Harvard University Press.
• –––, 1978, “Evolution”,Scientific Amen, 239(3): 46–55.doi:10.1038/scientificamerican0978-46
• –––, 1982, “Adaptation andSelection”,BiologischesZentralblatt, 101:161–174.
• McAdam, Andrew G., Dany Garant, and Alastair J. Wilson, 2014,“The Effects of Others’ Genes: Maternal and Other IndirectGenetic Effects”, inQuantitative Genetics in the Wild,Anne Charmantier, Dany Garant, and Loeske E. B. Kruuk (eds.), Oxford:Oxford University Press, 84–103 (ch. 6).doi:10.1093/acprof:oso/9780199674237.003.0006
• McAdam, Andrew G., Quinn M. R. Webber, Ben Dantzer, Jeffrey E.Lane, and Stan Boutin, 2022, “Social Effects on Annual Fitnessin Red Squirrels”,Journal of Heredity, 113(1):69–78. doi:10.1093/jhered/esab051
• McCauley, David E. and Michael J. Wade, 1980, “GroupSelection: The Genetic and Demographic Basis for the PhenotypicDifferentiation of Small Populations ofTriboliumcastaneum”,Evolution, 34(4): 813–821.doi:10.1111/j.1558-5646.1980.tb04020.x
• McFall-Ngai, Margaret, Michael G. Hadfield, Thomas C. G. Bosch,Hannah V. Carey, Tomislav Domazet-Lošo, Angela E. Douglas,Nicole Dubilier, Gerard Eberl, Tadashi Fukami, Scott F. Gilbert, etal., 2013, “Animals in a Bacterial World, a New Imperative forthe Life Sciences”,Proceedings of the National Academy ofSciences, 110(9): 3229–3236.doi:10.1073/pnas.1218525110
• McGlothlin, Joel W., Allen J. Moore, Jason B. Wolf, and Edmund D.Brodie III, 2010, “Interacting Phenotypes and the EvolutionaryProcess. III. Social Evolution”,Evolution, 64(9):2558–2574.
• McGlothlin, Joel W. and Laura F. Galloway, 2014a, “TheContribution of Maternal Effects to Selection Response: An EmpiricalTest of Competing Models”,Evolution, 68(2):549–558.
• McGlothlin, Joel W., Jason B. Wolf, Edmund D. Brodie III, andAllen J. Moore, 2014b, “Quantitative Genetic Versions ofHamilton’s Rule with Empirical Applications”,Philosophical Transactions of the Royal Society B: BiologicalSciences, 369(1642): 20130358.
• McKenna, Kenneth Z., Günter P. Wagner, and Kimberly L.Cooper, 2021, “A Developmental Perspective of Homology andEvolutionary Novelty”, inCurrent Topics in DevelopmentalBiology 141, Cambridge, MA: Academic Press, 1–38.doi:10.1016/bs.ctdb.2020.12.001
• Meincke, Anne Sophie and John Dupré (eds.), 2020,Biological Identity: Perspectives from Metaphysics and thePhilosophy of Biology, New York: Routledge.doi:10.4324/9781351066389
• Michod, Richard E. and Matthew D. Herron, 2006, “Cooperationand Conflict during Evolutionary Transitions in Individuality”,Journal of Evolutionary Biology, 19(5): 1406–1409.doi:10.1111/j.1420-9101.2006.01142.x
• Michod, Richard E. and Denis Roze, 1999, “Cooperation andConflict in the Evolution of Individuality. III. Transitions in theUnit of Fitness”, inMathematical and Computational Biology:Computational Morphogenesis, Hierarchical Complexity, and DigitalEvolution, Chrystopher L. Nehaniv (ed.), (Lectures on Mathematicsin the Life Sciences 26), Providence, RI: American MathematicalSociety, 47–92.
• –––, 2001, “Cooperation and Conflict inthe Evolution of Multicellularity”,Heredity, 86(1):1–7. doi:10.1046/j.1365-2540.2001.00808.x
• Mitchell, Sandra D., 1987, “Competing Units of Selection?: ACase of Symbiosis”,Philosophy of Science, 54(3):351–367. doi:10.1086/289388
• Molter, Daniel J., 2019, “On MycorrhizalIndividuality”,Biology & Philosophy, 34(5):article 52. doi:10.1007/s10539-019-9706-y
• Moore, Allen J., Edmund D. Brodie, and Jason B. Wolf, 1997,“Interacting Phenotypes and the Evolutionary Process: I. Directand Indirect Genetic Effects of Social Interactions”,Evolution, 51(5): 1352–1362.doi:10.1111/j.1558-5646.1997.tb01458.x
• Moran, Nancy A. and Daniel B. Sloan, 2015, “The HologenomeConcept: Helpful or Hollow?”,PLOS Biology, 13(12):e1002311. doi:10.1371/journal.pbio.1002311
• Morowitz, Harold J., 1992,Beginnings of Cellular Life:Metabolism Recapitulates Biogenesis, (Bio-Origins Series), NewHaven, CT: Yale University Press.
• Morrison, Monica Ainhorn, 2021, “The Models are Alright: aSocio-Epistemic Theory of the Landscape of Climate ModelDevelopment”, PhD thesis, Indiana University.
• Muir, William M., 1996, “Group Selection for Adaptation toMultiple-Hen Cages: Selection Program and Direct Responses”,Poultry Science, 75(4): 447–458.doi:10.3382/ps.0750447
• –––, 2005, “Incorporation of CompetitiveEffects in Forest Tree or Animal Breeding Programs”,Genetics, 170(3): 1247–1259.doi:10.1534/genetics.104.035956
• Munson, Ronald, 1971, “Biological Adaptation”,Philosophy of Science, 38(2): 200–215. Reprinted inTopics in the Philosophy of Biology, Marjorie Grene andEverett Mendelsohn (eds), (Boston Studies in the Philosophy of Science27), Dordrecht: Springer Netherlands, 330–350.doi:10.1086/288354 doi:10.1007/978-94-010-1829-6_15
• Mutic, Joshua J. and Jason B. Wolf, 2007, “Indirect GeneticEffects from Ecological Interactions in Arabidopsis Thaliana”,Molecular Ecology, 16(11): 2371–2381.doi:10.1111/j.1365-294X.2007.03259.x
• Nanay, Bence, 2011, “Replication Without Replicators”,Synthese, 179(3): 455–477.doi:10.1007/s11229-009-9702-x
• Nanninga, Gerrit B. and Andrea Manica, 2018, “LarvalSwimming Capacities Affect Genetic Differentiation and Range Size inDemersal Marine Fishes”,Marine Ecology ProgressSeries, 589: 1–12.
