Coda: The View From Where We Are

The first reading, Development: A Wider View, began with a familiar fact that becomes a problem once we try to explain it. Development is remarkably reliable, yet no two individuals develop in exactly the same way. Human infants do not develop just anything, and broad patterns recur. Infants reach familiar capacities at different ages and by different routes. As the first reading put it, we often handle this by quietly dividing the explanatory work: biology or maturation accounts for what is reliable, while experience and context account for variation. But the readings have pushed toward a different possibility. In each individual, the reliable pattern and the variation are not separate products with separate causes. They are produced together.

A second problem followed from this one. We can try to identify factors that predict later outcomes, such as genetic variants, brain measures, parenting, and schooling. Those predictors may be useful without explaining how a person came to be organized as they are. In the same way, calling something an influence on development does not yet tell us what it did within a changing developmental process.

The opening example of an infant reaching for a toy put these questions into a concrete form. Over the first year, posture, looking, arm movement, and hand use come together in new ways until the infant can pick up a small object between thumb and index finger. How does that new organization come about? Why does a recognizable pattern recur while the path differs from one infant to another? And how does what has already developed shape what can happen next?

The readings have changed what we should expect from an answer. We cannot explain development simply by naming another cause to put beside genes or environment.

Many people come to these questions with a version of the same answer, because it is accessible and it makes sense. Genes build the basic structure, and the environment adjusts what genes provide. Nature and nurture interact, and maybe nature provides the basics and nurture fills in the details. That answer is trying to capture something real about both reliability and variability, and development does depend on inherited resources as well as on the conditions in which it happens.

The problem is with the way the answer divides things up. Once genes and environment are treated as separate sources, the developing organism becomes the place where their contributions get added together. The readings in this resource have come at the problem from a different direction, asking how an already organized living system changes over time and how each change alters the conditions for what comes next.

We could add more molecular detail, but this does not get us away from the basic problem, because we can explain transcription, translation, and gene regulation in considerable detail and still fall back on the blueprint when someone asks what genes actually do in development. Experience can be described as modifying a structure that genes already supplied, and we can even talk about interactions without disturbing the basic division. The picture of separation that we started with ends up organizing how we interpret the details.

If we want to question the separation, we have to go back to what we have treated as the source of organization. Then we need to think about what we have left for development itself to explain. The readings suggest a different way of building an explanation, one that starts with the reconstruction of a life cycle and follows development as a trajectory rather than as a route toward an endpoint set in advance. Along the way, the organism turns out to be a participant in its own development, not just the place where outside influences meet. From there the implications reach well beyond embryos and brains, into how we think about human lives.

Reconstructing a Life Cycle

What gets repeated across generations is not an adult passed forward like a copy, but a whole cycle. The frog life cycle of egg, tadpole, adult, egg keeps going because it is rebuilt each time from resources that earlier cycles produced, including DNA, the organization of the egg, what the mother contributes, and the physical properties of tissues. DNA is very important, but it is not standing outside the process holding a description of the frog that is to come.

Nor does a new frog start out as unorganized material waiting for DNA to tell it what to do, since the fertilized egg already has membranes, proteins, a cytoskeleton, and a spatial layout inherited from the egg. Which genes get expressed first depends on that organization. As cells divide, each round of activity changes the conditions for the next: cells respond to their neighbors, to chemical gradients, and to mechanical forces, and their responses change what the next cells encounter. DNA is involved throughout, but the system is already organized, and DNA is one of the resources that the organized system draws on.

The frog organizer discussed in Brain, Body, World makes the same point in a more concrete way. If you ask where the organizer comes from, you are led to a neighboring region of the embryo that induced it. That region depended on changes in the distribution and activity of molecules already present in the egg after fertilization, and those molecules had accumulated or become localized as the egg formed in the mother’s body. You can keep asking what came before, but you never arrive at a set of instructions, only one organized structure acting on another. There is no first link in the chain because it was never really a chain, but a loop that runs back into the previous generation.

