T evolved independently or `convergently’ in two different clades; such traits can be Talmapimod site termed analogies (the more technical biological term `homoplasy’ refers to all shared traits that are not homologies, and includes analogy as a special case). Canonical examples of analogy include the independent evolution of wing from forelimbs in birds and bats, or the evolution of Y-27632 biological activity bipedalism (walking on two feet) in humans and birds. Neither wings nor bipedalism were present in the quadrupedal reptilian LCA of mammals and birds, but instead evolved convergently in each of these clades. Analogous and homologous traits play different roles in biology, but both are important. Homologous traits are those that are used in classification and taxonomy (for this purpose, analogous traits are just a nuisance variable). More relevant to bio-musicology, homologies often allow us to make inferences about traits that were present in an ancestral species, because a set of homologous traits in a particular clade are by definition inherited by descent from a common ancestor of that clade. Often, particularly for behavioural or cognitive capacities, homology-based phylogenetic inference is the only means we have of reconstructing these extinct ancestors, because behavioural traits typically leave no fossils (fossil footprints providing one exception). For example, although we will probably never find a fossilized Cretaceous stem mammal in the act of suckling her young, we can nonetheless infer, with great confidence, that the ancestral mammal did so from the fact that all living descendants of this species still do. Thus, a careful analysis of living species, combined with comparative inference, provides a sort of `evolutionary time machine’ to reconstruct the behaviour and physiology of long-extinct species. Analogous traits serve a different and complementary purpose: they provide a means for testing hypotheses using multiple independent data points. Although all of the more than 5000 existing species of mammals suckle their young, this ability derives from their evolutionary origin at the base of the clade, and thus statistically constitutes a single data point (not 5000). By contrast, convergently evolved traits are by definition independent evolutionary events, and each clade independently possessing a trait therefore represents an independent data point. Only a set of convergently evolved traits provides an adequate database for statistically valid tests of evolutionary hypotheses. This point is often ignored, even by biologists discussing music evolution (e.g. [23]). Fortunately, for many cases of convergent evolution, such as bipedalism or vocal learning, a trait has evolved independently enough times to provide a rich source of evidenceto test hypotheses concerning both evolution and mechanistic function. Thus, for example, we can test mechanistic hypotheses about the requirements of vocal learning by examining its neural correlates in the many species that have evolved this ability convergently (cf. [41]). Similarly, we can test functional hypotheses about why the capacity for vocal synchrony or antiphony is adaptive by examining the many bird, mammal, frog and insect species that have convergently evolved this ability [40]. While the conceptual distinction between homology and analogy is clear, recent discoveries in genetics and neuroscience suggest that in some cases a trait can be both homologous and analogous, depending on the level of explanation. For.T evolved independently or `convergently’ in two different clades; such traits can be termed analogies (the more technical biological term `homoplasy’ refers to all shared traits that are not homologies, and includes analogy as a special case). Canonical examples of analogy include the independent evolution of wing from forelimbs in birds and bats, or the evolution of bipedalism (walking on two feet) in humans and birds. Neither wings nor bipedalism were present in the quadrupedal reptilian LCA of mammals and birds, but instead evolved convergently in each of these clades. Analogous and homologous traits play different roles in biology, but both are important. Homologous traits are those that are used in classification and taxonomy (for this purpose, analogous traits are just a nuisance variable). More relevant to bio-musicology, homologies often allow us to make inferences about traits that were present in an ancestral species, because a set of homologous traits in a particular clade are by definition inherited by descent from a common ancestor of that clade. Often, particularly for behavioural or cognitive capacities, homology-based phylogenetic inference is the only means we have of reconstructing these extinct ancestors, because behavioural traits typically leave no fossils (fossil footprints providing one exception). For example, although we will probably never find a fossilized Cretaceous stem mammal in the act of suckling her young, we can nonetheless infer, with great confidence, that the ancestral mammal did so from the fact that all living descendants of this species still do. Thus, a careful analysis of living species, combined with comparative inference, provides a sort of `evolutionary time machine’ to reconstruct the behaviour and physiology of long-extinct species. Analogous traits serve a different and complementary purpose: they provide a means for testing hypotheses using multiple independent data points. Although all of the more than 5000 existing species of mammals suckle their young, this ability derives from their evolutionary origin at the base of the clade, and thus statistically constitutes a single data point (not 5000). By contrast, convergently evolved traits are by definition independent evolutionary events, and each clade independently possessing a trait therefore represents an independent data point. Only a set of convergently evolved traits provides an adequate database for statistically valid tests of evolutionary hypotheses. This point is often ignored, even by biologists discussing music evolution (e.g. [23]). Fortunately, for many cases of convergent evolution, such as bipedalism or vocal learning, a trait has evolved independently enough times to provide a rich source of evidenceto test hypotheses concerning both evolution and mechanistic function. Thus, for example, we can test mechanistic hypotheses about the requirements of vocal learning by examining its neural correlates in the many species that have evolved this ability convergently (cf. [41]). Similarly, we can test functional hypotheses about why the capacity for vocal synchrony or antiphony is adaptive by examining the many bird, mammal, frog and insect species that have convergently evolved this ability [40]. While the conceptual distinction between homology and analogy is clear, recent discoveries in genetics and neuroscience suggest that in some cases a trait can be both homologous and analogous, depending on the level of explanation. For.