davekinkead / thesis

A work-in-progess draft for my PhD thesis on computational methods in philosophy

Thesis Outline

One final attempt at this fella

My thesis is simple. I propose a novel approach to computational philosophy I call Simulation-as-Argument and demonstrate how it can make computational methodologies more philosophical.

My thesis is in three parts. In part one, I draw a history of how computers have been used in philosophy. I contrast digital philosophy, the use of computers to aid philosophical scholarship, from computational philosophy, the algorithmic generation of truth claims. I explore the epistemic nature of computer simulation and show how simulation expands the boundaries of know-ability, so long as we are aware of its limitations.

I then review the nature of philosophical scholarship and explore what sets philosophy apart from other disciplines. After exploring issues about subject matter and the open nature of philosophical questions, I argue that the defining features of philosophy are epistemic -- it is a priori and centered in argument.

Finally, I show how the current use of computers in philosophical work struggles to meet this philosophical criteria -- that the emerging sub-field of computational philosophy is closer to scientific scholarship, rather than philosophical.

In part two, I propose a novel approach to computational philosophy that I call Simulation-as-Argument. I explore the use of literate programming, the inversion of priority between code and comments first proposed by @kunth, and extend this concept with Javascript, a modern isomorphic programming language that can run on both web browser and computer server. I document how the choice of technology stack, programming language, and programming paradigm can transcend computer code from a set of instructions for a processor to execute to a form of argument in and of itself.

I then demonstrate Simulation-as-Argument in action with an example philosophical argument. I use Simulation-as-Argument (or maybe transcendental programming) to show how common inferences we make in public policy, such as inferring school performance

In part three, I argue that Simulation-as-Argument can make computational approaches to philosophy more philosophical. Not only does it extend the boundaries of know-ability like regular computer simulation can, but it does so while minimising many of the limitations faced by regular simulation.

I claim that Simulation-as-Argument represents an epistemic shift in the use of computer simulation by allowing an author to use computer algorithms as formal inference mechanisms rather than as a way to merely establish the truth claims of propositions. By embedding code within our philosophical scholarship, we can use computer simulation as a form of inference rather than testimony, a shift that makes our simulations philosophical.

Finally, I (might) briefly explore the rhetoric aspects of Simulation-as-Argument and show how the visceral experience of witnessing a counter-intuitive conclusion entail makes argument, when viewed as tools of rational persuasion, stronger.

Table of Contents

1. Introduction
2. Philosophy and Computing
3. Meta-philosophy
4. Simulation as Argument
5. School Performance
6. Transcendental Programming
7. Conclusion