The hard problem of meta-learning is what-to-learn.

Binz et al. highlight the potential of meta-learning to greatly enhance the flexibility of AI algorithms, as well as to approximate human behavior more accurately than traditional learning methods. We wish to emphasize a basic problem that lies underneath these two objectives, and in turn suggest another perspective of the required notion of "meta" in meta-learning: knowing what to learn.

Francis Galton's regression towards mediocrity and the stability of types.

A prevalent narrative locates the discovery of the statistical phenomenon of regression to the mean in the work of Francis Galton. It is claimed that after 1885, Galton came to explain the fact that offspring deviated less from the mean value of the population than their parents did as a population-level statistical phenomenon and not as the result of the processes of inheritance. Arguing against this claim, we show that Galton did not explain regression towards mediocrity statistically, and did not give up on his ideas regarding an inheritance process that caused offspring to revert to the mean. While the common narrative focuses almost exclusively on Galton's statistics, our arguments emphasize the anthropological and biological questions that Galton addressed. Galton used regression towards mediocrity to support the claim that some biological types were more stable than others and hence were resistant to evolutionary change. This view had implications concerning both natural selection and eugenics. The statistical explanation attributed to Galton appeared later, during the biometrician-mutationist debate in the early 1900s. It was in the context of this debate and specifically by the biometricians, that the development of the statistical explanation was originally attributed to Galton.

Richard Lewontin and the "complications of linkage".

During the 1960s and 1970s population geneticists pushed beyond models of single genes to grapple with the effect on evolution of multiple genes associated by linkage. The resulting models of multiple interacting loci suggested that blocks of genes, maybe even entire chromosomes or the genome itself, should be treated as a unit. In this context, Richard Lewontin wrote his famous 1974 book The Genetic Basis of Evolutionary Change, which concludes with an argument for considering the entire genome as the unit of selection as a result of linkage. Why did Lewontin and others devote so much intellectual energy to the "complications of linkage" in the 1960s and 1970s? We argue that this attention to linkage should be understood in the context of research on chromosomal inversions and co-adapted gene complexes that occupied mid-century evolutionary genetics. For Lewontin, the complications of linkage were an extension of this chromosomal focus expressed in the new language of models for linkage disequilibrium.

Friedrich Miescher's Discovery in the Historiography of Genetics: From Contamination to Confusion, from Nuclein to DNA.

In 1869, Johann Friedrich Miescher discovered a new substance in the nucleus of living cells. The substance, which he called nuclein, is now known as DNA, yet both Miescher's name and his theoretical ideas about nuclein are all but forgotten. This paper traces the trajectory of Miescher's reception in the historiography of genetics. To his critics, Miescher was a "contaminator," whose preparations were impure. Modern historians portrayed him as a "confuser," whose misunderstandings delayed the development of molecular biology. Each of these portrayals reflects the disciplinary context in which Miescher's work was evaluated. Using archival sources to unearth Miescher's unpublished speculations-including an analogy between the hereditary material and language, and a speculation that a series of asymmetric carbon atoms could account for hereditary variation-this paper clarifies the ways in which the past was judged through the lens of contemporary concerns. It also shows how organization, structure, function, and information were already being considered when nuclein was first discovered nearly 150 years ago.

Systems Thinking Versus Population Thinking: Genotype Integration and Chromosomal Organization 1930s-1950s.

