Matthew Kotzen

Most philosophers accept some version of the requirement of total evidence (RTE), which tells us to always update on our complete evidence, which often includes ‘background information’ about how that evidence was collected. But different philosophers disagree about how to implement that requirement. In this article, I argue against one natural picture of how to implement the RTE in likelihood arguments, and I argue in favor of a different picture. I also apply my picture to the controversy over the so-called ‘Objection from Anthropic Bias’ to the fine-tuning argument, and argue that the Objection from Anthropic Bias fails

Nico Silins has proposed and defended a form of Liberalism about perception that, he thinks, is a good compromise between the Dogmatism of Jim Pryor and others, and the Conservatism of Roger White, Crispin Wright, and others. In particular, Silins argues that his theory can explain why having justification to believe the negation of skeptical hypotheses is a necessary condition for having justification to believe ordinary propositions, even though (contra the Conservative) the latter is not had in virtue of the former. I argue that Silins's explanation is unsuccessful, and hence that we should prefer either White/Wright-style Conservatism (which can explain this necessary condition) or Pryor-style Dogmatism (which denies that this is a necessary condition).

This essay addresses the question of when evidence for a stronger claim H1 also constitutes evidence for a weaker claim H2. Although the answer “Always” is tempting, it is false on a natural Bayesian conception of evidence. This essay first describes some prima facie counterexamples to this answer and surveys some weaker answers and rejects them. Next, it proposes an answer, which appeals to the “Dragging Condition.” After explaining and arguing for its use of the Dragging Condition, the essay argues that the Dragging Condition provides a general account of, and solution to, the counterexamples with which the essay began. The essay briefly discusses the relevance of the Dragging Condition to the recently much-discussed topic of “transmission failure” in epistemology, applies the Dragging Condition to the problem of “bootstrapping” in epistemology, and discusses three important objections to the view defended in the essay.

In this paper, I aim to explicate the distinction between ‘unconditional relevance’ and ‘conditional relevance’ as those terms and related concepts are applied in the context of admissibility determinations in modern trials. I take the U.S. Federal Rules of Evidence to be my model in analyzing these concepts, though on my view any reasonable approach to legal evidence will have to distinguish between these concepts and make appropriate provisions for their separate treatment. I begin by explaining how the Federal Rules define and apply the concepts of relevance and conditional relevance, and I present an influential argument due to Vaughn Ball that threatens to undermine the distinction between the two concepts. I then argue that Ball's argument fails and I diagnose that failure. However, building on some insights from a variety of evidence scholars, I argue that the approach to conditional relevance adopted by the Federal Rules is crucially flawed for reasons entirely independent of the ones raised by Ball's argument. I identify the main constraints that, on my view, any reasonable approach to conditional admissibility must obey, and I argue for a specific proposal that obeys those constraints. On my positive view, two pieces of evidence should be admitted under a Conditional Admissibility Principle only when each piece of evidence would survive ordinary admissibility scrutiny, conditional on the admission of the other one. I conclude by considering the question of whether it should also be necessary for the two pieces of evidence to survive admissibility scrutiny together, as an ‘evidential package’; I argue that, though the issue may arise infrequently in practice, there is good reason to impose this additional requirement.