Levelt, reviewing the AGL field from a psycholinguistic perspective, identifies various gaps and makes a number of concrete suggestions for improving several currently used experimental designs. He raises the question whether artificial (and natural) grammar learning is about detecting ‘rules’, as is commonly assumed, or rather the detection of a set of ‘constraints’. He cautions the community to not ignore ‘semantics’, and recommends to consider less artificial tasks, that may be needed for learning more complex rules by human (...) or nonhuman animals. (shrink)

The paper examines constraints and preferences employed by people in learning decision rules from preclassified examples. Results from four experiments with human subjects were analyzed and compared with artificial intelligence (AI) inductive learning programs. The results showed the people's rule inductions tended to emphasize category validity (probability of some property, given a category) more than cue validity (probability that an entity is a member of a category given that it has some property) to a greater extent than (...) did the AI programs. Although the relative proportions of different rule types (e.g., conjunctive vs. disjunctive) changed across experiments, a single process model provided a good account of the data from each study. These observations are used to argue for describing constraints in terms of processes embodied in models rather than in terms of products or outputs. Thus AI induction programs become candidate psychological process models and results from inductive learning experiments can suggest new algorithms. More generally, the results show that human inductive generalizations tend toward greater specificity than would be expected if conceptual simplicity were the key constraint on inductions. This bias toward specificity may be due to the fact that this criterion both maximizes inferences that may be drawn from category membership and protects rule induction systems from developing over‐generalizations. (shrink)

A theoretical framework for the reinforcement learning of aesthetic biases was recently proposed based on brain circuitries revealed by neuroimaging. A model grounded on that framework accounted for interesting features of human aesthetic biases. These features included individuality, cultural predispositions, stochastic dynamics of learning and aesthetic biases, and the peak-shift effect. However, despite the success in explaining these features, a potential weakness was the linearity of the value function used to predict reward. This linearity meant that the (...) class='Hi'>learning process employed a value function that assumed a linear relationship between reward and sensory stimuli. Linearity is common in reinforcement learning in neuroscience. However, linearity can be problematic because neural mechanisms and the dependence of reward on sensory stimuli were typically nonlinear. Here, we analyze the learning performance with models including optimal nonlinear value functions. We also compare updating the free parameters of the value functions with the delta rule, which neuroscience models use frequently, vs. updating with a new Phi rule that considers the structure of the nonlinearities. Our computer simulations showed that optimal nonlinear value functions resulted in improvements of learning errors when the reward models were nonlinear. Similarly, the new Phi rule led to improvements in these errors. These improvements were accompanied by the straightening of the trajectories of the vector of free parameters in its phase space. This straightening meant that the process became more efficient in learning the prediction of reward. Surprisingly, however, this improved efficiency had a complex relationship with the rate of learning. Finally, the stochasticity arising from the probabilistic sampling of sensory stimuli, rewards, and motivations helped the learning process narrow the range of free parameters to nearly optimal outcomes. Therefore, we suggest that value functions and update rules optimized for social and ecological constraints are ideal for learning aesthetic biases. (shrink)

A significant part of everyday learning occurs incidentally — a process typically described as implicit learning. A central issue in this domain and others, such as language acquisition, is the extent to which performance depends on the acquisition and deployment of abstract rules. Shanks and colleagues [22], [11] have suggested (1) that discrimination between grammatical and ungrammatical instances of a biconditional grammar requires the acquisition and use of abstract rules, and (2) that training conditions — in particular whether (...) instructions orient participants to identify the relevant rules or not — strongly influence the extent to which such rules will be learned. In this paper, we show (1) that a Simple Recurrent Network can in fact, under some conditions, learn a biconditional grammar, (2) that training conditions indeed influence learning in simple auto-associators networks and (3) that such networks can likewise learn about biconditional grammars, albeit to a lesser extent than human participants. These findings suggest that mastering biconditional grammars does not require the acquisition of abstract rules to the extent implied by Shanks and colleagues, and that performance on such material may in fact be based, at least in part, on simple associative learning mechanisms. (shrink)