• Nicholson, Daniel J. and John Dupré (eds.), 2018,Everything Flows: Towards a Processual Philosophy of Biology,Oxford: Oxford University Press.doi:10.1093/oso/9780198779636.001.0001
• Nicholson, Jeremy K., Elaine Holmes, James Kinross, Remy Burcelin,Glenn Gibson, Wei Jia, and Sven Pettersson, 2012, “Host-GutMicrobiota Metabolic Interactions”,Science, 336(6086):1262–1267. doi:10.1126/science.1223813
• Nowak, Martin A., Corina E. Tarnita, and Edward O. Wilson, 2010,“The Evolution of Eusociality”,Nature,466(7310): 1057–1062. doi:10.1038/nature09205
• –––, 2011, “Nowak et al. Reply”,Nature, 471(7339): E9–E10. doi:10.1038/nature09836
• Odling-Smee, John and Kevin N. Laland, 2011, “EcologicalInheritance and Cultural Inheritance: What Are They and How Do TheyDiffer?”,Biological Theory, 6(3): 220–230.doi:10.1007/s13752-012-0030-x
• Ohta, Kuniyoshi, 1983, “Hierarchical Theory of Selection:The Covariance Formula of Selection and Its Applications”,Japanese Journal of Biometrics, 4: 25–33.doi:10.5691/jjb.4.25
• Okasha, Samir, 2004a, “Multi-level Selection, Covariance andContextual Analysis”,British Journal for the Philosophy ofScience, 55(3): 481–504. doi:10.1093/bjps/55.3.481
• –––, 2004b, “Multilevel Selection and thePartitioning of Covariance: A Comparison of Three Approaches”,Evolution, 58(3): 486–494.doi:10.1111/j.0014-3820.2004.tb01672.x
• –––, 2006,Evolution and the Levels ofSelection, New York: Oxford University Press.doi:10.1093/acprof:oso/9780199267972.001.0001
• –––, 2011, “Reply to Sober andWaters”,Philosophy and Phenomenological Research,82(1): 241–8. doi:10.1111/j.1933-1592.2010.00474.x
• –––, 2015, “The Relation Between Kin andMultilevel Selection: An Approach Using Causal Graphs”,British Journal for the Philosophy of Science, 67(2):435–470. doi:10.1093/bjps/axu047
• –––, 2018,Agents and Goals inEvolution, Oxford: Oxford University Press.doi:10.1093/oso/9780198815082.001.0001
• –––, 2022, “The Major Transitions inEvolution—A Philosophy-of-Science Perspective”,Frontiers in Ecology and Evolution, 10(February): 793824.doi:10.3389/fevo.2022.793824
• Okasha, Samir and Jun Otsuka, 2020, “The Price Equation andthe Causal Analysis of Evolutionary Change”,PhilosophicalTransactions of the Royal Society B: Biological Sciences,375(1797): 20190365. doi:10.1098/rstb.2019.0365
• Okasha, Samir and C. Paternotte, 2012, “Group Adaptation,Formal Darwinism and Contextual Analysis: Group Adaptation, FormalDarwinism and Contextual Analysis”,Journal of EvolutionaryBiology, 25(6): 1127–1139.doi:10.1111/j.1420-9101.2012.02501.x
• O’Malley, Maureen A., 2017, “From Endosymbiosis toHolobionts: Evaluating a Conceptual Legacy”,Journal ofTheoretical Biology, 434(December): 34–41.doi:10.1016/j.jtbi.2017.03.008
• O’Malley, Maureen A. and John Dupré, 2007,“Size Doesn’t Matter: Towards a More Inclusive Philosophyof Biology”,Biology & Philosophy, 22(2):155–191. doi:10.1007/s10539-006-9031-0
• O’Malley, Maureen A., Ingo Brigandt, Alan C. Love, John W.Crawford, Jack A. Gilbert, Rob Knight, Sandra D. Mitchell, and ForestRohwer, 2014, “Multilevel Research Strategies and BiologicalSystems”,Philosophy of Science, 81(5): 811–828.doi:10.1086/677889
• O’Malley, Maureen A. and Russell Powell, 2016, “MajorProblems in Evolutionary Transitions: How a Metabolic Perspective CanEnrich Our Understanding of Macroevolution”,Biology &Philosophy, 31(2): 159–189.doi:10.1007/s10539-015-9513-z
• Oster, George F. and Edward O. Wilson, 1978,Caste and Ecologyin the Social Insects, (Monographs in Population Biology 12),Princeton, NJ: Princeton University Press.
• Otsuka, Jun, 2016a, “Causal Foundations of EvolutionaryGenetics”,The British Journal for the Philosophy ofScience, 67(1): 247–269. doi:10.1093/bjps/axu039
• –––, 2016b, “A Critical Review of theStatisticalist Debate”,Biology & Philosophy,31(4): 459–482. doi:10.1007/s10539-016-9528-0
• –––, 2019a, “Ontology, Causality, andMethodology of Evolutionary Research Programs”, inEvolutionary Causation: Biological and PhilosophicalReflections, Tobias Uller and Kevin N. Laland (eds.), Cambridge,MA: MIT Press, 247–264.
• –––, 2019b,The Role of Mathematics inEvolutionary Theory, Cambridge: Cambridge University Press.doi:10.1017/9781108672115
• Otsuka, Jun, Trin Turner, Colin Allen, and Elisabeth A. Lloyd,2011, “Why the Causal View of Fitness Survives”,Philosophy of Science, 78(2): 209–224.doi:10.1086/659219
• Oyama, Susan, 1985,The Ontogeny of Information: DevelopmentalSystems and Evolution, Cambridge/New York: Cambridge UniversityPress.
• Oyama, Susan, Paul E. Griffiths, and Russell D. Gray (eds.), 2001,Cycles of Contingency: Developmental Systems and Evolution(Life and Mind), Cambridge, MA: MIT Press.
• Papale, François, 2021, “Evolution by Means ofNatural Selection without Reproduction: Revamping Lewontin’sAccount”,Synthese, 198(11): 10429–10455.
• Pie, Marcio R. and Andreas L. S. Meyer, 2017, “The Evolutionof Range Sizes in Mammals and Squamates: Heritability and DifferentialEvolutionary Rates for Low- and High-Latitude Limits”,Evolutionary Biology, 44(3): 347–355.doi:10.1007/s11692-017-9412-0
• Planer, Ronald J., 2015, “On the Free-Rider IdentificationProblem”,Biological Theory, 10(2): 134–144.doi:10.1007/s13752-015-0206-2
• Polly, Paul, Jesualdo Fuentes-Gonzalez, A. Michelle Lawing,Allison Bormet, and Robert G. Dundas, 2017, “Clade Sorting Has aGreater Effect than Local Adaptation on Ecometric Patterns inCarnivora”,Evolutionary Ecology Research,18(February): 61–95.
• Pradeu, Thomas, 2010a, “The Organism in DevelopmentalSystems Theory”,Biological Theory, 5(3):216–222. doi:10.1162/BIOT_a_00042
• –––, 2010b, “What Is an Organism? AnImmunological Answer”,History and Philosophy of the LifeSciences, 32(2–3): 247–267.