This article describes how empirical discoveries in the 1930s-1950s regarding population variation for chromosomal inversions affected Theodosius Dobzhansky and Richard Goldschmidt. A significant fraction of the empirical work I discuss was done by Dobzhansky and his coworkers; Goldschmidt was an astute interpreter, with strong and unusual commitments. I argue that both belong to a mechanistic tradition in genetics, concerned with the effects of chromosomal organization and systems on the inheritance patterns of species. Their different trajectories illustrate how scientists' commitments affect how they interpret new evidence and adjust to it. Dobzhansky was moved to revised views about selection, while Goldschmidt moved his attention to different genetic phenomena. However different, there are significant connections between the two that enrich our understanding of their views. I focus on two: the role of developmental considerations in Dobzhansky's thought and the role of neutrality and drift in Goldschmidt's evolutionary account. Dobzhansky's struggle with chromosomal variation is not solely about competing schools of thought within the selectionist camp, as insightfully articulated by John Beatty, but also a story of competition between selectionist thinking and developmental perspectives. In contraposition, Goldschmidt emphasized the role of low penetrance mutations that spread neutrally and pointed out that drift could result from developmental canalization. This account adds to the dominant story about Goldschmidt's resistance to the splitting of development from genetics, as told by Garland Allen and Michael Dietrich. The story I tell illustrates how developmental thinking and genetic thinking conflicted and influenced researchers with different convictions about the significance of chromosomal organization.

The genome as a developmental organ.

This paper applies the conceptual toolkit of Evolutionary Developmental Biology (evo-devo) to the evolution of the genome and the role of the genome in organism development. This challenges both the Modern Evolutionary Synthesis, the dominant view in evolutionary theory for much of the 20th century, and the typically unreflective analysis of heredity by evo-devo. First, the history of the marginalization of applying system-thinking to the genome is described. Next, the suggested framework is presented. Finally, its application to the evolution of genome modularity, the evolution of induced mutations, the junk DNA versus ENCODE debate, the role of drift in genome evolution, and the relationship between genome dynamics and symbiosis with microorganisms are briefly discussed.

The interplay of social identity and norm psychology in the evolution of human groups.

People's attitudes towards social norms play a crucial role in understanding group behaviour. Norm psychology accounts focus on processes of norm internalization that influence people's norm-following attitudes but pay considerably less attention to social identity and group identification processes. Social identity theory in contrast studies group identity but works with a relatively thin and instrumental notion of social norms. We argue that to best understand both sets of phenomena, it is important to integrate the insights of both approaches. Social status, social identity and social norms are considered separate phenomena in evolutionary accounts. We discuss assumptions and views that support this separation, and suggest an integrated view of our own. We argue that we should be open to the early origins of human social complexity, and conjecture that the longer that the human social world involved multi-level societies the more probable it is that norm psychology and social identity interacted in rich ways. This article is part of the theme issue 'Human socio-cultural evolution in light of evolutionary transitions'.

Human major transitions from the perspective of distributed adaptations.

Distributed adaptations are cases in which adaptation is dependent on the population as a whole: the adaptation is conferred by a structural or compositional aspect of the population; the adaptively relevant information cannot be reduced to information possessed by a single individual. Possible examples of human-distributed adaptations are song lines, traditions, trail systems, game drive lanes and systems of water collection and irrigation. Here we discuss the possible role of distributed adaptations in human cultural macro-evolution. Several kinds of human-distributed adaptations are presented, and their evolutionary implications are highlighted. In particular, we discuss the implications of population size, density and bottlenecks on the distributed adaptations that a population may possess and how they in turn would affect the population's resilience to ecological change. We discuss the implications that distributed adaptations may have for human collective action and the possibility that they played a role in colonization of new areas and niches, in seasonal migration, and in setting constraints for minimal inter-population connectivity. This article is part of the theme issue 'Human socio-cultural evolution in light of evolutionary transitions'.

A third way to the selected effect/causal role distinction in the great encode debate.

Since the ENCODE project published its final results in a series of articles in 2012, there is no consensus on what its implications are. ENCODE's central and most controversial claim was that there is essentially no junk DNA: most sections of the human genome believed to be «junk¼ are functional. This claim was met with many reservations. If researchers disagree about whether there is junk DNA, they have first to agree on a concept of function and how function, given a particular definition, can be discovered. The ENCODE debate centered on a notion of function that assumes a strong dichotomy between evolutionary and non-evolutionary function and causes, prevalent in the Modern Evolutionary Synthesis. In contrast to how the debate is typically portrayed, both sides share a commitment to this distinction. This distinction is, however, much debated in alternative approaches to evolutionary theory, such as the EES. We show that because the ENCODE debate is grounded in a particular notion of function, it is unclear how it connects to broader debates about what is the correct evolutionary framework. Furthermore, we show how arguments brought forward in the controversy, particularly arguments from mathematical population genetics, are deeply embedded in their particular disciplinary contexts, and reflect substantive assumptions about the evolution of genomes. With this article, we aim to provide an anatomy of the ENCODE debate that offers a new perspective on the notions of function both sides employed, as well as to situate the ENCODE debate within wider debates regarding the forces operating in evolution.