So what needs explaining is not how a stored plan gets carried out but how the conditions that make an organized developmental process possible get brought together and kept going. Those conditions constrain what can happen without specifying the outcome. Think of a twelve-month-old learning to walk. Earlier attempts have changed what the infant can do, but keeping balance also depends on the surface underfoot. A step that succeeds once may fail on the next attempt. As the infant moves, the space of possibilities changes, and each attempt leaves the infant organized a little differently for the next one.

In the people with six-fingered hands that Mehring and colleagues studied, discussed in Plasticity and Constraint, the extra finger had its own muscles and nerves, and the cortex developed a separate representation of it. That map was not simply written by input from the hand, because the cortex already had an organization and a history of its own. But there was also no instruction anywhere to map five fingers. Activity from the developing hand met a developing nervous system whose possibilities were structured but not closed, and the map organized itself around the hand that was actually there.

When a change in conditions occurs can matter as much as what changes, because early on, cells and circuits respond to conditions that will have much less effect later. Spemann could move tissue in an early newt embryo and it would take on the identity of its new neighbors, whereas later on the same tissue kept going with what it had already been becoming. Critical periods in vision work in a similar way. Early removal of dense congenital cataracts can reduce the risk of lasting visual problems, while an adult who develops cataracts can often regain good vision after surgery.

We often call this a loss of plasticity, but I think that description misses something, because the system is not so much losing capacity as settling into an organized state. A system that stayed open to everything would never get organized at all.

Reliability and Variability

Frog life cycles recur with extraordinary dependability, but reconstruction never happens under exactly the same circumstances twice, and each frog develops through its own sequence of encounters. Within the development of an individual, reliability and variability therefore do not need different explanations. The same developmental process can produce both because it is constrained enough for recognizable forms to recur while remaining responsive to the particular conditions and history through which each individual develops. Variation is part of the process from the start, not a deviation added afterward to a standard path.

This is a claim about how individuals develop, not about populations. Asking why outcomes are distributed as they are across a population, and which genetic or environmental differences go along with differences in those outcomes, is a legitimate question, but it is a different question, and answering it does not tell us how the organization of any particular individual came about (see Developmental Systems Theory).

In Gottlieb’s studies, newly hatched ducklings that had never met a mother still preferred their own species’ maternal call, so it seemed reasonable to call the preference innate. But duckling embryos vocalize in the egg before they hatch, and when Gottlieb prevented them from doing that, the preference changed. Under ordinary conditions nobody would have noticed, because every normally developing duckling hears its own calls, and it took removing that condition to show that it mattered.

Someone defending innateness could reply that “innate” never meant independent of all conditions, only that something develops reliably under the conditions a species normally encounters. That is a fair point about how the word gets used, but it leaves the developmental question where it was. Calling the preference innate does not tell us how it came about.

In the developing brain, waves of spontaneous activity sweep across the retina before mature patterned vision develops, helping to organize the visual pathways, and fetal movements generate patterned activity from muscles, joints, and skin that is thought to help organize the body map in the cortex. Neither of these fits neatly on one side of an intrinsic/extrinsic divide, since this is the nervous system’s own activity, and the body’s, organizing itself before the outside world has had much to say. The important question is not whether a source is intrinsic or extrinsic, but how a particular activity or condition becomes part of a process that is already underway and what it changes about what can happen next.

The duckling hears a call before it hatches because it makes one. The retina generates its own waves, and the fetus produces movements that are thought to help organize its body map. Some of the conditions that make development reliable are reliable because the developing organism supplies them itself. This is not the whole answer, since much of the body’s organization is laid down before activity of this kind begins, but it is an answer that does not simply assume the recurrence it is meant to explain.

Reliability therefore does not need the outcome to be written down in advance, but it does depend on constraints that keep recurring across many levels and timescales, from molecular interactions and the physical properties of tissues up through the activity of the organism, its relations with other organisms, and its evolutionary history. In humans, cultural practices, tools, languages, and institutions also reshape the conditions each new generation encounters. These conditions and relations do not matter equally to every outcome, and saying that they form a system does not explain any particular case. We have to identify the relations involved in a particular case and explain how they keep recurring or change.