• –––, 2011, “A Mixed Self: The Role ofSymbiosis in Development”,Biological Theory, 6(1):80–88. doi:10.1007/s13752-011-0011-5
• –––, 2016a, “Organisms or BiologicalIndividuals? Combining Physiological and EvolutionaryIndividuality”,Biology & Philosophy, 31(6):797–817. doi:10.1007/s10539-016-9551-1
• –––, 2016b, “The Many Faces of BiologicalIndividuality”,Biology & Philosophy, 31(6):761–773. doi:10.1007/s10539-016-9553-z
• –––, 2018, “Genidentity and BiologicalProcesses”, in Nicholson and Dupré 2018: 96–112.doi:10.1093/oso/9780198779636.003.0005
• –––, 2019, “Immunology andIndividuality”,ELife, 8(April): e47384. [Pradeu 2019 available online]
• Price, George R., 1970, “Selection and Covariance”,Nature, 227(5257): 520–521. doi:10.1038/227520a0
• –––, 1972, “Extension of CovarianceSelection Mathematics”,Annals of Human Genetics,35(4): 485–490. doi:10.1111/j.1469-1809.1957.tb01874.x
• Queller, David C., 1992, “Quantitative Genetics, InclusiveFitness, and Group Selection”,The American Naturalist,139(3): 540–558. doi:10.1086/285343
• –––, 2000, “Relatedness and the FraternalMajor Transitions”,Philosophical Transactions of the RoyalSociety of London. Series B: Biological Sciences, 355(1403):1647–1655. doi:10.1098/rstb.2000.0727
• –––, 2011, “Expanded Social Fitness andHamilton’s Rule for Kin, Kith, and Kind”,Proceedingsof the National Academy of Sciences, 108(supplement 2):10792–10799. doi:10.1073/pnas.1100298108
• Queller, David C. and Joan E. Strassmann, 2009, “BeyondSociety: The Evolution of Organismality”,PhilosophicalTransactions of the Royal Society B: Biological Sciences,364(1533): 3143–3155. doi:10.1098/rstb.2009.0095
• –––, 2016, “Problems of Multi-SpeciesOrganisms: Endosymbionts to Holobionts”,Biology &Philosophy, 31(6): 855–873.doi:10.1007/s10539-016-9547-x
• Raff, Rudolf A., 1996,The Shape of Life: Genes, Development,and the Evolution of Animal Form, Chicago, IL: University ofChicago Press.
• Raff, Rudolf A. and Elizabeth C. Raff (eds.), 1987,Development as an Evolutionary Process, Proceedings of aMeeting Held at the Marine Biological Laboratory in Woods Hole,Massachusetts, August 23 and 24, 1985, (MBL Lectures in Biology 8),New York: A.R. Liss.
• Raff, Rudolf A. and Thomas C. Kaufman, 1983,Embryos, Genes,and Evolution: The Developmental-Genetic Basis of EvolutionaryChange, New York/London: Macmillan.
• Ramsey, Grant and Robert Brandon, 2011, “Why ReciprocalAltruism Is Not a Kind of Group Selection”,Biology &Philosophy, 26(3): 385–400.doi:10.1007/s10539-011-9261-7
• Rees, Tobias, Thomas Bosch, and Angela E. Douglas, 2018,“How the Microbiome Challenges Our Concept of Self”,PLOS Biology, 16(2): e2005358.doi:10.1371/journal.pbio.2005358
• Ren, Tiantian, Stan Boutin, Murray M. Humphries, Ben Dantzer,Jamieson C. Gorrell, David W. Coltman, Andrew G. McAdam, and MartinWu, 2017, “Seasonal, Spatial, and Maternal Effects on GutMicrobiome in Wild Red Squirrels”,Microbiome, 5(1):article 163. doi:10.1186/s40168-017-0382-3
• Rice, Brian R. and Alexander E. Lipka, 2021, “DiversifyingMaize Genomic Selection Models”,Molecular Breeding,41(5): article 33. doi:10.1007/s11032-021-01221-4
• Robertson, Alan, 1966, “Artificial Selection in Plants andAnimals”,Proceedings of the Royal Society of London. SeriesB. Biological Sciences, 164(995): 341–349.doi:10.1098/rspb.1966.0036
• Rosenberg, Alexander, 2006,Darwinian Reductionism, or, How toStop Worrying and Love Molecular Biology, Chicago, IL: Universityof Chicago Press.
• Rosenberg, E and I Zilber-Rosenberg, 2020, “The HologenomeConcept of Evolution: Do Mothers Matter Most?”,BJOG: AnInternational Journal of Obstetrics & Gynaecology, 127(2):129–137. doi:10.1111/1471-0528.15882
• Roughgarden, Joan, Scott F. Gilbert, Eugene Rosenberg, IlanaZilber-Rosenberg, and Elisabeth A. Lloyd, 2018, “Holobionts asUnits of Selection and a Model of Their Population Dynamics andEvolution”,Biological Theory, 13(1): 44–65.doi:10.1007/s13752-017-0287-1
• Round, June L., Ryan M. O’Connell, and Sarkis K. Mazmanian,2010, “Coordination of Tolerogenic Immune Responses by theCommensal Microbiota”,Journal of Autoimmunity, 34(3):J220–J225. doi:10.1016/j.jaut.2009.11.007
• Rousset, François, 2004,Genetic Structure andSelection in Subdivided Populations, (Monographs in PopulationBiology 40), Princeton, NJ: Princeton University Press.
• Sampson, Timothy R. and Sarkis K. Mazmanian, 2015, “Controlof Brain Development, Function, and Behavior by the Microbiome”,Cell Host & Microbe, 17(5): 565–576.doi:10.1016/j.chom.2015.04.011
• Sapienza, Carmen, 2010, “Selection Does Operate Primarily onGenes: In Defense of the Gene as the Unit of Selection”, inContemporary Debates in Philosophy of Biology, FranciscoJosé Ayala and Robert Arp (eds.), (Contemporary Debates inPhilosophy 12), Chichester, UK/Malden, MA: Wiley-Blackwell,127–140.
