1. Introduction

On traditional views of organism life cycles, development and reproduction are phases of life trajectories culminating in the generation of offspring, which initiate new life trajectories. To the extent that stages of development in offspring are repetitions of stages in their parents, concatenated sequences of life trajectories constitute life cycles, each formed by processes of generation, development, and reproduction. This picture is a fair representation for simple life cycles in which development can be seen as relatively continuous growth and differentiation of shapes and sizes of parts along with the maturation necessary to reproduce. Complex life cycles involve relatively discontinuous and substantial changes of form, behavior, or environment, as when (1) larvae metamorphose into adults differing in form and life habits, (2) parasites move from one sort of host to others to complete development, or (3) branch points create alternative offspring trajectories, leading, say, to wingless form, aquatic lifestyle, or asexual reproduction rather than winged form, terrestrial lifestyle, or sexual reproduction. One general account describes complex life cycles as involving “discrete phases that exhibit contrasting morphological, physiological, behavioral, or ecological attributes” (Moran Reference Moran1994, 574).

Complex life cycles complicate relations between processes of development and reproduction to such an extent that even the meaning of ‘organism’ begins to break down. In colonial “organisms,” such as bryozoans, component zooids do not have the same degree of autonomy and integration as free-living “organisms,” yet they can separate from the colony to pursue different life trajectories. It is likewise open to question whether the growth of a sea mat, bacterial film, or yeast colony is a matter of either development or reproduction (Dawkins Reference Dawkins1983; Ereshefsky and Pedroso Reference Ereshefsky and Pedroso2013).

Peter Godfrey-Smith (Reference Godfrey-Smith2009) pursues a divide-and-conquer approach to these conceptual matters, making considerable progress toward a coherent understanding of the diversity of modes of development and reproduction and their role in a Darwinian picture of evolution. He divides “reproducers” (Griesemer Reference Griesemer2000a) into simple, collective, and scaffolded. The first and last are relatively coherent categories in traditional terms according to Godfrey-Smith. Simple reproducers “can reproduce largely ‘under their own steam’—or, more exactly, using their own machinery, in conjunction with external sources of energy and raw materials” (Godfrey-Smith Reference Godfrey-Smith2009, 88). Scaffolded reproducers, on the other hand, “are entities which get reproduced as part of the reproduction of some larger unit (a simple reproducer), or that are reproduced by some other entity. Their reproduction is dependent on an elaborate scaffolding of some kind that is external to them” (88). In between are the collective reproducers: “reproducing entities with parts that themselves have the capacity to reproduce, where the parts do so largely through their own resources rather than through the coordinated activity of the whole. Not all the parts need to be able to do this for an entity to count as a collective, the requirement is that some can” (87).

For this intermediate category, Godfrey-Smith conquers difficult cases by constructing a three-dimensional space of “reproduction-related” parameters for bottlenecks, germ/soma specializations, and integration. Each is a degree property, so intermediate cases are covered by degrees up to limiting cases of simple material reproduction, for example, in bacteria, where the entire body of the organism divides to form offspring, and cases of simple scaffolded reproduction, for example, in retroviruses, where none of the body of the virus entity provides material to offspring viruses; only information is conveyed from parent to offspring in formal reproduction mediated by extensive host machinery playing external scaffold.