The conditions that support a life cycle do not have to keep recurring, and in that sense life cycles are contingent: they depend on certain conditions being present, and they could be otherwise. Most species that have ever lived are now extinct, and their life cycles went with them once the conditions they depended on were no longer there. Oyama, Griffiths, and Gray (2001) called their edited book on developmental systems theory Cycles of Contingency. This title captures the central idea very well: Life cycles can be remarkably stable without being inevitable, and their stability is something achieved in each generation rather than evidence that the future was already contained in the beginning.

Development Has a History

Early in these readings we contrasted a coin that lands heads ten times with a developing system. The coin’s earlier tosses are part of its history, but they are not carried forward into the next toss. In development, what has already happened is retained in the organization of the system and changes what can happen next, and that is the sense in which I think developmental explanation is historical.

When people hear the word “history” they often think of a list of events in order, which for an embryo would be zygote, blastula, gastrula, neurula, and so on. That sequence is important, but listing it is not an explanation. Historians ask why events took the course they did and what could have gone differently. For development, this means asking how activity produced a change under particular constraints, and what that change made possible next. You cannot explain gastrulation without the cell movements and the signaling between tissues, so mechanisms remain important. The historical question places those mechanisms inside the trajectory where they arise and take on their role.

Looking backward from the present can also mislead us, because similar outcomes can be reached along different paths, and a small difference early on can be amplified, redirected, or absorbed as the system changes. When we look back from an adult characteristic, it is easy to see development as having been headed toward that characteristic all along. But development does not know where it is going. At each point it proceeds from what is available then, not from the endpoint that we can see afterward.

Development and evolution come together at this point, because natural selection does not by itself generate the forms it acts on; development does. The Evolution and Development reading drew on Scott Gilbert’s line that development is the artist and natural selection is the curator. The metaphor is not perfect, since development is not an independent creator and selection does not stand outside the process, picking among finished products. But the metaphor corrects an imbalance: Before a variant can persist across generations, a developmental system has to be able to produce it.

The axolotl becomes reproductively mature while keeping features that, in related salamanders, belong to an earlier phase of development. This is heterochrony, a change in the relative timing of developmental processes, and it shows that evolution can change what gets produced by changing when or for how long processes that are already underway keep running, without needing a new instruction for the resulting form.

The bat-wing work discussed in Evolution and Development gives a smaller example: swapping in the bat version of a regulatory sequence for one growth-related gene made mouse forelimbs slightly longer just before birth, without coming close to making a wing. In both cases, a change enters a process that is already organized, and what it does depends on the rest of that process.

Organisms also change the conditions they develop in, as when beavers build dams, or when humans cook food, make tools, and build practices and institutions that shape how the next generation develops. These are not part of an outer environment acting on a finished organism, but become part of the developmental system that gets reconstructed over time.

Lactase persistence is a familiar example. The ability to digest milk into adulthood only mattered once people were drinking milk, so dairying changed which genetic variants made a difference to people’s lives. The cultural practice preceded the widespread increase in the relevant genetic variants.

How selection worked is more complicated than the textbook story. Evershed and colleagues (2022) found that milk was used in prehistoric Europe for thousands of years before lactase persistence became common. They propose that the strongest selection may have come during famines and outbreaks of disease, when drinking milk without the variant was more dangerous. That complication strengthens the point, because what the variant did depended on the conditions of the lives it entered.

Seen historically, a pattern can stay stable because the conditions that sustain it keep recurring, not because it is fixed in its nature. Change does not need a complete break with the past either. It happens when activity alters some part of the organization or the conditions of reconstruction, which can open up possibilities that were not there before or close off paths that once were. What looks new often carries a lot of the past forward in a different arrangement.

Human Lives

Turning to human development after this long biological route should not suggest that everything so far was just preparation for the human case, because human lives were developmental all along. What changes is the set of assumptions we bring when we try to explain them.