• Sariola, Salla and Scott F. Gilbert, 2020, “Toward aSymbiotic Perspective on Public Health: Recognizing the Ambivalence ofMicrobes in the Anthropocene”,Microorganisms, 8(5):746. doi:10.3390/microorganisms8050746
• Sarkar, Sahotra, 1994, “The Selection of Alleles and theAdditivity of Variance”,PSA Proceedings of the BiennialMeeting of the Philosophy of Science Association, 1994(1):3–12. doi:10.1086/psaprocbienmeetp.1994.1.193006
• –––, 2008, “A Note on Frequency Dependenceand the Levels/Units of Selection”,Biology &Philosophy, 23(2): 217–228.doi:10.1007/s10539-007-9092-8
• Schneider, Tamar, 2021, “The Holobiont Self: UnderstandingImmunity in Context”,History and Philosophy of the LifeSciences, 43(3): article 99 (23 pages).doi:10.1007/s40656-021-00454-y
• Seeley, Thomas D., 1997, “Honey Bee Colonies Are Group-LevelAdaptive Units”,The American Naturalist, 150(S1):S22–S41. doi:10.1086/286048
• Şencan, Sinan, 2019, “A Tale of Two IndividualityAccounts and Integrative Pluralism”,Philosophy ofScience, 86(5): 1111–1122. doi:10.1086/705523
• Sepkoski, David, 2020,Catastrophic Thinking: Extinction andthe Value of Diversity from Darwin to the Anthropocene,(Science.Culture), Chicago, IL: The University of Chicago Press.
• Shaffer, Zachary, Takao Sasaki, Brian Haney, Marco Janssen,Stephen C. Pratt, and Jennifer H. Fewell, 2016, “TheFoundress’s Dilemma: Group Selection for Cooperation amongQueens of the Harvester Ant,Pogonomyrmexcalifornicus”,Scientific Reports, 6(1):29828. doi:10.1038/srep29828
• Shanahan, Timothy, 1997, “Pluralism, Antirealism, and theUnits of Selection”,Acta Biotheoretica, 45(2):117–126. doi:10.1023/A:1000377821347
• Shavit, Ayelet, 2005, “The Notion of ‘Group’ andTests of Group Selection”,Philosophy of Science,72(5): 1052–1063. doi:10.1086/508119
• Shavit, Ayelet and Roberta L. Millstein, 2008, “GroupSelection Is Dead! Long Live Group Selection?”,BioScience, 58(7): 574–575. doi:10.1641/B580702
• Shelton, Deborah E. and Richard E. Michod, 2014a, “Levels ofSelection and the Formal Darwinism Project”,Biology &Philosophy, 29(2): 217–224.doi:10.1007/s10539-013-9420-0
• –––, 2014b, “Group Selection and GroupAdaptation During Major Evolutionary Transition: Insights from theEvolution of Multicellularity in the Volvocine Algae”,Biological Theory, 9(4): 452–469.doi:10.1007/s13752-014-0159-x
• –––, 2020, “Group and Individual Selectionduring Evolutionary Transitions in Individuality: Meanings andPartitions”,Philosophical Transactions of the Royal SocietyB: Biological Sciences, 375(1797): 20190364.doi:10.1098/rstb.2019.0364
• Sih, Andrew and Jason V. Watters, 2005, “The Mix Matters:Behavioural Types and Group Dynamics in Water Striders”,Behaviour, 142(9–10): 1417–1431.doi:10.1163/156853905774539454
• Simon, Burton, 2014, “Continuous-Time Models of GroupSelection, and the Dynamical Insufficiency of Kin SelectionModels”,Journal of Theoretical Biology, 349(May):22–31. doi:10.1016/j.jtbi.2014.01.030
• Skillings, Derek, 2016, “Holobionts and the Ecology ofOrganisms: Multi-Species Communities or IntegratedIndividuals?”,Biology & Philosophy, 31(6):875–892. doi:10.1007/s10539-016-9544-0
• Slatkin, Montgomery, 1972, “On Treating the Chromosome asthe Unit of Selection”,Genetics, 72(1): 157–168.doi:10.1093/genetics/72.1.157
• –––, 1981a, “A Diffusion Model of SpeciesSelection”,Paleobiology, 7(4) : 421–425.doi:10.1017/S0094837300025471
• –––, 1981b, “PopulationalHeritability”,Evolution, 35(5): 859–871.doi:10.1111/j.1558-5646.1981.tb04949.x
• Slatkin, Montgomery and Michael J. Wade, 1978, “GroupSelection on a Quantitative Character”,Proceedings of theNational Academy of Sciences, 75(7): 3531–3534.doi:10.1073/pnas.75.7.3531
• Slobodkin, Lawrence B. and Anatol Rapoport, 1974, “AnOptimal Strategy of Evolution”,The Quarterly Review ofBiology, 49(3): 181–200. doi:10.1086/408082
• Smith, Eric Alden, 1994, “Semantics, Theory, andMethodological Individualism in the Group-SelectionControversy”,Behavioral and Brain Sciences, 17(4):636–637. doi:10.1017/S0140525X00036414
• Smith, Subrena E., 2017, “Organisms as Persisters”,Philosophy, Theory, and Practice in Biology, 9: article 14.doi:10.3998/ptb.6959004.0009.014
• Sober, Elliott, 1980, “Holism, Individualism, and the Unitsof Selection”,PSA: Proceedings of the Biennial Meeting ofthe Philosophy of Science Association, 1980(2): 93–121.doi:10.1086/psaprocbienmeetp.1980.2.192588
• –––, 1984,The Nature of Selection:Evolutionary Theory in Philosophical Focus, Cambridge, MA: MITPress.
• –––, 1990, “The Poverty of Pluralism: AReply to Sterelny and Kitcher”,The Journal ofPhilosophy, 87(3): 151–158. doi:10.2307/2026633
• Sober, Elliott and Richard C. Lewontin, 1982, “Artifact,Cause and Genic Selection”,Philosophy of Science,49(2): 157–180. doi:10.1086/289047
• Sober, Elliott and David Sloan Wilson, 1994, “A CriticalReview of Philosophical Work on the Units of Selection Problem”,Philosophy of Science, 61(4): 534–555.doi:10.1086/289821
• –––, 1998,Unto Others: The Evolution andPsychology of Unselfish Behavior, Cambridge, MA: HarvardUniversity Press.
• Spencer, Hamish G. and Judith C. Masters, 1992, “SexualSelection: Contemporary Debates”, in Keller and Lloyd 1992:294–301.
• Spencer, Hamish G., Martyn Kennedy, and Russell D. Gray, 1995,“Patch Choice with Competitive Asymmetries and PerceptualLimits: The Importance of History”,Animal Behaviour,50(2): 497–508. doi:10.1006/anbe.1995.0264
• Stanford, P. Kyle, 2001, “The Units of Selection and theCausal Structure of the World”,Erkenntnis, 54(2):215–233. doi:10.1023/A:1005641025742
• Stencel, Adrian and Javier Suárez, 2021, “Do SomaticCells Really Sacrifice Themselves? Why an Appeal to Coercion May Be aHelpful Strategy in Explaining the Evolution ofMulticellularity”,Biological Theory, 16(2):102–113. doi:10.1007/s13752-021-00376-9
• Stencel, Adrian and Dominika M. Wloch-Salamon, 2018, “SomeTheoretical Insights into the Hologenome Theory of Evolution and theRole of Microbes in Speciation”,Theory in Biosciences,137(2): 197–206. doi:10.1007/s12064-018-0268-3
• Sterelny, Kim, 1996a, “Explanatory Pluralism in EvolutionaryBiology”,Biology & Philosophy, 11(2):193–214. doi:10.1007/BF00128919
• –––, 1996b, “The Return of theGroup”,Philosophy of Science, 63(4): 562–584.doi:10.1086/289977
• –––, 2004, “Symbiosis, Evolvability andModularity”, inModularity in Development andEvolution, Gerhard Schlosser and Günter P. Wagner (eds.),Chicago, IL: University of Chicago Press, 490–516 (ch. 22).