A significant part of everyday learning occurs incidentally — a process typically described as implicit learning. A central issue in this domain and others, such as language acquisition, is the extent to which performance depends on the acquisition and deployment of abstract rules. Shanks and colleagues [22], [11] have suggested (1) that discrimination between grammatical and ungrammatical instances of a biconditional grammar requires the acquisition and use of abstract rules, and (2) that training conditions — in particular whether (...) instructions orient participants to identify the relevant rules or not — strongly influence the extent to which such rules will be learned. In this paper, we show (1) that a Simple Recurrent Network can in fact, under some conditions, learn a biconditional grammar, (2) that training conditions indeed influence learning in simple auto-associators networks and (3) that such networks can likewise learn about biconditional grammars, albeit to a lesser extent than human participants. These findings suggest that mastering biconditional grammars does not require the acquisition of abstract rules to the extent implied by Shanks and colleagues, and that performance on such material may in fact be based, at least in part, on simple associative learning mechanisms. (shrink)

The information explosion and digital modes of learning often combine to inform the quest for the best ways of transforming information in digital form for pedagogical purposes. This quest has become more urgent and pervasive with the ‘turn’ to online learning in the context of COVID-19. This can result in linear, asynchronous, transmission-based modes of teaching and learning which commodify, package and deliver knowledge for individual ‘customers’. The primary concerns in such models are often technical and economic (...) – technology as a cost-effective ‘solution’ to educational challenges. In this paper I argue for the importance of dialogic learning space in teaching and learning by means of Information and Communication Technologies, whether in the form of fully online learning, blended learning or face-to-face encounters using ICT affordances. Although the 20th Century theorists Mikhail Bakhtin (1895–1975) and Hans-Georg Gadamer (1900–2002) produced their seminal works before the advent of ICTs, they were both concerned with the quality and authenticity of human engagement with texts and with other persons and contexts. Besides a shared interest in dialogue as an ontological feature of human life and being, they both used spatiotemporal concepts for understanding and interpreting texts. The article draws on Gadamer’s notions of dialogue and horizon, and Bakhtin’s notions of dialogue and chronotope, to conceptualize dialogic possibilities for online education. Its purpose is to provide a framework, grounded in Bakhtin’s and Gadamer’s ideas, for a dialogic approach to online teaching and learning in higher education. (shrink)

Are morphological patterns learned in the form of rules? Some models deny this, attributing all morphology to analogical mechanisms. The dual mechanism model (Pinker, S., & Prince, A. (1998). On language and connectionism: analysis of a parallel distributed processing model of language acquisition. Cognition, 28, 73-193) posits that speakers do internalize rules, but that these rules are few and cover only regular processes; the remaining patterns are attributed to analogy. This article advocates a third approach, which uses multiple stochastic rules (...) and no analogy. We propose a model that employs inductive learning to discover multiple rules, and assigns them confidence scores based on their performance in the lexicon. Our model is supported over the two alternatives by new "wug test" data on English past tenses, which show that participant ratings of novel pasts depend on the phonological shape of the stem, both for irregulars and, surprisingly, also for regulars. The latter observation cannot be explained under the dual mechanism approach, which derives all regulars with a single rule. To evaluate the alternative hypothesis that all morphology is analogical, we implemented a purely analogical model, which evaluates novel pasts based solely on their similarity to existing verbs. Tested against experimental data, this analogical model also failed in key respects: it could not locate patterns that require abstract structural characterizations, and it favored implausible responses based on single, highly similar exemplars. We conclude that speakers extend morphological patterns based on abstract structural properties, of a kind appropriately described with rules. (shrink)

This addendum presents results that confound some commonly made claims about the sorts of environments in which the performance of TTB exceeds that of Franklin's rule, and vice versa.