The scheme is ingenious, clarifying much discussion in contemporary biology. Yet it relies on a conceptual grounding to which I seek alternatives, not because I oppose it as a research strategy, but from a commitment to discover robust scientific results with a multiplicity of models and perspectives over which to evaluate idealizing assumptions (Griesemer Reference Griesemer2000a, Reference Cartwright2014a; Wimsatt Reference Wimsatt2007). The conceptual gradient from simple to collective to scaffolded biological reproduction progressively sidelines materiality from full-bodied fission (simple), to partial-bodied passage through material bottlenecks of varying degrees from buffalo herds to slime molds (collective), to single-cell preformistic germ/soma bottlenecks (Bonner Reference Bonner1974; Dawkins Reference Dawkins1983; Buss Reference Buss1987; Grosberg and Strathmann Reference Grosberg and Strathmann1998), to completely formal reproduction where most or all material involvement is from the scaffolding sidelines and the “essence” of reproduction is transmission of information. Godfrey-Smith (Reference Godfrey-Smith2009, 79) notes the tension between his endorsement of a causal role for information distinct from matter and Oyama’s (Reference Oyama1985/2000) powerful critique, but he shrugs it off by suggesting that formal reproduction occurs in special circumstances only “below the levels of organisms and cells.” Contemporary biology’s privileging of genes in explanations of the evolutionary significance of bottlenecks, reproductive specialization, and degree of integration, however, makes it too easy to treat the informational role of genes in development and reproduction as foundational across all modes of reproduction.

There is nothing wrong with such programs in biology, understood Wimsatt’s (Reference Wimsatt2007) way: as heuristic scientific modeling strategies. Godfrey-Smith’s approach suggests modeling strategies based on the parameters he identifies. Related modeling approaches are suggested by Queller and Strathmann’s (Reference Queller and Strathmann2009) cooperation/conflict space for representing degrees of “organismality.” I argued for a heuristic alternative to these sorts of replicator- and information-grounded perspectives to explore not only difficult cases but also whole domains hotly contested as involving biological reproduction at all, for example, origins of precellular life or cultural evolution with nongenetic modes of inheritance (Griesemer Reference Griesemer2000a, Reference Griesemer, Jones and Cartwright2005; Wimsatt and Griesemer Reference Wimsatt, Griesemer, Sansom and Brandon2007). In this respect, my approach is in sympathy with developmental systems theory, epigenetics, evo-devo, eco-devo, and niche construction—all theories pushing outward from gene-centric biology. I want also to push beyond conventional boundaries of biology to explore development- and reproduction-like processes in chemistry and culture to reconceptualize the role of materiality in biological reproduction and discover new questions and modeling approaches in biology and beyond.

On the account of reproduction I favor, development is not a life cycle phase preceding reproduction; rather, developmental and reproduction processes are mutually embedding, entwined aspects of life made coherent by their intertwining. In this article, I seek to interpret complexities of life cycles in terms of this mutual embedding.

This is a big agenda—too big for one article. The goal for this article is to explore one aspect in one context: an alternative interpretation of scaffolding as part of the “genetic” contribution (in an expanded, beyond-the-genes sense) to a material mode of reproduction in the context of complex life cycles. On my account of reproduction, templating in the narrow sense of semiconservative gene replication is a form of scaffolding; scaffolding is a general aspect of development and also a particular mode of reproduction. Scaffolds are temporary structures that interact in assembly, construction, maintenance, or developmental processes with developing entities that more easily acquire or sustain a capacity, skill, or dynamic state at an appropriate time than they otherwise would, or which would otherwise be more costly to achieve (Bickhard Reference Bickhard, Winegar and Valsiner1992). In exemplary cases, scaffolding is removed or falls away from the finished “product,” or the developed individual becomes autonomous from the scaffold with respect to the skill or capacity scaffolded (Caporael, Griesemer, and Wimsatt Reference Caporael, Griesemer and Wimsatt2014). In other cases, the scaffold plays a constructional role while becoming “internalized” as part of the structure of the developing entity. DNA replication can be understood in terms of the internalization of scaffolding, in which one strand serves as a template or scaffold for the construction of a complementary strand and then remains attached to the new strand as part of an offspring double helix. In reverse transcription, an original RNA strand serves as a template but is “discarded” after a second round of templating from a cDNA construct (Griesemer Reference Cartwright2014a). The embedding relations between development and reproduction entail that scaffolding in reproduction is also general. Rather than seek degree properties that isolate scaffolded reproduction as a special case, I see it as the general situation of biological entities conveying developmental capacities to offspring through interaction with their “developmental environments” by means of material overlap of parts with their parent(s).

