Recent evidence appears to show a close connection between explanation and belief revision, specifically, the revision of graded beliefs. Insofar as this is also evidence of violations of Bayesian norms of reasoning, the question arises whether we are facing a new bias here, on a par with previously discovered biases in probabilistic reasoning. We consider an apparently successful attempt by Costello and Watts to explain away a number of known such biases in terms of sampling error, which makes those biases look entirely innocuous and compatible with the descriptive adequacy of Bayesian psychology in any but the most uninteresting way. Specifically, we query whether this attempt can be extended to neutralize the aforementioned evidence allegedly showing that explanatory considerations influence our reasoning in ways inconsistent with Bayesian prescriptions.

中文翻译：

---

可以解释解释性推理的证据吗？

最近的证据似乎显示了解释和信念修正之间的密切联系，特别是分级信念的修正。就这也是违反贝叶斯推理规范的证据而言，出现的问题是，我们是否面临新的偏见，与先前在概率推理中发现的偏见相提并论。我们认为 Costello 和 Watts 显然成功地尝试用抽样误差来解释许多已知的此类偏差，这使得这些偏差看起来完全无害，并且以最无趣的方式与贝叶斯心理学的描述充分性相容。具体来说，