Function, Teleology, Normativity
Eyes are for seeing, and wings are for flying. It seems intuitive to us to use the normal functions of biological traits to discuss their purposes. But what determines whether a particular level of visual capacity counts as normal for human eyes, and what level of flying capacity counts as normal for goose wings? The question about the normality of functions extends beyond biological performance. Our cognitive faculties, such as perception and memory, are also functions of our biological structures. Understanding biological function and normality is thus crucial for theories about knowledge, mental representation, and mental health.
Philosophers such as Millikan, Neander, and Garson have turned to evolutionary biology for answers, arguing that a trait's proper function is determined by what it was selected for in its evolutionary past. Their selected effect theories promise to ground normative standards for what biological traits should do or be like in the natural facts of selection history.
In my dissertation, I examine whether evolutionary theory can bear this normative weight. Drawing on conceptual analysis and recent developments in philosophy of biology, I argue that evolutionary proper functions cannot support many of the normative projects built upon them, for instance, proper functionalist epistemology, the view that a properly functioning truth-aimed cognitive faculty is necessary for knowledge.
Proper Functions as Etiological Functions of Predominant Traits.
Presented at APA Central Division Meeting, February 2025. [Paper]
​
I propose a revised definition of etiological proper function, arguing that proper functions should be attributed only to predominant traits — traits with a close-to-one frequency in the population. This restriction preserves the distinction between normal variations and dysfunctions, while better supporting the desired explanatory power and binary attribution of proper functions.
Evolutionary Function and Warranted Belief.
Presented at Philosophy of Science and Epistemology, HKUST, 2024.6.27-29.
​
I develop a critique of proper functionalism in epistemology. Specifically, I argue that evolutionary function cannot serve as a necessary condition for warrant, which is the epistemic good that makes a true belief knowledge. This is because cognitive improvements can be accidental, and those accidental cognitive improvements can later be selected and become a proper function. If knowledge is considered to reflect the agent’s ability or effort, a reliable belief produced by an accidental cognitive faculty should still qualify as knowledge, even if it is not a product of proper function.
Detaching Normativity from Evolutionary Functions.
Presented at Philosophy of Science Around the World, 2023.11.5-17.
​
Evolutionary functions are often associated with normativity: they describe how a trait should work. I argue that the normative aspect of evolutionary function is very minimal, if not non-existent at all. This is because evolutionary functions result from the contingent match between certain traits and the environment, either in the past evolutionary history or at present time. Therefore, it is misleading to use evolutionary functions as criteria or explanation for normative judgments, and there is no need to incorporate the normative element into an account of biological function.
Simulating Evolution, Understanding Possibilities
Simulating Evolution: How-Possibly Explanations and How They Explain.
Presented at ISHPSSB, 2025.7.20-25.
​
With the development and adoption of computer simulations, evolutionary biologists increasingly rely on how-possibly explanations provided by simulations to understand evolutionary processes. I propose an abductive way of understanding how-possibly explanations from simulations. One way a simulation offers an explanation is by providing a possible causal story, challenging the belief that certain events, causes, or mechanisms are impossible. Another way to understand explanations by simulations is to see it as an abductive inference: if the simulation’s input-output pairs match the observed ones, we may infer that the simulation’s mechanism resembles the real-world process, providing the best explanation for the match.
Possible Origin of Division of Labor in Multicellular Organisms. [Details]​​
Collaboration with Claudia Gadaleta, Jokubas Janusauskas, Sebastian Sander Oest, and Claudio Davini.
Presented at Philosophy and Biology Shop Talks, 2025.6.23-27.
Poster at PSA 29th Biennial Meeting, 2024.11.14-17. ​​
​The appearance of multicellular organisms with division of labor among cells marks a major evolutionary milestone. How does the transition from undifferentiated multicellularity to specialization of activities happen? We propose that it can begin with an aggregative phenomenon that occurs when cells share resources and are constrained by lifespan and spatial limits. We use an agent-based model to support this hypothesis: under certain conditions, some simulated cells specialize in one activity throughout their lifespan without genetic changes or selective processes.​
Levels of Selection
Units of Selection Question under the Statisticalist View.
Presented at ISHPSSB, 2023.7.9-15.
​
I argue that the units of selection question has its explanatory significance, even under the statisticalist view, which posits that models of evolutionary population dynamics explain population change by the appeal to the statistical structure of populations without adverting to the causes of survival and reproduction. The explanatory significance persists because the units of selection concerns the carriers of the trait. Although the carriers of the trait is a causal detail of the phenomena to be explained, it is crucial when applying evolutionary models to specific cases of population changes to offer a statistical explanation.