Perceptual learning and representational learning in humans and animals

Traditionally, perceptual learning in humans and classical conditioning in animals have been considered as two very different research areas, with separate problems, paradigms, and explanations. However, a number of themes common to these fields of research emerge when they are approached from the more general concept of representational learning. To demonstrate this, I present results of several learning experiments with human adults and infants, exploring how internal representations of complex unknown visual patterns might emerge in the brain. I provide evidence that this learning cannot be captured fully by any simple pairwise associative learning scheme, but rather by a probabilistic inference process called Bayesian model averaging, in which the brain is assumed to formulate the most likely chunking/grouping of its previous experience into independent representational units. Such a generative model attempts to represent the entire world of stimuli with optimal ability to generalize to likely scenes in the future. I review the evidence showing that a similar philosophy and generative scheme of representation has successfully described a wide range of experimental data in the domain of classical conditioning in animals. These convergent findings suggest that statistical theories of representational learning might help to link human perceptual learning and animal classical conditioning results into a coherent framework.     

从德布罗意波的波长和频率看波粒二象性

文章通过推导说明德布罗意波的波速、波长和频率三者之间,并不满足机械波和电磁波等经典波所遵循的关系,其波长、频率也不具有经典理论所描述的物理意义。其原因是德布罗意波并非经典波的简单再现,波粒二象性所指的“粒子性”和“波动性”都是具有独特含义的。     

A transfer of control test for contextual associations

A transfer of control experiment measured the associative properties of contextual stimuli from three standard classical conditioning paradigms. After baseline training on a Sidman avoidance schedule, dogs received aversive conditioning using excitatory, inhibitory, or truly random conditioning procedures in the presence of a manipulable background stimulus. As predicted by current theory (Rescorla & Wagner, 1972; Wagner & Rescorla, 1972), the contextual stimulus was excitatory after serving as the background during conditioning of a CS? and was neutral when it had been part of the background for conditioning of a CS+. The background to the truly random procedure was also neutral. This last result contrasts with Rescorla and Wagner’s theory.     

The effect of implicit theories on raters' inference in performance judgment: Consequences for the validity of student ratings of instruction

Performance judgment is a situation of incomplete information where raters' inference would play an important role. Consequently, the schematic nature of human cognition may introduce implicit personality theory bias in performance judgment. To demonstrate this, a causal model of performance rating judgment was framed from the theories of person perception and social cognition. The model yielded a good fit to the data obtained from a performance rating task where the availability of performance information was manipulated. The results supported the hypotheses that student raters' inferences are partly contaminated by their implicit theories of a good instructor. Student raters inferred traits and behaviors and provided ratings for corresponding items even when the instructor behavior was limited to a subset of performance data only. The findings imply that one aspect of invalidity in student ratings of instructors is the bias in human inference due to the implicit theories of effective instructional behavior.     

Causal Inferences in the Campbellian Validity System

The purpose of the present paper is to critically examine causal inferences and internal validity as defined by Campbell and co‐workers. Several arguments are given against their counterfactual effect definition, and this effect definition should be considered inadequate for causal research in general. Moreover, their defined independence between internal and construct validity is not meaningful. An alternative causal inference and its validity are proposed, where the causal effect is defined in factual terms, and where the causal inference includes constructs.     

西方两大派别学习理论发展过程的系统分析

西方学习理论可以分为学习的联结主义理论与学习的认知主义理论。学习的联结派认为，学习实质上是通过建立简单的联系而获得经验的过程，从桑代克的试误说起，到华生的经典性条件反射说，斯金纳的操作性条件反射说，再到班杜拉的观察学习说，联结派在本规范内经历了不断完善的过程。学习的认知派则认为，学习是经过复杂的加工活动建立认知结构而获得经验的过程，学习认知派理论从格式塔完形说、托尔曼的认知地图说开始，经历了从认知主义到建构主义的发展过程。由于两大派学习理论都无法解释学习的全野，因此也出现了企图融合两大派理论的折中主义学习理论。     

Blocking of conditioned suppression with 1 or 10 compound trials

It is generally agreed that the Kamin blocking effect provides a differential test of the model of classical conditioning proposed by Rescorla and Wagner (1972), on the one hand, and the models proposed by Mackintosh (1975) and Pearce and Hall (1980), on the other. Specifically, if the blocking effect occurs with 1 compound trial, Rescorla and Wagner are supported. Experiment 1 showed that prior training with one element of a simultaneous compound stimulus prevented the added element from becoming conditioned with 10 compound trials when compared with a nonpretrained (Kamin) control. A pseudoconditioning control, which received no compound training, showed substantial suppression only on the first trial. A contextual conditioning control, which received unsignaled shock prior to compound training, was suppressed, suggesting a failure of conditioning to contextual cues. Experiment 2 demonstrated 1-compound-trial blocking using a simultaneous presentation of the compound elements. The importance of this support for the Rescorla-Wagner model is discussed in the light of other supportive studies and those which support alternative models of classical conditioning.     