• –––, 2011, “Darwinian Spaces: PeterGodfrey-Smith on Selection and Evolution”,Biology &Philosophy, 26(4): 489–500.doi:10.1007/s10539-010-9244-0
• –––, 2012,The Evolved Apprentice: HowEvolution Made Humans Unique, Cambridge, MA: MIT Press.doi:10.7551/mitpress/9780262016797.001.0001
• Sterelny, Kim and Paul E. Griffiths, 1999,Sex and Death: AnIntroduction to Philosophy of Biology, (Science and ItsConceptual Foundations), Chicago, IL: University of ChicagoPress.
• Sterelny, Kim and Philip Kitcher, 1988, “The Return of theGene”,The Journal of Philosophy, 85(7): 339–361.doi:10.2307/2026953
• Sterner, Beckett, 2015, “Pathways to Pluralism aboutBiological Individuality”,Biology & Philosophy,30(5): 609–628. doi:10.1007/s10539-015-9494-y
• Stevens, Lori, Charles J. Goodnight, and Susan Kalisz, 1995,“Multilevel Selection in Natural Populations ofImpatienscapensis”,The American Naturalist, 145(4):513–526. doi:10.1086/285753
• Strassmann, Joan E. and David C. Queller, 2011, “Evolutionof Cooperation and Control of Cheating in a Social Microbe”,Proceedings of the National Academy of Sciences,108(supplement 2): 10855–10862. doi:10.1073/pnas.1102451108
• Strassmann, Joan E., David C. Queller, John C. Avise, andFrancisco J. Ayala, 2011, “In the Light of Evolution V:Cooperation and Conflict”,Proceedings of the NationalAcademy of Sciences, 108(supplement_2): 10787–10791.doi:10.1073/pnas.1100289108
• Suárez, Javier, 2018, “‘The Importance ofSymbiosis in Philosophy of Biology: An Analysis of the Current Debateon Biological Individuality and Its Historical Roots’”,Symbiosis, 76(2): 77–96.doi:10.1007/s13199-018-0556-1
• –––, 2020, “The Stability of TraitsConception of the Hologenome: An Evolutionary Account of HolobiontIndividuality”,History and Philosophy of the LifeSciences, 42(1): article 11. doi:10.1007/s40656-020-00305-2
• Suárez, Javier and Elisabeth A. Lloyd, 2023,Units ofSelection (Elements in the Philosophy of Biology), Cambridge:Cambridge University Press. doi:10.1017/9781009276429
• Suárez, Javier and Adrian Stencel, 2020, “APart-dependent Account of Biological Individuality: Why Holobionts AreIndividuals and Ecosystems Simultaneously”,BiologicalReviews, 95(5): 1308–1324. doi:10.1111/brv.12610
• Suárez, Javier and Vanessa Triviño, 2020,“What Is a Hologenomic Adaptation? Emergent Individuality andInter-Identity in Multispecies Systems”,Frontiers inPsychology, 11(March): 187. doi:10.3389/fpsyg.2020.00187
• Sukhoverkhov, Anton V. and Nathalie Gontier, 2021,“Non-Genetic Inheritance: Evolution above the OrganismalLevel”,Biosystems, 200(February): 104325 (6 pages).doi:10.1016/j.biosystems.2020.104325
• Szathmáry, Eors, 1999, “The First Replicators”,inLevels of Selection in Evolution, Laurent Keller (ed.),(Monographs in Behavior and Ecology), Princeton, NJ: PrincetonUniversity Press, 31–52.
• Szathmáry, Eors and John Maynard Smith, 1993, “TheOrigin of Genetic Systems”,Abstracta Botanica, 17:197–206.
• Theis, Kevin R., Nolwenn M. Dheilly, Jonathan L. Klassen, RobertM. Brucker, John F. Baines, Thomas C. G. Bosch, John F. Cryan, ScottF. Gilbert, Charles J. Goodnight, Elisabeth A. Lloyd, Jan Sapp,Philippe Vandenkoornhuyse, Ilana Zilber-Rosenberg, Eugene Rosenberg,and Seth R. Bordenstein, 2016, “Getting the Hologenome ConceptRight: An Eco-Evolutionary Framework for Hosts and TheirMicrobiomes”,MSystems, 1(2): e00028-16.doi:10.1128/mSystems.00028-16
• Thies, Christoph and Richard A. Watson, 2021, “IdentifyingCauses of Social Evolution: Contextual Analysis, the Price Approach,and Multilevel Selection”,Frontiers in Ecology andEvolution, 9(December): 780508 (14 pages).doi:10.3389/fevo.2021.780508
• Thoday, John M., 1953, “Components of Fitness”,Symposia of the Society for Experimental Biology, 7:96–113. [Thoday 1953 available online (pdf)]
• Thompson, Paul, 1989,The Structure of BiologicalTheories, (SUNY Series in Philosophy and Biology), Albany, NY:State University of New York Press.
• Traulsen, Arne, 2010, “Mathematics of Kin- andGroup-Selection: Formally Equivalent?”,Evolution,64(2): 316–323. doi:10.1111/j.1558-5646.2009.00899.x
• Triviño, Vanessa and Javier Suárez, 2020,“Holobionts: Ecological Communities, Hybrids, or BiologicalIndividuals? A Metaphysical Perspective on MultispeciesSystems”,Studies in History and Philosophy of Science PartC: Studies in History and Philosophy of Biological and BiomedicalSciences, 84: 101323. doi:10.1016/j.shpsc.2020.101323
• Tsuji, Kazuki, 1995, “Reproductive Conflicts and Levels ofSelection in the AntPristomyrmexpungens:Contextual Analysis and Partitioning of Covariance”,TheAmerican Naturalist, 146(4): 586–607.doi:10.1086/285816
• Tyler, Anna L., Folkert W. Asselbergs, Scott M. Williams, andJason H. Moore, 2009, “Shadows of Complexity: What BiologicalNetworks Reveal about Epistasis and Pleiotropy”,BioEssays, 31(2): 220–227.doi:10.1002/bies.200800022
• Uyenoyama, Marcy K., 1979, “Evolution of Altruism UnderGroup Selection in Large and Small Populations in FluctuatingEnvironments”,Theoretical Population Biology, 15(1):58–85. doi:10.1016/0040-5809(79)90027-3
• Uyenoyama, Marcy K. and Marcus W. Feldman, 1980, “Theoriesof Kin and Group Selection: A Population Genetics Perspective”,Theoretical Population Biology, 17(3): 380–414.doi:10.1016/0040-5809(80)90033-7
• Van der Steen, W. J. and H. A. Van der Berg, 1999,“Dissolving Disputes over Genic Selectionism”,Journalof Evolutionary Biology, 12(1): 184–187.doi:10.1046/j.1420-9101.1999.00021.x
• van Fraassen, Bas C. van, 1980,The Scientific Image,Oxford: Clarendon Press. doi:10.1093/0198244274.001.0001
• van Veelen, Matthijs, 2009a, “Group Selection, KinSelection, Altruism, and Cooperation: When Inclusive Fitness is Rightand When it Can be Wrong”,Journal of TheoreticalBiology, 259(3): 589–600.doi:10.1016/j.jtbi.2009.04.019
• –––, 2009b, “Does it Pay to be Good?Competing Evolutionary Explanations of Pro-Social Behaviour”,The Moral Brain: Essays on the Evolutionary and NeuroscientificAspects of Morality,185–200.