In the next section, I state my views on reproduction and development in order to identify a role for scaffolding. The following section articulates a concept of scaffolded development to argue that, owing to the general role of scaffolds, all life cycles are complex in the sense that life trajectories are interspersed with what might be called “developmental generations.” The degree property of complexity is not proposed as a parameter or variable, however, as it is neither explanans nor explanandum for my project. Rather, understanding the organization of scaffolding interactions is sought as a means to describe developmental systems in terms of scaffolded reproduction to explore new modes of evolutionary explanation. The final section sketches a developmental reaction norms perspective in which scaffolding interactions structure representations of development–environment relations and provide a potential focal point for modeling developmental evolution.

2. Reproduction, Development, and Inheritance

Reproducers are entities with the capacity to multiply (make more reproducers) because offspring bear relations of ‘material overlap’ with their parents (Griesemer Reference Griesemer2000a, Reference Cartwright2014a, Reference Cartwright2014b). “More-making” capacity does not entail a process without inputs of energy, raw materials, or assistance of various kinds. Material overlap means that reproduction involves bonds of material continuity, not merely resemblance or formal information transmission. Otherwise, we might well describe reproductive more-making as production, as in a factory making many similar products, or communication, as in conveying information with speech or other signals. Griesemer (Reference Griesemer, Jones and Cartwright2005) argues that information copying or merely formal relations are problematic as stand-alone concepts of inheritance in abstraction from the material conditions of reproduction. The latter, not the former, determine the causal pathways of hereditary relations; the flow of genetic information depends on materially intimate connections of “sender” and “receiver.” Most importantly, in reproduction at least some material overlapping parts convey or confer developmental capacities on offspring via transfer of material parts—a propagule generation or ‘progeneration’ (Griesemer Reference Griesemer2000a). An issue that divides me and Godfrey-Smith is what counts as a salient material bond: he sees merely formal or informational relations in, for example, retroviral replication, whereas I see material overlap due to RNA strand hybridization guiding and channeling flows of genetic information (Griesemer Reference Cartwright2014a).

Development is the recursive acquisition, refinement, or maintenance of a capacity to reproduce. Reproductive capacity is realized in diverse ways and modes of development in extant life-forms on earth. The double recursion of development and reproduction bottoms out in ‘null development’, in which progenerated entities are born ready-made with a capacity to develop rather than having to acquire it. Development in multicellular organisms, for example, requires reproduction of cells, reproduction of cells requires development of cellular capacities, and cellular development requires autocatalysis (null development) of molecular constituents.

Inheritance is reproduction further specified in terms of particular qualities of developmental mechanisms involved (Griesemer Reference Griesemer2000a). Inheritance is a process in which evolved mechanisms of development are propagated in reproduction. This concept is inclusive, admitting epigenetic and nongenetic mechanisms. Replication, on my view, is a special form of inheritance in which evolved scaffolding mechanisms of development include template coding mechanisms, such as semiconservative DNA and RNA strand replication.