Applying Konorski’s model of classical conditioning to signal-centered behavior in the rat: Some functional similarities between hunger CRs and sign-tracking

Two experiments were conducted to investigate functional similarities between “hunger CRs” of Konorski’s (1967) model of appetitive classical conditioning and sign-tracking behavior in rats. Konorski’s model predicts that hunger CRs will be facilitated (1) when a nonrein-forced stimulus similar to the reinforced CS is introduced, and (2) when some CS presentations are unexpectedly nonreinforced. In Experiment 1, hungry rats acquired a leverpress response to a retractable lever that was paired with response-independent food. Following this training, a second lever was introduced whose presentation was not followed by food. The effect of the presence of this second lever was to facilitate responding to the original lever. In Experiment 2, single-lever autoshaping training was followed by a shift from 100% pairing of the lever with food to only 50% of the lever presentations being followed by food. The introduction of partial reinforcement produced an immediate and durable increase in leverpressing. The findings of both experiments are consistent with predictions from Konorski’s model of classical conditioning if sign-tracking is considered as a “hunger CR.”     

Evaluating the TD model of classical conditioning

The temporal-difference (TD) algorithm from reinforcement learning provides a simple method for incrementally learning predictions of upcoming events. Applied to classical conditioning, TD models suppose that animals learn a real-time prediction of the unconditioned stimulus (US) on the basis of all available conditioned stimuli (CSs). In the TD model, similar to other error-correction models, learning is driven by prediction errors-the difference between the change in US prediction and the actual US. With the TD model, however, learning occurs continuously from moment to moment and is not artificially constrained to occur in trials. Accordingly, a key feature of any TD model is the assumption about the representation of a CS on a moment-to-moment basis. Here, we evaluate the performance of the TD model with a heretofore unexplored range of classical conditioning tasks. To do so, we consider three stimulus representations that vary in their degree of temporal generalization and evaluate how the representation influences the performance of the TD model on these conditioning tasks.     

含有预设的推理与推理的有效性

推理的全部前提可以看作这一推理的预设。日常的具体推理往往是省略的推理，有些前提并未明确给出，在这种情况下，对预设作全面的分析就尤为重要。与此相关，考察经典逻辑和模态逻辑中的某些推理规则，讨论这些推理规则在日常推理中的有效性问题，是十分必要的。     

从逻辑学到心理学——归纳推理的心理学意义初探

逻辑学对归纳推理的理论阐述有诸多困难,如逻辑学对归纳推理的理论界定本身就包含着矛盾、逻辑学也没有在归纳推理和类比推理之间作出区分、逻辑学希望为归纳推理制定规则的努力也被证明是行不通的等。而心理学对归纳推理的研究则采取了和逻辑学的研究不同的策略,如心理学对归纳推理作了精确的定义、心理学描述归纳推理的内在心理规律而不是制定规则、心理学也在努力建立归纳推理的理论模型等。     

Discriminated avoidance vs classical conditioning: A two-phase model analysis

The acquisition of discriminated avoidance and classically conditioned leg flexions in dogs and of discriminated avoidance leg flexion in cats was examined via a two-phase model of conditioning. Classically conditioned Ss uniformly required two operators during Phase 2, with the limit of the CR operator being greater than that of the CR operator. Not all avoidance-trained Ss required two operators during Phase 2, but those who did also had a higher CR than CR operator limit. The CR operator limit was higher for avoidance than for classical training, while the change rate was greater for classical training. The avoidance behavior of cats paralleled that of the dogs. It was concluded that the data generally conformed to views of Hilgard and Marquis and, more recently, of Bolles, but were contrary to what would be expected from two-factor theory.     

后格赖斯语用学理论的发展

会话含意理论为解释应用中的语言开辟了新的途径,但同时存在着明显的不足之处, 各国学者在“古典”格赖斯理论之后,相继提出新的理论,其中包括斯帕伯和威尔逊的“关联理论”、 荷恩两原则、列文森三原则,以及徐盛桓在其1991年的推理模式的基础上,根据列文森三原则建立 的语用推理机制。这些理论和欧洲大陆的“宏观语用学”一起,推动了语用学理论的发展和完善。     

Resistance to discrimination and subsequent resistance to extinction as a function of the sequence of partial S+ reward in differential conditioning

The effects of transitions from nonrewarded (N) to rewarded (R) trials (N-R transitions) on discriminative behavior in differential conditioning and subsequent resistance to extinction were investigated in two experiments. In Experiment 1, groups given N-R transitions within S+ were more resistant to discrimination (ran fast in S?) and extinction than were groups given a partial reinforcement (PRF) schedule in S+ devoid of N-R transitions. Experiment 2 indicated that N-R transitions that occur when an N trial in S? is followed by an R trial in S+ are as effective in increasing resistance to discrimination, but not resistance to extinction, as are N-R transitions that occur within S+. The sequential effects obtained here were highly similar to those in conventional PRF and support the view that differential conditioning and PRF are highly interrelated phenomena. The results are discussed in terms of the extension of sequential theory to differential conditioning and the importance of internal reward-produced cues in discrimination learning.     

