Random Sampling, Offspring Distribution, and Genetic Drift

Yoichi Ishida

In population genetics, genetic drift as a phenomenon refers to non-directional or random changes in the frequency of types, such as alleles or genotypes, in a population. In its pure form, drift occurs in a population of finite size in the absence of any other evolutionary factors, such as selection, mutation, and migration. Both in biology and philosophy of biology, the common way of thinking about drift is in terms of random sampling. For example, each generation of a population of diploid organisms (like us) is thought to be a relatively small sample drawn randomly from an infinitely large pool of gametes produced by the previous generation of organisms. Because each generation may be an unrepresentative sample from the large pool of gametes, the allele frequency in a population may change randomly from generation to generation. In the paper I’m co-authoring with Alirio Rosales (UBC), we challenge this sampling-based thinking and suggest a viable alternative. This WIP talk will present our basic idea, and I will try to make it accessible to those without any detailed knowledge of population genetics or philosophy of biology.

Operationalizing Consciousness: Verbal Report and Task Performance

Trey Boone

The determination of a suitable operationalization for consciousness is of paramount importance to the project of unearthing the neural correlates of consciousness. For more than a century, we have been studying circumstances in which subjects seem to be able to differentially respond to stimuli in the performance of some task in spite of an inability to indicate those differences through subjective report. The empirical work along these lines (e.g. subliminal perception, blindsight, etc.) seem to support an operational definition of consciousnessbased on opposition between these two measures — task performance and subjective report — typically identifying consciousness with the latter. Lau (2008) represents a culmination of this approach consciousness, arguing not only that task performance is opposed to subjective report, but further that task performance should be seen as a potential confound in the study ofconsciousness. In other words, Lau maintains that, in order to identify the neural correlates of consciousness, we need to utilize tasks in which task performance may be held constant while there is nonetheless a difference in verbal report. I will argue that Lau and the approach he represents are misguided. Rather than thinking of subjective report and task performance as opposed measures, we should instead embrace an approach to consciousness in which these measures are seen as complementary.

The argument I intend to make is two-fold: one element is empirical, the other conceptual. The empirical argument draws on the work of Overgaard et al. (2008) to present evidence that the apparent tension between task performance and subjective report is, in many cases, an artifact of flawed experimental designs soliciting binary instead of graded subjective reports. I argue that these problematic experimental designs likely stem from a pretheoretical assumption that consciousness is a binary property of mental/neural states. I argue further that subjective report should not be seen as a measure of consciousness that is directly opposed to task performance. Rather, what is needed is an integrated approach where a combination of psychophysical task performance along with our understanding of the neurophysiology of perceptual systems can then shape subjective report tasks to be more sensitive to the types of properties subjects may be experiencing, but unable to report (in the standard paradigm).