By a developmental trajectory I do not just mean a series of measurements over time, but the history that gets produced as development moves from one organization to the next. Activity changes the developing person, and it can change their surroundings too, and the result becomes part of the starting point for what follows. Earlier development is not left behind, but it is not preserved unchanged either. It is carried forward in the ways it has reorganized the person and their relations with the world.

Some decades ago, Annette Karmiloff-Smith (1998) made this point about developmental disorders when she argued that performance that looks similar does not mean the organization behind it developed in the same way. A capacity that seems relatively spared in an older child or adult might be read as a normally developed module sitting alongside impaired ones, but an early difference can change the development of several systems and how they relate to each other. An adult profile cannot tell us by itself which paths produced it.

In the default framework, a developmental disorder is often treated as a mistake in the genetic program: a gene is changed, something is missing, and development goes wrong. Clinical genetics already knows this is too simple, since the same variant can lead to very different outcomes in different people.

The more basic question is what a diagnosis is taken to explain. From the perspective of these readings, a developmental difference is a different path through a space of possibilities. That does not make it any less real, and it does not mean every path is equally good for the person taking it. A developmental path can involve serious suffering or limit what a person can do. Recognizing that cost does not require treating a typical outcome as the destination fixed in advance. What the framework rules out is treating distance from an average as failure in itself.

A diagnosis can identify a present pattern without explaining how that pattern developed in this particular person. Similar difficulties can rest on different underlying organization, and the same early difference can lead into several possible trajectories. Genetic findings, neural mechanisms, behavioral assessments, and what we know about relationships and social conditions can all help, but none of them makes the developmental question go away. What remains is to explain how those difficulties arose in this person and what now keeps them going.

Any intervention enters a history that is already underway, so its effects depend on timing, on what has already developed, and on the relationships and conditions through which it is taken up. The same intervention can have different consequences for different people, or for the same person at different points in life. Cataract surgery is a simple case, since the same operation means something quite different in early infancy than it does in adulthood. I do not think this makes the practical questions less scientific. It gives us specific questions about what an intervention changes in the person’s activity and how those changes can be sustained.

Human developmental histories extend beyond the individual. Infants arrive into ways of living that already have long histories, with opportunities that differ sharply between families and communities. Children take part in these arrangements and sometimes change them, but they do not begin from conditions of their own choosing. A developmental account of a human life therefore has to connect a person’s activity with the historically organized relations that make that activity possible and meaningful. The biological and the social are not separate layers that can be finished independently and then put together. They meet in the life of a developing person.

What Is Not Settled

We should be clear that none of this gives us a new overarching explanation to replace genes, environment, natural selection, or the brain. A “developmental systems” approach can become just as empty an answer as “genetic program” if naming the system takes the place of explaining how a particular organization is produced. Talking about the importance of constraints only helps if we can say how they structure a process, and historical explanation needs more than the observation that everything has a past. Appealing to activity explains very little unless we can say what is acting and what its activity changes.

A developmental explanation has to account for how one organization becomes another through activity that changes the conditions for what happens next. That changes where we are willing to stop an explanation, because every explanation has to start from something it does not itself explain. In development that starting point is always an organization that already exists, whether a cell, a body, or a nervous system with a history of its own, and there is nothing wrong with that.

What an explanation cannot do is start from a description of the very thing whose development it is supposed to explain, whether that description is placed in a gene, in a module, or in a mechanism characterized by the job it will eventually do (Marshall, 2026). The little person in the sperm failed this test, and so does the blueprint, since both put a description of the outcome at the beginning.

Where exactly that line falls is not always easy to say, and it is not settled here. Gene regulatory networks that reliably lead to particular cell fates, or molecular gradients that lay out broad regions of the cortex, are legitimate starting points for explaining later steps. The question in each case is whether the starting point is being used to explain a transformation, or whether it is the later organization described again and placed earlier.

