This paper introduces a special issue of Cognitive Science initiated on the 25th anniversary of the publication of Parallel Distributed Processing (PDP), a two-volume work that introduced the use of neural network models as vehicles for understanding cognition. The collection surveys the core commitments of the PDP framework, the key issues the framework has addressed, and the debates the framework has spawned, and presents viewpoints on the current status of these issues. The articles focus on both historical roots and contemporary (...) developments in learning, optimality theory, perception, memory, language, conceptual knowledge, cognitive control, and consciousness. Here we consider the approach more generally, reviewing the original motivations, the resulting framework, and the central tenets of the underlying theory. We then evaluate the impact of PDP both on the field at large and within specific subdomains of cognitive science and consider the current role of PDP models within the broader landscape of contemporary theoretical frameworks in cognitive science. Looking to the future, we consider the implications for cognitive science of the recent success of machine learning systems called “deep networks”—systems that build on key ideas presented in the PDP volumes. (shrink)

It is proposed that our memory is made up of individual, unconnected Records, to each of which is attached a Heading. Retrieval of a Record can only be accomplished by addressing the attached Heading, the contents of which cannot itself be retrieved. Each Heading is made up of a mixture of content in more or less literal form and context, the latter including specification of environment and of internal states (e.g. drug states and mood). This view of memory allows an (...) easy account of a number of natural memory phenomena as well as a variety of laboratory findings such as the differences between recall and recognition. The theory further proposes that Headed Records can neither be deleted nor modified. Data apparently against such a hypothesis can be accounted for in terms of the retrieval process. (shrink)

McRae, Brown and Elman argue against the view that events are structured as frequently‐occurring sequences of world stimuli. They underline the importance of temporal structure defining event types and advance a more complex temporal structure, which allows for some variance in the component elements.

This study presents a process model of very long‐term episodic memory. The process presented is a reconstructive process. The process involves application of three kinds of reconstructive strategies—component‐to‐context instantiation strategies, component‐instantiation strategies, and context‐to‐context instantiation strategies. The first is used to direct search to appropriate conceptual categories in memory. The other two are used to direct search within the chosen conceptual category. A fourth type of strategy, called executive search strategies, guide search for concepts related to the one targeted for (...) retrieval.A conceptual memory organization implied by human reconstructive memory is presented along with examples which motivate it. A basic retrieval algorithm is presented for traversing that stucture. Retrieval strategies arise from failures in that algorithm. The memory organization and retrieval processes are implemented in a computer program called CYRUS which stores events in the lives of former Secretaries of State Cyrus Vance and Edmund Muskie and answers questions posed in English concerning that information. Examples which motivate the process model are drawn from protocols of human memory search. Examples of CYRUS's behavior demonstrate the implemented process model. Conclusions are drawn concerning retrieval failures and the relationship of episodic and semantic memory. (shrink)

As new unanticipated items are added to a memory, it must be able to reorganize itself, integrating the new items into its structure. The reorganization process must maintain the memory's structure and also build up the knowledge retrieval strategies need to search that structure. This study will present an algorithm for knowledge‐based memory reorganization. Included in that algorithm are processes for directed generalization and generalization refinement. A fact retrieval system called CYRUS which uses the algorithm is also presented. Conclusions are (...) drawn about maintaining accessibility in a conceptual memory, organizing generalized knowledge with respect to specific knowledge, and expected retrieval failures due to change over time in the memory's organization. (shrink)

