Preference change and conservatism: comparing the Bayesian and the AGM models of preference revision

Richard Bradley’s Bayesian model of preference kinematics is compared with Sven Ove Hansson’s AGM-style model of preference revision. Both seek to model the revision of preference orders as a consequence of retaining consistency when some preferences change. Both models are often interpreted normatively, as giving advice on how an agent should revise her preferences. I raise four criticisms of the Bayesian model: it is unrealistic; it neglects an important change mechanism; it disregards endogenous information relevant to preference change, in particular about similarity and incompleteness; and its representational framework, when expanded with similarity comparisons, may give misleading advice. These criticisms are based on a principle of conservatism, and on two proposals of similarity metrics for the Bayesian model. The performance of the Bayesian model, with and without the similarity metrics, is then tested in three different cases of preference change, and compared to the performance of the AGM model.     

In defence of a perspectival semantics for ‘know’

Relativism offers an ingenious way of accommodating most of our intuitions about ‘know’: the truth-value of sentences containing ‘know’ is a function of parameters determined by a context of use and a context of assessment. This sort of double-indexing provides a more adequate account of the linguistic data involving ‘know’ than does standard contextualism. However, relativism has come under recent attack: it supposedly cannot account for the factivity of ‘know’, and it entails, counterintuitively, that circumstances of evaluation have features that cannot be shifted by any intensional operator. I offer replies to these objections on behalf of the relativist. I then argue that a version of contextualism can account for the same data as relativism without relativizing sentence truth to contexts of assessment. This version of contextualism is thus preferable to relativism on methodological grounds.     

The impact of accountability on teachers’ assessments of student performance: a social cognitive analysis

Research on teachers’ judgments of student performance has demonstrated that educational assessments may be biased or may more correctly take the achievements of students into account depending on teachers’ motivations while making the judgment. Building on research on social judgment formation the present investigation examined whether the accountability of teachers has an influence on judgment formation. We predicted that unaccountable teachers would activate social categories and use them for the assessment, whereas accountable teachers’ attention would be directed to individual attributes of students. Using secondary school teachers as participants, three studies investigating teachers’ assessments, inferences and memory for students’ attributes supported these hypotheses. Thus, accountability appears to be a moderator of social information processing and judgment formation in the domain of educational assessments.     

Herrschen und Lieben als Grundmotive der philosophischen Weltanschauungen

Ohne Zusammenfassung     

Against Emotional Dogmatism

It may seem that when you have an emotional response to a perceived object or event that makes it seem to you that the perceived source of the emotion possesses some evaluative property, then you thereby have prima facie, immediate justification for believing that the object or event possesses the evaluative property. Call this view ‘dogmatism about emotional justification’. We defend a view of the structure of emotional awareness according to which the objects of emotional awareness are derived from other experiences such as bodily sensation, inner awareness, sensory perception, memory, and imagination. On this basis, we argue that dogmatism about emotional justification is an untenable position, regardless of whether the special feature of an immediate justifier that makes it an immediate justifier is its presentational phenomenology or its evidence insensitivity.