A virtue of this reproducer perspective is to admit a wide range of developmental systems in which not all hereditary, even informational, resources need be located in genes, cells, or even organisms, and which may extend to parts of reproducer environments (see Oyama Reference Oyama1985/2000; Griffiths and Gray Reference Griffiths and Gray1994; Oyama, Griffiths, and Gray Reference Oyama, Griffiths and Gray2001; Gilbert and Epel Reference Gilbert and Epel2009; Griffiths and Stotz Reference Griffiths and Stotz2013). Elsewhere, I consider developmental systems that may not even be composed solely of living constituents, but may include “prostheses” such as hermit crab houses and twigs used as tools by birds, host molecules used by parasites as protection from host immune system attack, substrates such as rocks for marine holdfasts, and artifacts used by human makers of “culture” (Wimsatt and Griesemer Reference Wimsatt, Griesemer, Sansom and Brandon2007; Griesemer Reference Cartwright2014a). I do not regard every element of environment or ecological interaction, however, as constituting part of a developmental system or process. Only certain kinds of hybridized systems, in which some parts play the role of developmental scaffolding toward other parts, count as developmental systems. I call them hybrids to signal that the material bonds that form hybrid entities function in an expanded sense to convey or confer developmental capacities in analogy with the function of hybridization of genetic entities formed in syngamy. Nonliving environmental scaffolds form a different kind of material bond with developing organisms than developmental capacities conveyed through syngamy and template DNA replication. The difference is salient to the particular mechanisms by which developmental capacities are conveyed through scaffolding interactions but not to the fact of material connection in hybrids of both sorts. Nor do the disanalogies undermine the shared fact of the material role of scaffolding in both cases in the propagation, through material overlap, of developmental capacities. Indeed, the material bond between paternal and maternal genomes in a zygote is no closer than that between barnacle foot and rocky substrate; both involve noncovalent chemical interaction (Wiegemann, Kowalik, and Hartwig Reference Wiegemann, Kowalik and Hartwig2006). On Godfrey-Smith’s taxonomy of simple, collective, and scaffolded reproducers, only entities like viruses depending on extensive host replication machinery and metabolism count as scaffolded reproducers, while typical multicellular eukaryotes, which nevertheless may depend on parents, siblings, rivals, teachers, other species, artifacts, and many other persistent or recurrent nonliving resources for their development, are called “collective” reproducers.

In my view, most or all life cycles are complex because most or all are scaffolded. Scaffolding is inherently collective, even if the collected entities are not necessarily (all) cells or even living. What differentiates kinds of reproducers on my account is modes and mechanisms of scaffolded development rather than the fact (or not) of it. My focus in the next section is how developmental scaffolding in hybrid systems works in the context of complex life cycles.

3. Scaffolded Development and Complex Life Cycles

To model evolution, biologists “cut” life cycles at their generational joints between parents and offspring to identify individual organisms as fitness-bearers and distinguish generations across which to iterate descriptions of change. But what are parents and offspring? In my view, reproduction is a process spanning the whole life trajectory in its embedding relation with development. Parent- and offspring-hood are delimited by material overlap relations. When material parts of any complex system fission or fuse, there is a generic sort of reproduction or ‘general progeneration’ (Griesemer Reference Griesemer2000a). What avoids trivializing reproduction as just any change of parts is that reproduction involves the conveyance or conferral of developmental capacities. Not every mereological change achieves that. Moreover, since development is the acquisition of a capacity to reproduce, only lineage-forming (or terminating) mereological changes in development count.

In modern biology, the cell bottleneck identified with meiotic gamete formation (in multicellular sexual creatures like us) is used to demarcate generations. Gamete formation in creatures like us involves several coincident features that together give us confidence in calling a zygote a new being: diploid parent cells form haploid gametes in meiosis to which diploidy is restored in offspring by syngamy. Subsequent mitotic cell divisions count descendant cells, if not too spatially isolated from one another, as parts of a developing body rather than as reproduction of new organism individuals. Gametogenesis is a form of what Godfrey-Smith calls “simple” cellular reproduction, involving whole-body fission, so there is no discrepancy in the behavior of parts of the collection constituting the ancestral cell: all material parts go to one offspring cell or another, even if in females several of the four grand-offspring cells form polar bodies rather than eggs. Since the zygote is also a single cell, all higher-level organization into tissues and organs is lost in gametogenesis, to be redeveloped only after gametes form viable zygotes. Developmental “resetting” coincides with ploidy change, whole-body fission, spatial separation, and a degree of metabolic autonomy of gamete, zygote, larva, or juvenile from parents to give us a robust sense of the demarcation of reproductive generations.