Philosophers and scientists have across the ages been amazed about the fact that development and learning often lead to not just a merely incremental and gradual change in the learner but sometimes to a result that is strikingly different from the learner’s original situation: amazed, but at times also worried. Both philosophical and cognitive neuroscientific insights suggest that experts appear to perform ‘different’ tasks compared to beginners who behave in a similar way. These philosophical and empirical perspectives give some insight (...) into what happens when a novice is transitioning to a stage of expertise. Generally, this implies that increased skill and expertise support better results and a more flexible performance, in part because these allow an agent to withdraw part of her attention and other cognitive resources from the tasks involved, enabling her to devote those resources to supporting, or completely different, tasks. As positive as these developments appear, these changes have also raised concerns. The main concern is whether gaining expertise is like raising a ‘cognitive monster’ which escapes the individual’s conscious control and influences her actions with undesirable automatisms. If so, we should ask ourselves whether experts are capable of taming this monster. The answer appears initially not to be positive. Indeed, it has been noted that since it is difficult for experts to withhold automatic responses this can lead to inflexibility or performance that is only optimal under certain conditions, because it is limited to a specific domain, often context-dependent, biased and inflexible. In what follows, I will consider this challenge of protecting expertise and harnessing this brittleness from philosophical and cognitive neuroscientific perspectives. Taking into account that action is in general determined by a multitude of factors, with learning and development affecting how these factors exert their influence, a philosophical question is how this complex and dynamic process can be explained and subsequently, how controlling it might be understood. First, though, I will present the issue at hand more closely: should we appreciate expertise if it is similar to growing a ‘cognitive monster’? Second, I will introduce the framework of a “Sculpted Space of Actions” (Keestra, 2014), which I developed in order to explain how the challenge of selecting an adequate option for action is facilitated by expertise as it helps to constrain the space of potential action options. Subsequently, the question is raised of how such a Sculpted Space of Actions influences an expert’s engagement with specific situations, like teaching students in a classroom setting. It will be argued that a well prepared expert—teacher or otherwise— is not only able to rely upon her routines but will at the same time be more perceptive and attentive to unforeseen events and actions, according to the recent cognitive neuroscientific theory of Predictive Processing. Integrating the theory of Predictive Processing with the Sculpted Space of Actions framework, I conclude that expertise contributes to adaptive and flexible responses to specific contexts, yet only if it is associated with explicit planning and articulation of situation specific intentions—the latter effectively putting the cognitive monster at rest for a while. (shrink)

The paper examines locative expressions and shows that an adequate account of their meaning must be based on two essential understandings: First, the simple spatial relation, often given as the meaning of the spatial prepositions, is only an “ideal” from which there are deviations in context; second, a level of “geometric conceptualization” mediates between “the world as it is” and language. Pragmatic “near principles” are formulated to explain some deviations from the ideal and several other apparent irregularities of prepositional use. (...) A set of “use types” of the ideal meaning is proposed to account for conventional aspects of locative meaning. The paper concludes with a discussion of the consequences of this description of locative expressions for artificial intelligence and linguistics. (shrink)

This research presents a computer model called EUREKA that begins with novice‐like strategies and knowledge organizations for solving physics word problems and acquires features of knowledge organizations and basic approaches that characterize experts in this domain. EUREKA learns a highly interrelated network of problem‐type schemas with associated solution methodologies. Initially, superficial features of the problem statement form the basis for both the problem‐type schemas and the discriminating features that organize them in the P‐MOP (Problem Memory Organization Packet) network. As EUREKA (...) solves more problems, the content of the schemas and the discriminating features change to reflect more fundamental physics principles. This changing network allows EUREKA to shift from a novicelike means‐ends strategy to a more expertlike “knowledge development” strategy in which the presence of abstract concepts are triggered by problem features. In this model, the strategy shift emerges as a natural consequence of the evolving expertlike organization of problem‐type schemos. EUREKA captures many of the descriptive models of novice expert differences, and also suggests a number of empirically testable assumptions regarding problem‐solving strategies and the representation of problem‐solving knowledge. (shrink)

Roger Schank's research in AI takes seriously the ideas that understanding natural language involves mapping its expressions into an internal representation scheme and that these internal representations have a syntax appropriate for computational operations. It therefore falls within the computational approach to the study of mind. This paper discusses certain aspects of Schank's approach in order to assess its potential adequacy as a (partial) model of cognition. This version of the Language of Thought hypothesis encounters some of the same difficulties (...) that arise for Fodor's account. (shrink)