Dialogue is a seminal concept within the work of the Brazilian adult education theorist, Paulo Freire, and the Russian literary critic and philosopher, Mikhail Bakhtin. While there are commonalities in their understanding of dialogue, they differ in their treatment of dialectic. This paper addresses commonalities and dissonances within a Bakhtin-Freire dialogue on the notions of dialogue and dialectic. It then teases out some of the implications for education theory and practice in relation to two South African contexts of learning (...) that facilitate the access to education of disadvantaged groups, one in higher education and the other in early childhood education. (shrink)

This paper argues that Lemmon-style proof systems (those that consist of only introduction and elimination inference rules) have several pedagogical benefits over Copi-style systems (those that make use of inference rules and equivalence rules). It is argued that Lemmon-style systems are easier to learn as they do not require memorizing as many rules, they do not require learning the subtle distinction between a rule of inference and a rule of replacement, and deriving material conditionals is more straightforward. (...) Finally, it is argued that the need for learning provisional assumptions in Lemmon-style rules is not a significant enough reason for choosing the Copi-style system. (shrink)

This paper argues that Lemmon-style proof systems (those that consist of only introduction and elimination inference rules) have several pedagogical benefits over Copi-style systems (those that make use of inference rules and equivalence rules). It is argued that Lemmon-style systems are easier to learn as they do not require memorizing as many rules, they do not require learning the subtle distinction between a rule of inference and a rule of replacement, and deriving material conditionals is more straightforward. (...) Finally, it is argued that the need for learning provisional assumptions in Lemmon-style rules is not a significant enough reason for choosing the Copi-style system. (shrink)

Does the human mind resemble the machines that can behave like it? Biologically inspired machine-learning systems approach “human-level” accuracy in an astounding variety of domains, and even predict human brain activity—raising the exciting possibility that such systems represent the world like we do. However, even seemingly intelligent machines fail in strange and “unhumanlike” ways, threatening their status as models of our minds. How can we know when human–machine behavioral differences reflect deep disparities in their underlying capacities, vs. when such (...) failures are only superficial or peripheral? This article draws on a foundational insight from cognitive science—the distinction between performance and competence—to encourage “species-fair” comparisons between humans and machines. The performance/competence distinction urges us to consider whether the failure of a system to behave as ideally hypothesized, or the failure of one creature to behave like another, arises not because the system lacks the relevant knowledge or internal capacities (“competence”), but instead because of superficial constraints on demonstrating that knowledge (“performance”). I argue that this distinction has been neglected by research comparing human and machine behavior, and that it should be essential to any such comparison. Focusing on the domain of image classification, I identify three factors contributing to the species-fairness of human–machine comparisons, extracted from recent work that equates such constraints. Species-fair comparisons level the playing field between natural and artificial intelligence, so that we can separate more superficial differences from those that may be deep and enduring. (shrink)

Given that artificial moral agents—such as autonomous vehicles, lethal autonomous weapons, and automated financial trading systems—are now part of the socio-ethical equation, we should morally evaluate their behavior. How should artificial moral agents make decisions? Is one moral theory better suited than others for machine ethics? After briefly overviewing the dominant ethical approaches for building morality into machines, this paper discusses a recent proposal, put forward by Don Howard and Ioan Muntean (2016, 2017), for an artificial moral agent based on (...) virtue theory. While the virtuous artificial moral agent has various strengths, this paper argues that a rule-based utilitarian approach (in contrast to a strict act-utilitarian approach) is superior because it can capture the most important features of the virtue-theoretic approach while realizing additional significant benefits. Specifically, a 2-level utilitarian artificial moral agent incorporating both established moral rules and a utility calculator is especially well-suited for machine ethics. (shrink)

This paper reports completeness results for dyadic deontic logics in the tradition of Hansson’s systems. There are two ways to understand the core notion of best antecedent-worlds, which underpins such systems. One is in terms of maximality, and the other in terms of optimality. Depending on the choice being made, one gets different evaluation rules for the deontic modalities, but also different versions of the so-called limit assumption. Four of them are disentangled, and compared. The main observation of this paper (...) is that, even in the partial order case, the contrast between maximality and optimality is not as significant as one could expect, because the logic remains the same whatever notion of best is used. This is established by showing that, given analogous properties for the betterness relation, the same system is sound and complete with respect to its intended modelling. The chief result of this paper concerns Åqvist’s system F supplemented with the principle of cautious monotony. It is established that, under the maximality rule, F + is sound and complete with respect to the class of models in which the betterness relation is required be reflexive and smooth . From this, a number of spin-off results are obtained. First and foremost, it is shown that a similar determination result holds for optimality; that is, under the optimality rule, F + is also sound and complete with respect to the class of models in which the betterness relation is reflexive and smooth . Other spin-off results concern classes of models in which further constraints are placed on the betterness relation, like totalness and transitivity. (shrink)