• van Veelen, Matthijs, Julián García, Maurice W.Sabelis, and Martijn Egas, 2012, “Group Selection and InclusiveFitness Are Not Equivalent; the Price Equation vs. Models andStatistics”,Journal of Theoretical Biology,299(April): 64–80. doi:10.1016/j.jtbi.2011.07.025
• van Veelen, Matthijs, Shishi Luo, and Burton Simon, 2014, “ASimple Model of Group Selection That Cannot Be Analyzed with InclusiveFitness”,Journal of Theoretical Biology,360(November): 279–289. doi:10.1016/j.jtbi.2014.07.004
• van Vliet, Simon and Michael Doebeli, 2019, “The Role ofMultilevel Selection in Host Microbiome Evolution”,Proceedings of the National Academy of Sciences, 116(41):20591–20597. doi:10.1073/pnas.1909790116
• Vehrencamp, Sandra L. and Jack W Bradbury, 1984, “MatingSystems and Ecology”, inBehaviouralEcology: AnEvolutionary Approach, second edition, John R. Krebs and NicholasB. Davies (eds.), Sunderland, MA: Sinauer, pp. 251–278.
• Veigl, Sophie J., Javier Suárez, and Adrian Stencel, 2022,“Rethinking Hereditary Relations: The Reconstitutor as theEvolutionary Unit of Heredity”,Synthese, 200(5):article 367. doi:10.1007/s11229-022-03810-y
• Vrba, Elisabeth S., 1983, “Macroevolutionary Trends: NewPerspectives on the Roles of Adaptation and Incidental Effect”,Science, 221(4608): 387–389.doi:10.1126/science.221.4608.387
• –––, 1984, “What Is SpeciesSelection?”,Systematic Zoology, 33(3): 318–328.doi:10.2307/2413077
• –––, 1989, “Levels of Selection andSorting with Special Reference to the Species Level”,OxfordSurveys in Evolutionary Biology, Volume 6, Paul H. Harvey (ed.),Oxford: Oxford University Press, 111–168.
• Vrba, Elisabeth S. and Niles Eldredge, 1984, “Individuals,Hierarchies and Processes: Towards a More Complete EvolutionaryTheory”,Paleobiology, 10(2): 146–171.doi:10.1017/S0094837300008149
• Vrba, Elisabeth S. and Stephen Jay Gould, 1986, “TheHierarchical Expansion of Sorting and Selection: Sorting and SelectionCannot Be Equated”,Paleobiology, 12(2): 217–228.doi:10.1017/S0094837300013671
• Waddington, C. H., 1956, “Genetic Assimilation of theBithorax Phenotype”,Evolution, 10(1):1–13. doi:10.1111/j.1558-5646.1956.tb02824.x
• Wade, Michael J., 1976, “Group Selections among LaboratoryPopulations ofTribolium”,Proceedings of theNational Academy of Sciences, 73(12): 4604–4607.doi:10.1073/pnas.73.12.4604
• –––, 1977, “An Experimental Study of GroupSelection”,Evolution, 31(1): 134–153.doi:10.1111/j.1558-5646.1977.tb00991.x
• –––, 1978, “A Critical Review of theModels of Group Selection”,Quarterly Review ofBiology, 53(2): 101–114. doi:10.1086/410450
• –––, 1979, “Sexual Selection and Variancein Reproductive Success”,The American Naturalist,114(5): 742–747. doi:10.1086/283520
• –––, 1980a, “Kin Selection: ItsComponents”,Science, 210(4470): 665–667.doi:10.1126/science.210.4470.665
• –––, 1980b, “An Experimental Study of KinSelection”,Evolution, 34(5): 844–855.doi:10.1111/j.1558-5646.1980.tb04023.x
• –––, 1985, “Soft Selection, HardSelection, Kin Selection, and Group Selection”,The AmericanNaturalist, 125(1): 61–73. doi:10.1086/284328
• –––, 1994, “The Biology of the ImportedWillow Leaf Beetle, Plagiodera versicolora (Laicharting)”,Novel Aspects of the Biology of Chrysomelidae:541–547.
• –––, 1995, “The Ecology of SexualSelection: Mean Crowding of Females and Resource-DefencePolygyny”,Evolutionary Ecology, 9(1): 118–124.doi:10.1007/BF01237701
• –––, 1996, “Adaptation in SubdividedPopulations: Kin Selection and Interdemic Selection”, inAdaptation, Michael R. Rose and George V. Lauder (eds.), SanDiego, CA: Academic Press, 381–405.
• –––, 2000, “Opposing Levels of SelectionCan Cause Neutrality: Mating Patterns and Maternal-FetalInteractions”,Evolution, 54(1): 290–292.doi:10.1111/j.0014-3820.2000.tb00029.x
• –––, 2007, “The Co-Evolutionary Geneticsof Ecological Communities”,Nature Reviews Genetics,8(3): 185–195. doi:10.1038/nrg2031
• –––, 2014, “Genotype-by-EnvironmentInteractions and Sexual Selection: Female Choice in a ComplexWorld”, inGenotype-by-Environment Interactions and SexualSelection, John Hunt and David Hosken (eds.), Chichester, UK:John Wiley & Sons, Ltd, 1–18.doi:10.1002/9781118912591.ch1
• –––, 2016,Adaptations inMetapopulations, Chicago, IL: University of Chicago Press.