Scaffolding and complex life cycles complicate this picture. Unlike the unique fission and fusion of genomes in gametogenesis and syngamy in seemingly simple life cycles of creatures like us, genome reorganization and scaffolding interactions can happen at many points in complex life cycles. Contemporary biology grapples with complex life cycles in which the simple picture of genome fissioning and fusion breaks down, but the floodgates are open, even for creatures like us, if we consider other forms of scaffolding besides the role one DNA or RNA strand plays in the template scaffolding construction of another or the role of host scaffold “machinery” in the production of offspring development.

Life cycles are simple if the entire developing life trajectory is individuated by generation and reproduction “events” involving only a single “phase” throughout, in Moran’s (Reference Moran1994, 574) sense of maintaining “morphological, physiological, behavioral, or ecological attributes,” for example, a haploid genome, or a tailed body, or living in a single habitat or niche, rather than exhibiting several discrete phases with contrasting attributes. Life cycles are complex if a developing individual’s life trajectory includes distinct phases, for example, diploid as well as haploid genomes, larval as well as adult forms, mobile as well as sedentary habits, aquatic as well as terrestrial niches. A life cycle is multiple (multigenerational) if a complete turn of the cycle returning to an earlier set of attributes involves multiple reproductive generations of developing individuals that collectively instantiate distinct phases. If, instead of a single individual metamorphosing from aquatic larva to terrestrial adult, a life cycle involves a simple aquatic life trajectory in a parent generation followed by a simple terrestrial life trajectory in the offspring generation, returning to the aquatic trajectory in grand-offspring, then the life cycle is complex in the multiple sense. Many plant species have haploid (gametophyte) individuals in one reproductive generation and diploid (sporophyte) individuals in the next generation, so one turn of such plant life cycles includes two reproductive generations as demarcated by ploidy and cell bottleneck criteria. Multivoltine insects display alternative adult phenotypes in different seasons, for example, winged versus nonwinged forms. Parasites using different hosts in different reproductive generations and alternation of generations between sexual and parthenogenetic reproductive modes are also kinds of multiple, complex life cycles (Moran Reference Moran1994; Gilbert and Epel Reference Gilbert and Epel2009; see also Godfrey-Smith Reference Godfrey-Smith2016; Herron Reference Herron2016; O’Malley Reference O’Malley2016).

What distinguishes discrete phases within developmental life trajectories from discrete phases among reproductive generations? “How many different ‘lives’ [may] an animal … fit into its cycle?” (Minelli Reference Minelli2003, 67). Simple life cycles have one developmental phase per trajectory demarcated from the next in the life cycle by a single reproductive “event.” Developmental stages are traditionally viewed as stages of single reproductive individuals, while reproduction marks a distinction between one (or more) parent and offspring individuals. In some of the complex life cycles mentioned above, there is more than one developmental phase per reproductive “event.” In others, development takes more than a single life trajectory to complete. The traditional distinction of developmental phases in simple or complex life cycles from reproductive generations rests, I contend, on the coincidence of cell bottlenecks with changes of genetic ploidy or other markers of reproductive generations. Godfrey-Smith’s collective reproducers present difficulties because these criteria need not coincide at reproduction. There is no ploidy change or cell bottleneck when buffalo herds fission, yet some conceptualize group fissioning as reproduction (Griesemer and Wade Reference Griesemer and Wade2000). Some argue that biofilms exhibit “individuality” yet do not involve any of the traditional markers of individuation by reproductive events (Ereshefsky and Pedroso Reference Ereshefsky and Pedroso2013).

Cellular bottlenecks are important to the ways evolutionary biologists conceptualize ‘organism’ due to two kinds of consequences: (1) restriction of the reproducing material “propagule” to one or a few cells eliminates new genetic variation arising in somatic cell lines, reducing within-organism competition among organism parts for reproductive advantage, thus enhancing cooperation among parts that maintains and enhances coherence and integration of organisms (Grosberg and Strathmann Reference Grosberg and Strathmann1998); and (2) somatic development from a propagule generated with recombined genetic material allows development that has been “reset” by the propagule to generate new phenotypic variants as effects of the recombined material (Bonner Reference Bonner1974; Dawkins Reference Dawkins1983). Natural selection on phenotypic variants thus results in genetic evolution.