Inflectional morphology has been taken as a paradigmatic example of rule-governed grammatical knowledge (Pinker, 1999). The plausibility of this claim may be related to the fact that it is mainly based on studies of English, which has a very simple inflectional system. We examined the representation of inflectional morphology in Serbian, which encodes number, gender, and case for nouns. Linguists standardly characterize this system as a complex set of rules, with disagreements about their exact form. We present analyses of (...) a large corpus of nouns which showed that, as in English, Serbian inflectional morphology is quasiregular: It exhibits numerous partial regularities creating neighborhoods that vary in size and consistency. We then asked whether a simple connectionist network could encode this statistical information in a manner that also supported generalization. A network trained on 3,244 Serbian nouns learned to produce correctly inflected phonological forms from a specification of a word's lemma, gender, number, and case, and generalized to untrained cases. The model's performance was sensitive to variables that also influence human performance, including surface and lemma frequency. It was also influenced by inflectional neighborhood size, a novel measure of the consistency of meaning to form mapping. A word-naming experiment with native Serbian speakers showed that this measure also affects human performance. The results suggest that, as in English, generating correctly inflected forms involves satisfying a small number of simultaneous probabilistic constraints relating form and meaning. Thus, common computational mechanisms may govern the representation and use of inflectional information across typologically diverse languages. (shrink)

In a morally diverse society, moral agents inevitably run up against intractable disagreements. Civility functions as a valuable constraint on the sort of behaviors which moral agents might deploy in defense of their deeply held moral convictions and generally requires tolerance of other views and political liberalism, as does pluralism. However, most visions of civility are exceptionless: they require civil behavior regardless of how strong the disagreement is between two members of the same society. This seems an excellent idea (...) when those required to do the tolerating might otherwise smash us. However, the demands of civility are universal and fall upon everyone, including ourselves. They may seem to require us to tolerate the intolerable, leading us not into pluralism but rather into functional relativism, and also require the powerless to moderate their demands for redress. They also place moral agents in a very difficult position with respect to realizing our deeply held moral values. Isaiah Berlin’s pluralism, by contrast, allows us to violate tolerance when we come up against values which, put into practice, are incompatible with a form of life we can tolerate. Despite the many fronts on which civility and pluralism align, they are also pitted against each other. Only a qualified (not exceptionless) civility based in respect for persons can cohere with pluralism and thus resolve the double bind in which the moral agent seemed to be placed by exceptionless civility. I develop a rule for Accepted Exceptions that helps to explain how moral agents can be civil, value pluralism, demand redress, and maintain their own deeply held moral commitments. (shrink)

Digital transformation doesn't just raise ethical issues, it-in itself-is an ethical shift. Business leaders today are struggling to manage conflicting imperatives, those of the emerging digital world and those of the bureaucratic world of the past. The act of digital transformation requires a deep change in the moral outlook and ethical assumptions of a business. But how do we get there? Enterprise strategist and author Mark Schwartz shows how we need to learn to think differently about relationships with customers and (...) employees. That the ethics of digital transformation is a matter of cultivating and applying virtues rather than applying rules. Ethics is not just a matter of refraining from doing bad things. It's a matter of building the world we want, and it's the job of company executives. Featuring a chapter on bullshit, a handy chart of excuses for bad behavior, and Schwartz's typically paradoxical blend of deep insight and pasta jokes, this book guides business leaders as they struggle to adapt their bureaucratic framework of ethics to the emerging landscape of the digital world. By the end of the book, business leaders will rethink what it takes to be an ethical organization. (shrink)

This paper examines and finds wanting the arguments against van Fraassen’s voluntarism, the view that the only constraint of rationality is consistency. Foundationalists claim that if we have no grounds or rationale for a belief or rule, rationality demands that we suspend it. But that begs the question by assuming that there have to be grounds or a rationale. Instead of asking, why should we hold a basic belief or rule, the question has to be: why should (...) not we be committed as we are? Within a system we can sometimes find internal reasons. But, short of assuming foundationalism from the outset, when it comes to our evolving system as a whole there are no grounds for abandoning the commitments that we experience so strongly. Along the way the paper develops a systematic way of talking about terms that cause confusion because of variation in usage: foundationalism, relativism, basic beliefs and rules, voluntarism, etc. (shrink)

