Higher-Order Evidence

In science and in everyday life we collect evidence to acquire justified beliefs and to see whether what we believe is supported by the evidence. Recently, a special kind of evidence, higher-order evidence (HOE), has been getting more and more attention in epistemology. Ordinary, or first-order, evidence is assumed to inform directly about the world and HOE is taken to be evidence about ordinary evidence. Popular examples that have been presented as instances of HOE are evidence about doxastic disagreement or evidence about the unreliability of human reasoning when it comes to deductive inference. Despite the great importance of HOE for discussions of the justificatory status of beliefs there is no commonly shared characterization and taxonomy of HOE. Hitherto, there has been no agreement on the exact influence of HOE. The main aim of the project is to develop a theory of HOE and to investigate its epistemological role for rational reasoning. The project is divided into three sub-projects: Sub-Project 1: Characterization and Taxonomies of HOE. Sub-Project 2: HOE and Doxastic Disagreement Sub-Project 3: HOE and Deductive Inference. The working hypothesis of Sub-Project 1 is that HOE is best understood as evidence that describes or prescribes the epistemic state of an agent. The working hypotheses of Sub-Project 2 are: First, HOE about the evidential support relation of other agents does not necessarily play an important role for rational reasoning, because one can continue to rationally believe a proposition even if an epistemic peer believes its negation in the light of the same evidence. Second, if one possesses HOE that the other agent is more or less like one`s better-informed future-self, then one should adopt her degree of belief. The working hypothesis of Sub-Project 3 is that HOE cannot require us to lower our degree of belief in logical truths or and propositions that are implied by the propositions that we believe. Instead the role of HOE for rational reasoning in these cases requires the agent to suspend judgment on these propositions in the sense that the agent does not assign any degree of belief to these propositions. The project has the potential to revolutionize our conceptions of rational reasoning by emphasizing the importance of HOE. The project will develop a theory of HOE and will take the crucial steps to integrate it with our theories about rational reasoning with ordinary first-order evidence. This project discusses topics from traditional and formal epistemology and uses the methods of both. The methods of traditional epistemology relevant for the project include conceptual analyses and explications. The methods of formal epistemology relevant for the project include logic and probability theory. The researcher responsible for the project is Anna- Maria Asunta Eder.

Higher-order evidence (HOE) is one of the most relevant topics in contemporary Epistemology. It challenges an ordinary conception of the connection between evidence and rational belief. According to it, rational belief is mainly based on evidence about the world. This picture in accordance with popular characterizations of rationality but is challenged by HOE. Roughly speaking, first-order evidence is evidence about the world, HOE is evidence about, for instance, evidence itself, cognitive performance, or evidential relations. Many of our practices in politics, in court, and in science involve HOE. When politicians decide on policies they often do not seek observational evidence on some topic. Instead they are interested in HOE about what experts believe. In science, we often rely on HOE about our peers belief and about whether our own evidence have been obtained reliably and our reasoning is correct. In my research project, I explored different forms of HOE and their significance for debates in epistemology. In the following I summarize some if my results. It is commonly taken for granted that theorizing about epistemically ideal agents is helpful for investigating epistemic rationality. I show that this theorizing is problematic, which is due to the fact that we can acquire HOE that makes it impossible that our total evidence is the total evidence of ideal agents. This result is highly relevant for general debates on rationality. Epistemic rationality is commonly considered to be valuable. This is often explained by reference to positions that assume that rationality is conducive to epistemic ends such as truth. I explore the prospects and limitations of specifications of such positions from an evidentialist perspective and introduce my specification that is radical in nature. The specification that does not involve a commitment to aiming at the truth depends on how one deals with defeating HOE. Enjoying great popularity in decision theory, epistemology, and philosophy of science, Bayesianism is concerned with epistemically ideal rationality. It assumes a tight connection between evidential probability and ideally rational credence, and usually interprets evidential probability in terms of such credence. The evidential probability of a proposition on some body of evidence captures how strongly the body of evidence supports the proposition Timothy Williamson challenges Bayesianism. His argument is based on the fact that we posses HOE about our failing to recognize some logical truths. I show how to save Bayesianism.