A reliable developmental outcome therefore still requires an account of how the conditions producing it recur, and a strong genetic association still acquires its consequences through an organized developmental process. An evolutionary account of why a characteristic may have been favored does not by itself explain how that characteristic develops in an individual, and showing that a neural circuit is necessary for a behavior now does not explain how that circuit came to be organized in the developing individual. None of this is a reason to abandon genetic, evolutionary, or mechanistic research, but it is a reason to ask what question each kind of evidence answers and what remains to be explained.

These readings have also only taken the argument so far, staying close to biological and neural development, with most of the brain material coming from sensory and motor cortex. Other parts of the cortex, and capacities like language, may develop in quite different ways. The readings have said much less about how cultural practices, institutions, relationships, and participation with other people enter into the organization of human lives. To develop a fuller account of human development, we would need to follow these relations as closely as we have followed embryonic and neural processes.

From Development to Life

The first section of About Development covered some big questions: why development recurs so reliably, what gives it direction, how a history is carried forward in a changing individual, and how processes capable of producing an organized individual arose at all. The readings could not answer those questions, but they changed what an answer would look like. Reliability comes from conditions that keep being reconstructed; direction from constraints and from the organism’s own activity, which together narrow and redirect what can happen next; history from an organism that is itself changed by what it does; and the capacity for all of this from a long evolutionary history of bodies that hold themselves together.

Go back to the frog’s egg one last time. DNA was there at the beginning, along with a cell, a history, and a life cycle already underway. None of these contained a miniature adult or a complete description of the path ahead. Through activity that kept changing the conditions for what came next, the developing frog drew on forms of organization much older than itself and produced a life that had never happened before.

Living organization keeps itself going through activity, including neural activity. But nervous systems evolved in bodies that were already moving, and developing nervous systems generate their own activity before any patterned input from the world arrives. The brain is not simply taking in information. What a stimulus does depends on the organization it meets, and that organization is already in motion.

The word “activity” has been used quite loosely in these readings, for cells dividing and migrating as well as for a duckling calling in the egg or an infant reaching for a toy. Activity happens at every level of a developing organism, but only some of it is at the level of the organism that maintains itself and engages with a world. Developmental systems theory tells us why the organism cannot be separated from the relations through which it develops, but it does not, on its own, explain what makes those relations matter to the organism. What difference does it make that the system whose history we are explaining has to keep maintaining the organization that makes its activity possible? How does activity come to have a perspective, from which some things matter and others do not?

We do not make the future from nothing, and we do not receive it already formed. Living things proceed from an organization and a history, and in acting from where they are they change what can happen next. What kind of being can do this, and how a world comes to have significance from inside such a life, are the questions taken up in a separate resource, Enacting Life.

Sources

Studies discussed again here, including Gottlieb’s ducklings, Spemann’s embryological experiments, and Mehring and colleagues’ six-fingered hands, are cited in the readings where they first appear.

Evershed, R. P., Davey Smith, G., Roffet-Salque, M., Timpson, A., Diekmann, Y., Lyon, M. S., Cramp, L. J. E., Casanova, E., Smyth, J., Whelton, H. L., Dunne, J., Brychova, V., Šoberl, L., Gerbault, P., Gillis, R. E., Heyd, V., Johnson, E., Kendall, I., Manning, K., … Thomas, M. G. (2022). Dairying, diseases and the evolution of lactase persistence in Europe. Nature, 608, 336–345.https://doi.org/10.1038/s41586-022-05010-7

Karmiloff-Smith, A. (1998). Development itself is the key to understanding developmental disorders. Trends in Cognitive Sciences, 2(10), 389–398. https://doi.org/10.1016/S1364-6613(98)01230-3

Marshall, P. J. (2026). Psychology in 2125: A developmental science of living sense-making. Perspectives on Psychological Science. https://doi.org/10.1177/17456916261477319

Oyama, S., Griffiths, P. E., & Gray, R. D. (Eds.). (2001). Cycles of contingency: Developmental systems and evolution. MIT Press.