Languages are transmitted from person to person and generation to generation via a process of iterated learning: people learn a language from other people who once learned that language themselves. We analyze the consequences of iterated learning for learning algorithms based on the principles of Bayesian inference, assuming that learners compute a posterior distribution over languages by combining a prior (representing their inductive biases) with the evidence provided by linguistic data. We show that when learners sample languages from this posterior (...) distribution, iterated learning converges to a distribution over languages that is determined entirely by the prior. Under these conditions, iterated learning is a form of Gibbs sampling, a widely-used Markov chain Monte Carlo algorithm. The consequences of iterated learning are more complicated when learners choose the language with maximum posterior probability, being affected by both the prior of the learners and the amount of information transmitted between generations. We show that in this case, iterated learning corresponds to another statistical inference algorithm, a variant of the expectation-maximization (EM) algorithm. These results clarify the role of iterated learning in explanations of linguistic universals and provide a formal connection between constraints on language acquisition and the languages that come to be spoken, suggesting that information transmitted via iterated learning will ultimately come to mirror the minds of the learners. (shrink)

There are several proposals to resolve the problem of epistemic peer disagreement which concentrate on the question of how to incorporate evidence of such a disagreement. The main positions in this field are the equal weight view, the steadfast view, and the total evidence view. In this paper we present a new argument in favour of the equal weight view. As we will show, this view results from a general approach of forming epistemic attitudes in an optimal way. By this, (...) the argument for equal weighting can be massively strengthened from reasoning via epistemic indifference to reasoning from optimality. (shrink)

This paper examines conceptual issues that arise in applications of Darwinian natural selection to cultural systems. I argue that many criticisms of cultural selectionist models have been based on an over-detailed reading of the analogy between biological and cultural units of selection. I identify five of the most powerful objections to cultural selection theory and argue that none cuts to its heart. Some objections are based on mistaken assumptions about the simplicity of the mechanisms of biological heredity. Other objections are (...) attributable, rather, to mistaken inferences from observations of biological subject matter to what is essential in natural selection. I argue that such features are idiosyncratic of biological systems, but not essential for natural selection. My arguments throughout are illustrated by examples from biological and cultural evolution, and counter-factual illustrations from the history of theoretical biology. Introduction Cultural Selection Theory First Objection: Lamarckianism Second Objection: Genotype–Phenotype Distinction Third Objection: Common Hereditary Architecture Fourth Objection: Biological Analogue for Cultural Units 6.1 Regarding strict analogues 6.2 Regarding the trait analogue 6.3 Regarding the virus analogue Fifth Objection: Environmental Interaction Conclusion CiteULike Connotea Del.icio.us What's this? (shrink)

Recent debates about memetics have revealed some widespread misunderstandings about Darwinian approaches to cultural evolution. Drawing from these debates, this paper disputes five common claims: (1) mental representations are rarely discrete, and therefore models that assume discrete, gene-like particles (i.e., replicators) are useless; (2) replicators are necessary for cumulative, adaptive evolution; (3) content-dependent psychological biases are the only important processes that affect the spread of cultural representations; (4) the “cultural fitness” of a mental representation can be inferred from its successful (...) transmission; and (5) selective forces only matter if the sources of variation are random. We close by sketching the outlines of a unified evolutionary science of culture. (shrink)

This article suggests a ‘best alternative' justification of induction (in the sense of Reichenbach) which is based on meta-induction . The meta-inductivist applies the principle of induction to all competing prediction methods which are accessible to her. It is demonstrated, and illustrated by computer simulations, that there exist meta-inductivistic prediction strategies whose success is approximately optimal among all accessible prediction methods in arbitrary possible worlds, and which dominate the success of every noninductive prediction strategy. The proposed justification of meta-induction is (...) mathematically analytical. It implies, however, an a posteriori justification of object-induction based on the experiences in our world. *Received November 2005; revised March 2008. †To contact the author, please write to: Department of Philosophy, University of Duesseldorf, Universitaetsstrasse 1, Geb. 23.21, Duesseldorf, Germany D-40225; e-mail: [email protected]. (shrink)

While evolutionary thinking is increasingly becoming popular in fields of investigation outside the biological sciences, it remains unclear how helpful it is there and whether it actually yields good explanations of the phenomena under study. Here we examine the ontology of a recent approach to applying evolutionary thinking outside biology, the generalized Darwinism approach proposed by Geoffrey Hodgson and Thorbjørn Knudsen. We examine the ontology of populations in biology and in GD, and argue that biological evolutionary theory sets ontological criteria (...) that GD fails to meet. We suggest two options to revise the population concept in GD: reformulating the concept in terms of inheritance and reproduction such that it comes to pick out individuals similar to evolving populations, or trying to build an adequate population concept on a principle of differential retention instead of differential reproduction. 1 Introduction2 Generalized Darwinism2.1 What is generalized Darwinism?2.2 Darwinian principles3 The Ontology of Generalized Darwinism: What Are Populations?3.1 The population concept of generalized Darwinism3.2 The population concept in evolutionary theory4 Locating Evolving Systems in Generalized Darwinism5 Conclusion. (shrink)

Fitness is a central concept in evolutionary theory. Just as it is central to biological evolution, so, it seems, it should be central to cultural evolutionary theory. But importing the biological fitness concept to CET is no straightforward task—there are many features unique to cultural evolution that make this difficult. This has led some theorists to argue that there are fundamental problems with cultural fitness that render it hopelessly confused. In this essay, we defend the coherency of cultural fitness against (...) those who call it into doubt. (shrink)