cal practice: the enterprise of specifying information about the world for use in computer systems. Knowledge representation as a field also encompasses conceptual results that call practitioners’ attention to important truths about the world, mathematical results that allow practitioners to make these truths precise, and computational results that put these truths to work. This chapter surveys this practice and its results, as it applies to the interpretation of natural language utterances in implemented natural language processing systems. For a broader perspective (...) on such technical practice, in all its strengths and weaknesses, see (Agre 1997). Knowledge representation offers a powerful general tool for the science of language. Computational logic, a prototypical formalism for representing knowledge about the world, is also the model for the level of logical form that linguists use to characterize the grammar of meaning (Larson and Segal 1995). And researchers from (Schank and Abelson 1977) to (Shieber 1993) and (Bos to appear) have relied crucially on such representations, and the inference methods associated with them, in articulating accounts of semantic relations in language, such as synonymy, entailment, informativeness and contradiction. The new textbooks (Blackburn and Bos 2002a, Blackburn and Bos 2002b) provide an excellent grounding in this research, and demonstrate how deeply computational ideas from knowledge representation can inform pure linguistic study. In this short chapter, I must leave much of.. (shrink)

We describe an experimental approach toward implementing a commonsense "microtheory" for buying and selling. Our prototype system characterizes how intelligent agents hold items and money, how they buy and sell items, and the way money and items are transferred. The ontology of the system includes money (cash, check, credit card), agents (people, organizations), items (movable, real estate, service), barter, and the notions of transfer, loan, buying by installments, profit, and loss.

The principal critical objective of this dissertation is to examine three contemporary theories about the mind-body problem: dualism, anomalous monism, and functionalism. The dualism examined is closely linked to Cartesian dualism, while functionalism is a form of materialism. Anomalous monism is a kind of dual aspect view. All these theories have a long tradition with different formulations and exponents. I examine three contemporary exponents of these views: Saul Kripke, a dualist, Donald Davidson, an anomalous monist, and Daniel Dennett, a functionalist. (...) ;I argue that each of these theories is incorrect or has serious difficulties. The chief difficulty for dualism is in giving an account of mind-body interaction. The chief difficulty for anomalous monism is that internal tensions within the theory relating to Davidson's conception of physical events and scientific laws have the consequence that anomalous monism applies to the entities of physics and chemistry but not to events in the brain. The most serious difficulty for functionalism is in accounting for the qualities of subjective experience. ;I then discuss two criteria for identifying mental phenomena: incorrigibility and intentionality. I consider the problem of mind-body interaction as a problem for dualism. I argue that the mind-body problem is misconceived in that it is supposed that a univocal definition of 'physical' can be given. I state three different definitions of this term: one containing a scientific conception, one an objective conception, and one a personal conception. ;In my final chapter, I formulate a plural aspect view that accounts for mind-body interaction. On this view, mind is an aspect of body and body is an aspect of mind. This plural aspect view can be derived from a more general theory I call Viewpoint Pluralism. (shrink)

How can we explain the intentional nature of an expert’s actions, performed without immediate and conscious control, relying instead on automatic cognitive processes? How can we account for the differences and similarities with a novice’s performance of the same actions? Can a naturalist explanation of intentional expert action be in line with a philosophical concept of intentional action? Answering these and related questions in a positive sense, this dissertation develops a three-step argument. Part I considers different methods of explanations in (...) cognitive neuroscience (Bennett & Hacker’s philosophical, conceptual analysis; Marr’s three levels of explanation; Neural Correlates of Consciousness research; mechanistic explanation), defending ‘mechanistic explanation’ as a method that provides the necessary tools for integrating interdisciplinary insights into human action. Furthermore, a dynamic, explanatory mechanism allows the assessment of the impact of learning and development on expert action in a valuable way that other methods don’t. Part II continues by scrutinizing several cognitive neuroscientific theories of learning and development (neuroconstructivism; dual-processing theories; simulation theory; extended mind/cognition hypothesis), arguing for the complex interactions between different types of processing and different action representations involved in expert action performances. Moreover, according to our discussion of a particular ‘simulation theory’ these interactions can be influenced in several ways with the use of language, allowing an agent to configure a specific action representation for performance at a later stage. The results of Parts I and II are then applied in Part III to a parallel discussion of philosophical analyses of intentional action (discussing i.a. Frankfurt, Bratman, Pacherie and Ricoeur) and cognitive neuroscientific insights in it. Both approaches are found to converge in emphasizing the importance for an expert to develop stable patterns of actions that comply maximally not only with his intentions, but also with his motor expertise and with situational conditions. Consequently, his actions – automatic, or not – rely on this ‘sculpted space of actions’. (shrink)