In lieu of an abstract, here is a brief excerpt of the content:Learning from Examples of Civic Responsibility:What Community-Based Art Centers Teach Us about Arts EducationJessica Hoffmann Davis (bio)Introduction/QuestionThroughout the United States, beyond school walls, there struggles and soars a sprawling field of community art centers dedicated to education.1 Most frequently clustered on either coast in bustling urban communities, these centers provide arts training that enriches or exceeds what is offered in schools. They serve artists who need space for (...) work or performance, students who crave instruction and direction, and the broader community that enjoys attendant cultural enrichment. At the core, they create safe havens for arts learning that has been marginalized elsewhere.Unfettered by the demands and constraints of school administrations, these self-designed centers have been of considerable interest to arts education researchers over the last twenty years.2 Largely conceived and staffed by artists, they demonstrate artful alternatives to mainstream education: creative visions of teaching, learning, and assessment. Moreover, the positive impact these centers have had on students considered at risk across a number of variables has caught the attention of researchers looking for better ways to serve our youth.3The location of these centers beyond school walls, like their operation beyond the hours of the school day, would naturally seem to cast them as positive opportunities for extending arts learning that goes on in school. But with school art curricula scarce or in danger of elimination, supplementary options can seem threatening to school arts specialists. Administrators have regarded the opportunities that outside artists and centers provide as sufficient substitutes for a school-based course of sequential instruction in the arts. While vital funding sources encourage and sustain collaboration, a tension persists between the autonomy of outsider arts centers and the integrity [End Page 82] of insider school arts programs. For students disenfranchised from schools, the "otherness" of centers has a certain cachet. For centers unable to afford their own space, school classrooms in after-school hours offer a home base for their offerings and easy access for students. The back and forth across school walls raises opportunities and concerns that have resulted in a range of creative solutions.4 But my interest here is not in the cross-wall meeting of services, but rather in the differences in arts learning on either side of school walls. In the pages that follow, I pose the question, What can schools learn from the arts education provided in the community?In addressing this question, I first mark the differences between arts learning in schools and in the community as a need for justification versus an opportunity for demonstration. I then describe the research from which my thinking and discussion draw and offer a brief overview of the history of community art centers as inspired by and illustrative of civic responsibility. With an eye to what educators and advocates can learn from these community-based efforts, I focus next on three common objections to prioritizing the arts in our schools: value, measurement, and autonomy. Each objection is introduced from a school-based perspective and then reconsidered in light of community-based attitudes and activities.5 In conclusion, I delineate the insights derived from a consideration of community arts education as stimulus for the redirection of educational objectives and advocacy rationales for arts learning in schools.Justification vs. DemonstrationIn its struggle for recognition and place in school settings, arts education has sought justification in terms of its potential to serve the broader educational scene. We ask of classes that provide mounting skills in visual arts, music, or theater, "Yes, but what else does arts learning do? Can it improve student skills in reading, mathematics, or any other subjects that are more traditionally valued?" Such demands may seem unfair. When has math ever been defended for its ability to improve arts performance? Beyond a sense of injustice, claims for arts learning's causal impact on measurable extraneous outcomes (such as improvements in test scores in nonarts academic subjects) are at best most difficult to prove.6In community centers where art is the chosen heart of the matter, students come specifically to study the arts and to benefit from their enhancement of more generalized experience... (shrink)