• Wade, Michael J., Piter Bijma, Esther D. Ellen, and William Muir,2010, “Group Selection and Social Evolution in DomesticatedAnimals: Group Selection and Social Evolution”,EvolutionaryApplications, 3(5–6): 453–465.doi:10.1111/j.1752-4571.2010.00147.x
• Wade, Michael J. and Felix Breden, 1981, “Effect ofInbreeding on the Evolution of Altruistic Behavior by KinSelection”,Evolution, 35(5): 844–858.doi:10.1111/j.1558-5646.1981.tb04948.x
• Wade, Michael J. and Devin M. Drown, 2016,“Nuclear–Mitochondrial Epistasis: A Gene’s Eye Viewof Genomic Conflict”,Ecology and Evolution, 6(18):6460–6472. doi:10.1002/ece3.2345
• Wade, Michael J. and Charles J. Goodnight, 1998,“Perspective: The Theories of Fisher and Wright in the Contextof Metapopulations: When Nature Does Many Small Experiments”,Evolution, 52(6): 1537–1553.doi:10.1111/j.1558-5646.1998.tb02235.x
• Wade, Michael J. and James R. Griesemer, 1998, “PopulationalHeritability: Empirical Studies of Evolution inMetapopulations”,The American Naturalist, 151(2):135–147. doi:10.1086/286107 [Wade and Griesemer 1998 available online]
• Wade, Michael J. and Susan Kalisz, 1990, “The Causes ofNatural Selection”,Evolution, 44(8): 1947–1955.doi:10.1111/j.1558-5646.1990.tb04301.x
• Wade, Michael J. and David E. McCauley, 1980, “GroupSelection: The Phenotypic and Genotypic Differentiation of SmallPopulations”,Evolution, 34(4): 799–812.doi:10.1111/j.1558-5646.1980.tb04019.x
• Wagner, Günter P., 2015, “Evolutionary Innovations andNovelties: Let Us Get down to Business!”,ZoologischerAnzeiger: A Journal of Comparative Zoology, 256(May):75–81. doi:10.1016/j.jcz.2015.04.006
• –––, 2016, “What Is ‘HomologyThinking’ and What Is It For?”,Journal ofExperimental Zoology Part B: Molecular and DevelopmentalEvolution, 326(1): 3–8. doi:10.1002/jez.b.22656
• Wagner, Günter P., Koryu Kin, Louis Muglia, and MihaelaPavli, 2014, “Evolution of Mammalian Pregnancy and the Origin ofthe Decidual Stromal Cell”,The International Journal ofDevelopmental Biology, 58(2–3–4): 117–126.doi:10.1387/ijdb.130335gw
• Waldron, Anthony, 2007, “Null Models of Geographic RangeSize Evolution Reaffirm Its Heritability”,The AmericanNaturalist, 170(2): 221–231. doi:10.1086/518963
• Walsh, Denis M., Tim Lewens, and André Ariew, 2002,“The Trials of Life: Natural Selection and Random Drift”,Philosophy of Science, 69(3): 452–473.doi:10.1086/342454
• Warrener, Anna, 2023, “The Multifactor Pelvis: AnAlternative to the Adaptationist Approach of the ObstetricalDilemma”,Evolutionary Anthropology: Issues, News, andReviews32 (5): 260–274.
• Waters, C. Kenneth, 1986, “Models of Natural Selection fromDarwin to Dawkins”, Ph.D. dissertation, Indiana University.
• –––, 1991, “Tempered Realism about theForce of Selection”,Philosophy of Science, 58(4):553–573. doi:10.1086/289640
• –––, 2005, “Why Genic and MultilevelSelection Theories Are Here to Stay”,Philosophy ofScience, 72(2): 311–333. doi:10.1086/432426
• Watson, Richard A. and Eörs Szathmáry, 2016,“How Can Evolution Learn?”,Trends in Ecology &Evolution, 31(2): 147–157.doi:10.1016/j.tree.2015.11.009
• Wein, Tanita, Devani Romero Picazo, Frances Blow, ChristianWoehle, Elie Jami, Thorsten B.H. Reusch, William F. Martin, and TalDagan, 2019, “Currency, Exchange, and Inheritance in theEvolution of Symbiosis”,Trends in Microbiology,27(10): 836–849. doi:10.1016/j.tim.2019.05.010
• Weinig, Cynthia, Jill A. Johnston, Charles G. Willis, and Julin N.Maloof, 2007, “Antagonistic Multilevel Selection on Size andArchitecture in Variable Density Settings”,Evolution,61(1): 58–67. doi:10.1111/j.1558-5646.2007.00005.x
• Wenseleers, Tom, Andy Gardner, and Kevin R. Foster, 2010,“Social Evolution Theory: A Review of Methods andApproaches”, inSocial Behaviour: Genes, Ecology andEvolution, Tamás Székely, Allen J. Moore, and JanKomdeur (eds.), Cambridge: Cambridge University Press, 132–158.doi:10.1017/CBO9780511781360.013
• West, Stuart A., Claire El Mouden, and Andy Gardner, 2011,“Sixteen Common Misconceptions about the Evolution ofCooperation in Humans”,Evolution and Human Behavior,32(4): 231–262. doi:10.1016/j.evolhumbehav.2010.08.001
• West, Stuart A. and Andy Gardner, 2013, “Adaptation andInclusive Fitness”,Current Biology, 23(13):R577–R584. doi:10.1016/j.cub.2013.05.031
• West, Stuart A., A. S. Griffin, and Andy Gardner, 2007,“Social Semantics: Altruism, Cooperation, Mutualism, StrongReciprocity and Group Selection”,Journal of EvolutionaryBiology, 20(2): 415–432.doi:10.1111/j.1420-9101.2006.01258.x
• –––, 2008, “Social Semantics: How UsefulHas Group Selection Been?: Group Confusion”,Journal ofEvolutionary Biology, 21(1): 374–385.doi:10.1111/j.1420-9101.2007.01458.x
• West-Eberhard, Mary Jane, 1992, “Adaptation: CurrentUsages”, in Keller and Lloyd 1992: 13–18.
• Westneat, David F., 2012, “Evolution in Response to SocialSelection: The Importance of Interactive Effects of Traits onFitness”,Evolution, 66(3): 890–895.doi:10.1111/j.1558-5646.2011.01490.x
• Whitham, Thomas G., Gerard J. Allan, Hillary F. Cooper, andStephen M. Shuster, 2020, “Intraspecific Genetic Variation andSpecies Interactions Contribute to Community Evolution”,Annual Review of Ecology, Evolution, and Systematics, 51:587–612. doi:10.1146/annurev-ecolsys-011720-123655
• Williams, George C., 1966,Adaptation and Natural Selection: ACritique of Some Current Evolutionary Thought, Princeton, NJ:Princeton University Press.
• –––, 1985, “A Defense of Reductionism inEvolutionary Biology”,Oxford Surveys in EvolutionaryBiology, Volume 2, Richard Dawkins and Mark Ridley (eds),Oxford/New York: Oxford University Press, 1–27.
• –––, 1990, “Review ofThe Structureand Confirmation of Evolutionary Theory, by Elisabeth A.Lloyd”,The Quarterly Review of Biology, 65(4):504–504. doi:10.1086/416987
• –––, 1992,Natural Selection: Domains,Levels, and Challenges, (Oxford Series in Ecology and Evolution4), New York: Oxford University Press.