Human socio-cultural evolution in light of evolutionary transitions: introduction to the theme issue.

Human societies are no doubt complex. They are characterized by division of labour, multiple hierarchies, intricate communication networks and transport systems. These phenomena and others have led scholars to propose that human society may be, or may become, a new hierarchical level that may dominate the individual humans within it, similar to the relations between an organism and its cells, or an ant colony and its members. Recent discussions of the possibility of this major evolutionary transition in individuality (ETI) raise interesting and controversial questions that are explored in the present issue from four different complementary perspectives. (i) The general theory of ETIs. (ii) The unique aspects of cultural evolution. (iii) The evolutionary history and pre-history of humans. (iv) Specific routes of a possible human ETI. Each perspective uses different tools provided by different disciplines: biology, anthropology, cultural evolution, systems theory, psychology, economy, linguistics and philosophy of science. Altogether, this issue provides a broad and rich application of the notion of ETI to human past, present and perhaps also future evolution. It presents important case studies, new theoretical results and novel questions for future research. This article is part of the theme issue 'Human socio-cultural evolution in light of evolutionary transitions'.

Before Watson and Crick in 1953 Came Friedrich Miescher in 1869.

In 1869, the young Swiss biochemist Friedrich Miescher discovered the molecule we now refer to as DNA, developing techniques for its extraction. In this paper we explain why his name is all but forgotten, and his role in the history of genetics is mostly overlooked. We focus on the role of national rivalries and disciplinary turf wars in shaping historical memory, and on how the story we tell shapes our understanding of the science. We highlight that Miescher could just as correctly be portrayed as the person who understood the chemical nature of chromatin (before the term existed), and the first to suggest how stereochemistry might serve as the basis for the transmission of hereditary variation.

Conceptual and methodological biases in network models.

Many natural and biological phenomena can be depicted as networks. Theoretical and empirical analyses of networks have become prevalent. I discuss theoretical biases involved in the delineation of biological networks. The network perspective is shown to dissolve the distinction between regulatory architecture and regulatory state, consistent with the theoretical impossibility of distinguishing a priori between "program" and "data." The evolutionary significance of the dynamics of trans-generational and interorganism regulatory networks is explored and implications are presented for understanding the evolution of the biological categories development-heredity, plasticity-evolvability, and epigenetic-genetic.

Hopeful heretic--Richard Goldschmidt's genetic metaphors.

Richard Goldschmidt famously rejected the notion of atomic and corpuscular genes, arranged on the chromosome like beads-on-a-string. I provide an exegesis of Goldschmidt's intuition by analyzing his repeated and extensive use of metaphorical language and analogies in his attempts to convey his notion of the nature of the genetic material and specifically the significance of chromosomal pattern. The paper concentrates on Goldschmidt's use of metaphors in publications spanning 1940-1955.

Integrating evolution and development: from theory to practice.

This volume joins a growing list of books, monographs, and proceedings from scientific meetings that attempt to consolidate the wide spectrum of approaches emphasizing the role of development in evolution into a coherent and productive synthesis, often called evo-devo. Evo-devo is seen as a replacement or amendment of the modern synthesis that has dominated the field of evolution since the 1940s and which, as even its architects confessed, was fundamentally incomplete because development remained outside its theoretical framework (Mayr and Provine 1980). As the volume attests, there is now a strong feeling that the time is ripe for the consolidation of evo-devo, and that the field is mature enough so that mapping the theoretical terrain and experimental approaches is both feasible and scientifically productive. Now is an appropriate time to try to weave the strands of reasoning leading to the developmental perspective and offer a synthesis.