Business ethics can be regarded as a field dealing with corporate _self-regulation_ as it relates to the treatment of stakeholders. However, a concern for corporate stakeholders need not take a corporate-centric perspective, as shown by recent efforts (especially Singer in Bus Ethics Q 25(1):65–92, 2015) to situate corporate conduct within Rawls’ political theory. Although Rawls was largely mute on the subject himself, his theory has implications for business ethics and corporate governance more specifically. Given an understanding of a “Rawlsian society” (...) as a whole—where corporations as associations are a part—this paper addresses how a Rawlsian perspective would safeguard against corporate harms in society. We argue that a Rawlsian society would primarily regulate corporate conduct through exogenous constraints in the form of legislation. To the extent that business ethics is concerned with endogenous constraints in the form of corporate-centric self-regulation regarding _stakeholders_, to adopt a Rawlsian perspective is to assume instead a society-centric perspective and to impose exogenous constraints on corporate conduct in the form of legislation for the benefit of _citizens_. In the context of Rawls’ political liberalism, normative concerns in business are accounted for through legislation and the system of background justice. In a clear departure from Singer (Bus Ethics Q 25(1):65–92, 2015, Bus Ethics J Rev 6(3):11–17, 2018a), we further develop our argument to propose that Rawls' theory can be interpreted as providing a rule for corporate governance. The rule—which is imposed exogenously for the good of society—states: After choosing the corporate constraint mechanism (exogenous vs. endogenous) that best promotes the Liberty Principle, choose the corporate control regime (shareholder vs. stakeholder) that maximizes economic efficiency. (shrink)

Over the past decade, sequence learning has gradually become a central paradigm through which to study implicit learning. In this chapter, we start by briefly summarizing the results obtained with different variants of the sequence learning paradigm. We distinguish three subparadigms in terms of whether the stimulus material is generated either by following a fixed and repeating sequence (e.g., Nissen & Bullemer, 1987), by relying on a complex set of rules from which one can produce several alternative (...) deterministic sequences (e.g., Lewicki, Hill & Bizot, 1988; Stadler, 1989), or by following the output of a probabilistic set of rules such as instantiated by noisy finite-state grammars (Cleeremans & McClelland, 1991; Jiménez, Mendéz & Cleeremans, 1996). Next, we focus on the processes involved in sequence representation and acquisition. We suggest that the sensitivity to the sequential structure observed in the probabilistic subparadigm can only be a result of the acquisition of a representation of the statistical constraints of the material, and that this sensitivity emerges through the operation of mechanisms that are well instantiated by connectionist models such as the Simple Recurrent Network (Elman, 1990; Cleeremans, 1993b). We present new simulation work meant to explore to what extent the model can also account for specific data obtained in a paradigmatic instance of deterministic, rule-based sequence learning task: Lewicki et al. (1988)'s situation. Finally, we report on the results of an experiment that compares learning on otherwise similar deterministic and probabilistic structures, and we show that learning of both types of structures is equivalent only under conditions that maximally hinder explicit acquisition. Taken together, these simulation and experimental data lend support to the claim that implicit learning in all three sequence learning subparadigms can amount to a form of statistical sequence learning. They also suggest that distinguishing among several theories of sequence representation and acquisition may require us to analize the data in great detail. Hopefully, however, some truth can be found in such details.. (shrink)

This paper examines the relationship between learning and practice, rule and invention, in Japanese art. Drawing on Chinese precedent, learning through the close observation of conventional models for technical mastery or stylistic construction, underpinned training in almost all of the arts and crafts in Japan. The practice of building individually inventive projects was usually developed only after the successful completion of long apprenticeships in studio settings. The pictorial engagements of Edo, today's Tokyo, form the principal focus for (...) this examination of how individually forged learning pathways developed within the formal constraints of Tokugawa period art-training arrangements. This paper... (shrink)

Subjects exposed to strings of letters generated by a finite state grammar can later classify grammatical and nongrammatical test strings, even though they cannot adequately say what the rules of the grammar are (e.g., Reber, 1989). The MINERVA 2 (Hintzman, 1986) and Medin and Schaffer (1978) memory‐array models and a number of connectionist outoassociator models are tested against experimental data by deriving mainly parameter‐free predictions from the models of the rank order of classification difficulty of test strings. The importance of (...) different assumptions regarding the coding of features (How should the absence of a feature be coded? Should single letters or digrams be coded?), the learning rule used (Hebb rule vs. delta rule), and the connectivity (Should features be predicted only by previous features in the string, or by all features simultaneously?) is investigated by determining the performance of the models with and without each assumption. Only one class of connectionist model (the simultaneous delta rule) passes all the tests. It is shown that this class of model can be regarded by abstracting a set of representative but incomplete rules of the grammar. (shrink)

