Contraction: On the Decision-Theoretical Origins of Minimal Change and Entrenchment

We present a decision-theoretically motivated notion of contraction which, we claim, encodes the principles of minimal change and entrenchment. Contraction is seen as an operation whose goal is to minimize loses of informational value. The operation is also compatible with the principle that in contracting A one should preserve the sentences better entrenched than A (when the belief set contains A). Even when the principle of minimal change and the latter motivation for entrenchment figure prominently among the basic intuitions in the works of, among others, Quine and Ullian (1978), Levi (1980, 1991), Harman (1988) and Gärdenfors (1988), formal accounts of belief change (AGM, KM – see Gärdenfors (1988); Katsuno and Mendelzon (1991)) have abandoned both principles (see Rott (2000)). We argue for the principles and we show how to construct a contraction operation, which obeys both. An axiom system is proposed. We also prove that the decision-theoretic notion of contraction can be completely characterized in terms of the given axioms. Proving this type of completeness result is a well-known open problem in the field, whose solution requires employing both decision-theoretical techniques and logical methods recently used in belief change.     

生成式人工智能对于医患关系的伦理重构

基于对生成式人工智能医学能力、风险以及第三者效应的综合考量,分析生成式人工智能介入医患关系可能构成的四种伦理转向：即“医−人工智能−患”共同决策模式（可能Ⅰ）、“患−人工智能”深度协作与医生保守立场（可能Ⅱ）、恢复传统“医−患”关系（可能Ⅲ）以及“医−人工智能”深度协作与患者保守立场（可能Ⅳ）。其中“医−人工智能−患”共同决策模式的交互本质可能进一步形成两种相互背离的发展取向：一是以患者为中心的医疗实践模式,二是计算机主义的医疗家长主义。

     

TRAINING FUNCTIONAL EXPRESSIVE LANGUAGE: THE ACQUISITION AND GENERALIZATION OF FIVE-ELEMENT SYNTACTIC RESPONSES1

A single-subject multiple-baseline design using within- and across-subject replication was employed to study the acquisition of expanded “agent-action-object” sentences and the spontaneous generation of this form in the natural environment. Three young language-delayed subjects were trained to describe various agent-action-object relationships with a five-element syntactical form. The language training strategy was a synthesis of the developmental-psycholinguistic and behavioral models. Dynamic interactions between familiar persons and objects were the stimulus events that the children mapped. A five-element syntactical form, previously absent from the children's language repertoires, was trained during individual sessions. Concurrent with baseline, training, and followup, each subject's language was monitored in another setting, the classroom during free play. After onset of training, the core elements of the complex syntactic form were spontaneously emitted by the child in its natural environment. The free-play data reflect individual differences in the emergence and frequency of each child's spontaneous use of the agent-action-object form. The acquisition and maintenance of the specific lexicon and syntax trained were tested by posttraining probes and responses to videotape presentations. These probes revealed generalization and maintenance of both the lexical and syntactical forms acquired in treatment. The main purpose of any language-training procedure should be to provide language that is functional for the child in the natural environment. This study, which documented the spontaneous usage of the core agent-action-object syntactical form in the natural environment, effectively trained a functional syntactical rule.     

A Conceptual Toolkit for Intensive Early Behavioral Intervention Teachers

Intensive early behavioral intervention (IEBI) for children with autism requires effective team work involving professional therapists, parents, teachers and others. It is important for all team members to employ a common conceptual and procedural vocabulary so that communication among them is clear, unambiguous, and effective. This article outlines the rationale of a small and practical vocabulary or conceptual toolkit that we have found useful in working with teachers and families in designing, implementing, and analyzing the details of behavioral programs to help autistic children learn to speak and understand language. Special attention is given to a heuristic device called the language matrix. Illustrative data describing the course of generative language acquisition by a child with autism taught according to the language matrix approach are included.     

Darwinizing debunking arguments

To ‘Darwinize’ a debunking argument is to broaden and thereby strengthen it in ways inspired by Charles Darwin. It is to employ Darwinian strategies that converge on the conclusion that certain putative phenomena – the reality of stance‐independent moral properties, for instance – are illusory or epistemically problematic for animals like us. The aim of this essay is to defend one such strategy and illustrate its power relative to most evolutionary debunking arguments currently on offer.