The reproducer account loosens the grip of replicator thinking by abstracting from gene and cell demarcation criteria to broader conditions for conveying developmental capacities. Whether a particular form of material bottleneck is needed to “reset” development as a means to the evolution of complex adaptations is a dynamical question depending on empirical details. I propose to interpret material developmental scaffolding interactions, regardless of the genetic status of the scaffold, as a form of developmental “setting” (if not resetting). Each scaffolding event, to count as salient in development, facilitates acquiring, revising, or maintaining a developmental capacity a developing entity otherwise would have found difficult to acquire, revise, or maintain. When a scaffolding interaction leads to capacity acquisition, we can describe the interaction as a reproductive “event,” insofar as there is a “fusion” (scaffold + developing entity), a conferral or emergence of a new developmental capacity, and the persistence of that capacity beyond the end of the scaffolding interaction.

In a sense, each scaffolding interaction involves the production of an “offspring” (scaffold + developing entity) from “parents” (scaffold, developing entity) and the production of “grand-offspring” (released scaffold, developing entity with new capacity). Compare this to adult humans as parents, gametes as offspring, and the gamete-fused zygote as grand-offspring. Transmission genetics does not always do its accounting this way, but Herron (Reference Herron2016) makes a forceful argument that failure to represent the neglected gamete phase or “generation” misses out on important aspects of evolutionary dynamics in many cases of complex life cycles. I add to Herron’s argument that developmental scaffolding processes are reproduction according to the reproducer account. They fulfill the further conditions for inheritance processes along the lines of some accounts of epigenetic inheritance (Jablonka and Lamb Reference Jablonka and Lamb2005) and developmental systems theory (Oyama Reference Oyama1985/2000; Griffiths and Gray Reference Griffiths and Gray1994) insofar as the interactions between scaffolds and developers are evolved. They will rarely qualify as replicator systems because few developmental scaffolds function as template coding mechanisms. In their account of evolutionary transitions, Maynard Smith and Szathmáry (Reference Maynard Smith and Szathmáry1995) accorded replicator status only to nucleic acid coding and language-based communication (Griesemer Reference Griesemer2000b). In order not to obliterate all traditional usage, let us call the generations marked by developmental scaffolding events “developmental generations” and acknowledge that these demarcations are of reproduction processes of an unorthodox kind where one of the parents is an organism by the lights of traditional, contemporary biology, and the other is a scaffold that may be quite disparate in kind: organism, physical “prosthesis,” artifact, or substrate.

The formation, transformation, and dissolution of such hybrid entities conforms to the reproducer account sketched above, provided that we view the interaction of scaffold and developing entity in the material terms of a fusion and their separation as a fission, such that each fusion–fission pair involves a material overlap and some developmental capacity is acquired by at least one of the fission products in the process. A consequence is that there are “developmental generations” of progenerated “developmental individuals” insinuated as stages within a single “reproductive” organism. This contrasts with traditional notions of organisms as genetic entities, in which “one or two parent organisms are considered the genetic ancestors of each offspring organism” (Griesemer Reference Cartwright2014b, 198).

While a single-cell bottleneck is sufficient for developmental resetting on Dawkins’s presupposition that genes are the sole biological replicators, it is neither necessary nor sufficient if other channels of reproduction, inheritance, or replication function as scaffolding regulators or facilitators of development.