Theories of language production propose that utterances are constructed by a mechanism that separates linguistic content from linguistic structure, Linguistic content is retrieved from the mental lexicon, and is then inserted into slots in linguistic structures or frames. Support for this kind of model at the phonological level comes from patterns of phonological speech errors. W present an alternative account of these patterns using a connectionist or parallel distributed proceesing (PDP) model that learns to produce sequences of phonological features. The (...) model's errors exhibit some of the properties of human speech errors, specifically, properties that have been attributed to the action of phonological rules, frames, or other structural generalizations. (shrink)

Two key research issues in the field of causal learning are how people acquire causal knowledge when observing data that are presented sequentially, and the level of abstraction at which learning takes place. Does sequential causal learning solely involve the acquisition of specific cause-effect links, or do learners also acquire knowledge about abstract causal constraints? Recent empirical studies have revealed that experience with one set of causal cues can dramatically alter subsequent learning and performance with entirely (...) different cues, suggesting that learning involves abstract transfer, and such transfer effects involve sequential presentation of distinct sets of causal cues. It has been demonstrated that pre-training can modulate classic causal learning phenomena such as forward and backward blocking. To account for these effects, we propose a Bayesian theory of sequential causal learning. The theory assumes that humans are able to consider and use several alternative causal generative models, each instantiating a different causal integration rule. Model selection is used to decide which integration rule to use in a given learning environment in order to infer causal knowledge from sequential data. Detailed computer simulations demonstrate that humans rely on the abstract characteristics of outcome variables to select a causal integration rule, which in turn alters causal learning in a variety of blocking and overshadowing paradigms. When the nature of the outcome variable is ambiguous, humans select the model that yields the best fit with the recent environment, and then apply it to subsequent learning tasks. Based on sequential patterns of cue-outcome co-occurrence, the theory can account for a range of phenomena in sequential causal learning, including various blocking effects, primacy effects in some experimental conditions, and apparently abstract transfer of causal knowledge. (shrink)

The paper suggests that there are two different ways in which a legal system restricts an individual’s rights. It can either grant a power that revokes the legal protection of the right or it can acknowledge the infringement of a legal right and yet justify such an infringement by means of a criminal law justification. The distinction proposed by the paper has both expressive and practical implications and is useful in solving dilemmas arising in emergencies when constitutional constraints make it (...) impossible to grant the power to revoke legal protection of a basic right. In some of these situations a criminal law justification might support infringement of such a right. This claim is demonstrated by analyzing the ruling of the German Constitutional Court concerning the shooting down of a hijacked airplane in circumstances similar to those of September 11. (shrink)

Connected devices, such as smartphones and tablets, are exposed to a large variety of attacks. Their protection is often challenged by their resource constraints in terms of CPU, memory and energy. Security chains, composed of security functions such as firewalls, intrusion detection systems and data leakage prevention mechanisms, offer new perspectives to protect these devices using software-defined networking and network function virtualization. However, the complexity and dynamics of these chains require new automation techniques to orchestrate them. This chapter describes an (...) automated orchestration methodology for security chains in order to secure connected devices and their applications. This methodology exploits process learning to establish behavioral models and infer security constraints represented as logical predicates. It then generates and merges a set of chains of security functions on the basis of these predicates. These chains are finally compiled into low-level configuration rules and deployed into the network, optimizing for the underlying topology. The benefits and limits of such a methodology combining machine learning and verification techniques are evaluated by a set of experimental results. (shrink)

The qualitative nature of co-deployed or “associated” metaphorical source domains in discourse has been extensively researched, often in terms of whether they share common conceptual roots. There are however few empirical studies on the strength of these associations and their implications. In mental healthcare activities like psychological interviews and counseling, for example, strongly associated sources may suggest unique “conceptual assemblages” that highlight underexplored (dis)similarities between clients or client groups, beyond the typical focus on isolated sources and frequencies. Our case study (...) compares source domain associations produced by two interviewee groups – one meeting the diagnostic threshold for acute stress disorder (ASD), the other not – relating their first-hand experiences of traumatic events during the 2019 Hong Kong social unrest. The machine learning method of association rule mining is used to extract and quantify the top association rules in both groups. Results show i) a shared rule suggesting a highly schematic construal of trauma that nevertheless varies in instantiating details, and ii) several distinct rules that corroborates and lends further insight into the “agentive vs. non-agentive” characteristics exhibited by ASD vs. non-ASD individuals. Implications and future directions, including potential extensions to other discourse contexts, are discussed. (shrink)