• Williams, Hywel T. P. and Timothy M. Lenton, 2008,“Environmental Regulation in a Network of Simulated MicrobialEcosystems”,Proceedings of the National Academy ofSciences, 105(30): 10432–10437.doi:10.1073/pnas.0800244105
• Wilson, David Sloan, 1975, “A Theory of GroupSelection”,Proceedings of the National Academy ofSciences, 72(1): 143–146. doi:10.1073/pnas.72.1.143
• –––, 1980,The Natural Selection ofPopulations and Communities, Menlo Park, CA: BenjaminCummings.
• –––, 1983, “The Group SelectionControversy: History and Current Status”,Annual Review ofEcology and Systematics, 14: 159–187.doi:10.1146/annurev.es.14.110183.001111
• –––, 1997, “Biological Communities asFunctionally Organized Units”,Ecology, 78(7):2018–2024.doi:10.1890/0012-9658(1997)078[2018:BCAFOU]2.0.CO;2
• Wilson, David Sloan and Robert K. Colwell, 1981, “Evolutionof Sex Ratio in Structured Demes”,Evolution, 35(5):882–897. doi:10.1111/j.1558-5646.1981.tb04952.x
• Wilson, David Sloan and Edward O. Wilson, 2007, “Rethinkingthe Theoretical Foundation of Sociobiology”,The QuarterlyReview of Biology, 82(4): 327–348. doi:10.1086/522809
• Wilson, Edward O., 2008, “One Giant Leap: How InsectsAchieved Altruism and Colonial Life”,BioScience,58(1): 17–25. doi:10.1641/B580106
• –––, 2012,The Social Conquest ofEarth, New York: W.W. Norton.
• Wilson, Edward O. and Martin A. Nowak, 2014, “NaturalSelection Drives the Evolution of Ant Life Cycles”,Proceedings of the National Academy of Sciences, 111(35):12585–12590. doi:10.1073/pnas.1405550111
• Wilson, Robert A., 2003, “Pluralism, Entwinement, and theLevels of Selection”,Philosophy of Science, 70(3):531–552. doi:10.1086/376784
• Wimsatt, William C., 1980a, “Reductionistic ResearchStrategies and Their Biases in the Units of SelectionControversy”, inScientific Discovery: Case Studies,Thomas Nickles (ed.), (Boston Studies in the Philosophy of Science60), Dordrecht: Springer Netherlands, 213–259.doi:10.1007/978-94-009-9015-9_13
• –––, 1980b, “Units of Selection and theStructure of the Multi-Level Genome”,PSA: Proceedings ofthe Biennial Meeting of the Philosophy of Science Association,1980(2): 122–183.doi:10.1086/psaprocbienmeetp.1980.2.192589
• –––, 1997, “Aggregativity: ReductiveHeuristics for Finding Emergence”,Philosophy ofScience, 64(S4): S372–S384. doi:10.1086/392615
• –––, 2007,Re-Engineering Philosophy forLimited Beings: Piecewise Approximations to Reality, Cambridge,MA: Harvard University Press.
• Wolf, Jason B., 2003, “Genetic Architecture and EvolutionaryConstraint When the Environment Contains Genes”,Proceedingsof the National Academy of Sciences, 100(8): 4655–4660.doi:10.1073/pnas.0635741100
• Wolf, Jason B., Edmund D. Brodie III, James M. Cheverud, Allen J.Moore, and Michael J. Wade, 1998, “Evolutionary Consequences ofIndirect Genetic Effects”,Trends in Ecology &Evolution, 13(2): 64–69.doi:10.1016/S0169-5347(97)01233-0
• Wolf, Jason B., Edmund D. Brodie III, and Allen J. Moore, 1999,“Interacting Phenotypes and the Evolutionary Process. II.Selection Resulting from Social Interactions”,The AmericanNaturalist, 153(3): 254–266. doi:10.1086/303168
• Wolf, Jason B., Edmund D. Brodie, and Michael John Wade (eds.),2000,Epistasis and the Evolutionary Process, Oxford/NewYork: Oxford University Press.
• Woodward, James, 1993, “Capacities and Invariance”, inPhilosophical Problems of the Internal and External Worlds: Essayson the Philosophy of Adolf Grünbaum, John Earman, Allen I.Janis, Gerald J. Massey, and Nicholas Rescher (eds.),(Pittsburgh-Konstanz Series in the Philosophy and History of Science),Pittsburgh: University of Pittsburgh Press, 283–328.
• Wright, Larry, 1978, “The Ins and Outs of Teleology: ACritical Examination of Woodfield”,Inquiry,21(1–4): 223–237. doi:10.1080/00201747808601859
• Wright, Sewall, 1931, “Evolution in MendelianPopulations”,Genetics, 16(2): 97–159.doi:10.1093/genetics/16.2.97
• –––, 1938, “Size of Population andBreeding Structure in Relation to Evolution”,Science,87(2263): 430–431.
• –––, 1945, “Tempo and Mode in Evolution: ACritical Review”,Ecology, 26(4): 415–419.doi:10.2307/1931666
• –––, 1980, “Genic and OrganismicSelection”,Evolution, 34(5): 825–843.doi:10.1111/j.1558-5646.1980.tb04022.x
• –––, 1984,Evolution and the Genetics ofPopulations, Volume 4: Variability Within and Among NaturalPopulations,Chicago: Chicago University Press.
• Wynne-Edwards, Vero Copner, 1962,Animal Dispersion inRelation to Social Behavior, Edinburgh: Oliver and Boyd.
• Zacaï, Axelle, Emmanuel Fara, Arnaud Brayard, RémiLaffont, Jean-Louis Dommergues, and Christian Meister, 2017,“Phylogenetic Conservatism of Species Range Size Is the CombinedOutcome of Phylogeny and Environmental Stability”,Journalof Biogeography, 44(11): 2451–2462.doi:10.1111/jbi.13043
• Zilber-Rosenberg, Ilana and Eugene Rosenberg, 2008, “Role ofMicroorganisms in the Evolution of Animals and Plants: The HologenomeTheory of Evolution”,FEMS Microbiology Reviews, 32(5):723–735. doi:10.1111/j.1574-6976.2008.00123.x

Academic Tools

	How to cite this entry.
	Preview the PDF version of this entry at theFriends of the SEP Society.
	Look up topics and thinkers related to this entry at the Internet Philosophy Ontology Project (InPhO).
	Enhanced bibliography for this entryatPhilPapers, with links to its database.

Other Internet Resources

• Goodnight, Charles J., 2014, “Epistasis and Long Term Selection Experiments”, conference poster presented atEvolution 2014.

Related Entries

altruism: biological |biological individuals |biology: philosophy of |developmental biology: evolution and development |evolution |game theory: evolutionary |gene |genetics: ecological |macroevolution, philosophy of |natural selection |replication and reproduction