4. Conclusion: Toward a New Perspective on Developmental Reaction Norms

I have proposed that developmental scaffolding interactions take the form of reproduction processes and that some developmental stages can be interpreted as distinct developmental generations distinguishing developmental individuals within life trajectories. To the extent that modes and varieties of scaffolding events coincide in a few, major scaffolding interactions, developmental individuals are robust, having a high “degree” of individuality, as when morphological, habitual, and behavioral interactions distinguish the developmental individuality of pelagic larvae from sessile adults. This interpretation of scaffolded development as reproduction further suggests that development involves a sequence of developmental environments, marked by scaffolding interactions.

Traditionally, developmental stages are demarcated in terms of observable changes in morphology (such as topological changes in tissue organization in embryogenesis), habit (such as aquatic to terrestrial), behavior (such as nesting behavior), or even conventional clock time (as in normal tables of embryonic development). This article supports an additional strategy for demarcating developmental stages in terms of developmental scaffolding interactions. If each new developmental capacity marks a stage, then stage transitions are marked by scaffolding interactions. Since I have called these stages “developmental generations,” and because scaffolds can be viewed initially as parts of the environments of the developing entities, a further possibility is opened up by this account: developmental change due to scaffolding interactions can be interpreted in terms of reaction norm diagrams.

Models of development are often taken to map genotypes into phenotypes (Lewontin Reference Lewontin1974; Wagner and Altenberg Reference Wagner and Altenberg1996). This abstraction worked well enough when genetics was treated as a conceptual foundation rather than an empirical tool and genetic experiments regulated developmental variability by controlling environmental factors to reveal clear genotype/phenotype relationships. But it is now evident that plasticity of response across alternative environments, instability of response to constant environments, and variable ontogenetic trajectories over sequences of developmental environments are important properties of organisms (Gilbert and Epel Reference Gilbert and Epel2009; Sultan Reference Sultan2015). Reaction norm models extend genotype/phenotype mapping ecologically by considering alternative environments in which organisms might develop (Schlichting and Pigliucci Reference Schlichting and Pigliucci1998). A modest extension considers alternative sequences of developmental environments marked by any of a broad class of scaffolding interactions to articulate a developmental reaction norms perspective from which to model scaffolded development of hybrid reproducers within developmental systems.

Conventional reaction norm diagrams map genotypes in alternative environments into phenotypes. Schlichting and Pigliucci (Reference Schlichting and Pigliucci1998) review the evidence that environments change over the course of development and thus argue that reaction norms must consider sequences of different, changing developmental environments in order to model variable ontogenetic responses. I have argued here for an interpretation of developmental stage transitions in terms of scaffolding interactions that form and dissolve hybrids. This suggests a particular way to diagram reaction norms by considering developmental environments as sequences marked by scaffolding events mapping developmental resources such as genotypes into phenotypic trajectories. The alternative environments of conventional reaction norm diagrams can be interpreted as alternative sequences of developmental environments.

One way to evaluate scaffolding interactions in evolutionary explanations is to interpret scaffolds as fitness modulators (Bickhard Reference Valsiner2005). Scaffolds lower fitness costs associated with phenotypic transitions in development, altering developmental trajectories from what they otherwise would have been, such as when teaching or juvenile play, protected by watchful adults, lowers the fitness cost of learning. Fitness modulation offers a single currency through which one might model developmental reaction norms structured by scaffolding interactions.

On this developmental reaction norms view of life, all life cycles are complex, involving sequences of scaffolding developmental environments reproducing developmental individuals in distinct developmental generations. Progeneration individuates numerically distinct life trajectories in reproductive generations different in character but not necessarily in kind from developmental generations. It does not follow that traditional population genetic models cannot track sequences of reproductive generations, but it does follow that there can be nontraditional population-developmental evolution models tracking sequences of developmental generations. Because development, on the reproducer account, is acquisition of a reproductive capacity, conceptually it matters not whether the realization takes one or several turns of a traditionally interpreted complex to count as a unit in dynamic models of scaffolded reproduction, development, or evolution. Developmental reaction norms that track scaffolding events modulating fitness may be a promising eco-devo approach to complex life cycle evolution.