Using data of China’s listed companies from 2000 to 2016, we employ a staggered difference-in-difference approach to identify the causal effects of CEO turnover on corporate innovation. First, we find that listed companies with CEO turnover experienced an average increase of 9.5% in the quantity of innovation and 8.9% in innovation quality after the change. The dynamic effect test supports the parallel trend condition, and the placebo test rules out the nonrandom selection issue. Second, the heterogeneity tests show that CEO (...) turnover plays a more prominent role in promoting innovation for listed firms with CEO duality, high financial constraints, and in high-tech industries. Third, CEO turnover affects corporate innovation by driving top management team reorganization and promoting R&D input. This paper has important implications for the understanding of the role of CEO turnover in companies’ innovation, as well as for strategy formulation and implementation. (shrink)

밀의 규범철학이 행위 공리주의인지 규칙 공리주의인지의 논쟁이 있다. 논문은 최근 부각된 규칙 공리주의적 해석이 근거한 문구들을 검토하여 그 해석이 부적절함을 보이고자 한다. 그리고 행위 공리주의적 측면, 즉 개별 행위의 합산된 결과만이 그 행위의 평가 기준이라는 관점도 밀에서 행위 정당화의 한 측면이라고 주장한다. 논문은 밀의 공리주의는 다양한 평가 대상과 차원들을 가진다는 해석을 옹호한다. 예를 들어 행위는 그 결과의 측면에서 평가될 수도 있고 행위를 낳은 성격의 측면에서 정당화될 수도 있다. 후자의 정당화는 성격이 행위를 통해 일반적으로 낳는 결과에 의존하는데, 일부 경우 일반적 효용성이 (...) 있는 성격이 최적이 아닌 결과를 가지는 행위를 산출하기도 하지만 실수가 가장 적은 성격에 의한 행위라는 측면에서 행위는 정당화될 수 있다. 논문은 이러한 다층적 해석의 근거를 밀의 주요 저작을 통해 옹호한다. (shrink)

Traditionally, learning organizations face certain constraints related to both exogenous and endogenous factors. This paper models three well-established constraints that employees face while being part of their organizations. One is an explicit constraint on their natural behavior, and two implicit constraints on their endeavor to acquire new knowledge and perform new actions. The implicit constraints, which are elaborated, are related to their relative performance in acquiring new knowledge and by their consecutive actions based on the new knowledge gained. (...) Therefore, this paper attempts to underline such limitations which the agents face under organizational culture and suggest possible strategic initiatives that would effectively counteract such binding limitations to stimulate positive performances from their end. (shrink)

In On Virtue Ethics I offered a criterion for a character trait's being a virtue according to which a virtuous character trait must conduce to, or at least not be inimical to, four ends, one of which is the continuance of the human species. I argue here that this does not commit me to homosexuality's being a vice, since homosexuality is not a character trait and hence not up for assessment as a virtue or a vice. Vegetarianism is not up (...) for such assessment either, for the same reason, but, as a practice, may well be required by the virtue of compassion, and sacrificing one's life for an animal or alien may be required by courage. The clause about the continuance of the human species in my criterion does not specify a foundational value, because, following McDowell, I reject foundationalism. (shrink)

Proponents of the dominant paradigm in evolutionary psychology argue that a viable evolutionary cognitive psychology requires that specific cognitive capacities be heritable and “quasi-independent” from other heritable traits, and that these requirements are best satisfied by innate cognitive modules. We argue here that neither of these are required in order to describe and explain how evolution shaped the mind.

We evaluate the asymptotic performance of boundedly-rational strategies in multi-armed bandit problems, where performance is measured in terms of the tendency (in the limit) to play optimal actions in either (i) isolation or (ii) networks of other learners. We show that, for many strategies commonly employed in economics, psychology, and machine learning, performance in isolation and performance in networks are essentially unrelated. Our results suggest that the appropriateness of various, common boundedly-rational strategies depends crucially upon the social context (if (...) any) in which such strategies are to be employed. (shrink)