逻辑推论模型在研究生思想政治理论课教学中的应用

逻辑推论模型是分析问题和论证问题不可缺少的一种方法论手段,不仅能架起自然语言与逻辑语言之间的桥梁,澄清科学研究问题的语境,培养研究生做问题的意识和能力,更为重要的是,借助逻辑推论模型,研究生能够更加清晰地理解和掌握以习近平同志为核心的党中央方针政策理念的真理性、科学性和深远的方法论意义。运用逻辑分析方法,对2018版研究生思想政治理论课中"自然辩证法概论""中国特色社会主义理论与实践研究""中国马克思主义与当代"的相关知识点进行逻辑分析,力图展示逻辑推论模型在研究生思想政治理论课教学中的应用。     

马克思劳动二重性学说的理论来源

马克思特别重视劳动二重性学说,认为这一学说是对事实的全部理解的基础,而这一学说能够产生,在思想史上有两个作为前提的关键的环节,一个是劳动产生价值的思想,一个是对劳动进行一般和特殊的区分。这两个关键环节也就是马克思劳动二重性学说的理论来源。这两个关键的思想基础分别由古典经济学和黑格尔提出的,但是古典经济学和黑格尔由于哲学上的根本局限性,都没有从这些理论基础上进一步前进。马克思吸收了这两种思想作为基础,并且在哲学上产生了根本性的超越,从而产生劳动二重性学说,这一学说既使得经济学发生根本变革,也蕴含着一种新的哲学。     

A context-specific detrimental effect of UCS preexposure on place conditioning with morphine

The effect of morphine preexposure on place conditioning with morphine was investigated. In the first experiment, five injections of 10 mg/kg morphine were administered to rats prior to place conditioning or taste-aversion training with morphine. Although this number of preexposures retarded taste-aversion learning, there was no effect on place conditioning. In subsequent experiments we investigated the role of context blocking in UCS preexposure in place conditioning. In one experiment, preexposure to five morphine injections prior to place conditioning resulted in a reduced place preference, compared with preexposure and place conditioning in different contexts. However, the overall detrimental effect of morphine preexposure was questionable, because the rats that were preexposed were only marginally different from those that were not preexposed. In a final experiment we examined the effect of a context change from preexposure to place conditioning with 15 preexposures and demonstrated a detrimental effect of preexposure on place conditioning that was context specific. These results support a role of classical conditioning in place-preference conditioning with morphine.     

Conditioning history and inhibitory instrumental stimulus control: Independent-groups and within-subjects measures

The effects of excitatory conditioning history on establishing inhibitory stimulus control have been investigated in classical conditioning, but not in the free-operant paradigm. The present experiments address this question within the context of discriminated free-operant avoidance in which rats’ barpressing postponed shock. When a stimulus with only a history of signaling safety was combined, on a summation test, with a stimulus that maintained avoidance, avoidance rate was reduced, on average, by 60%. In comparison, after a stimulus acquired an excitatory free-operant avoidance history, nonreinforcement alone was not adequate to make it a predictable and effective inhibitor of avoidance on a summation test. These results, consistent with the classical conditioning literature, were produced by both between-group (Experiment 1) and within-subject (Experiment 2) comparisons. These findings are discussed in terms of (1) Konorski’s distinction between “primary” and “secondary” inhibitory stimuli, (2) the Rescorla-Wagner model, (3) the potential contribution of the “reinstatement of fear” to the outcome of summation tests, and (4) their implications for assaying the effectiveness of behavior-modification treatments of phobias.     

话语推理:从语用学到认知语用学

话语推理经历了从语用学到认知语用学的发展进程。Grice的会话含意理论及其合作原则首开了从语用学的角度研究话语推理的先河,在此基础上发展起来的Sperber和Wilson的关联原则探讨了作为主体的人对话语推理过程所起的重要作用。Thornburg和Panther提出的言语行为转喻理论则从语用学和认知的角度对转喻在言语行为中的语用功能做了较好的阐释,为话语推理的研究开创了新的思路。     

回溯推理独立存在的合理性问题探微

回溯推理是一种独立于演绎、归纳的推理类型,但有学者对此提出非难。回溯推理独立存在的合理性,在国内外学术界关于该推理的几种典型模式中,从哲学、逻辑学、心理学的不同视角提供了理论支持证据。科学发现、医疗诊断和刑事侦察等在实践上提供了经验支持证据。