Representations of task context play a crucial role in shaping human behavior. While the nature of these representations remains poorly understood, existing theories share a number of basic assumptions. One of these is that task representations are discrete, independent, and non-overlapping. We present here an alternative view, according to which task representations are instead viewed as graded, distributed patterns occupying a shared, continuous representational space. In recent work, we have implemented this view in a computational model of routine sequential action. In the present article, we focus specifically on this model's implications for understanding task representation, considering the implications of the account for two influential concepts: (1) cognitive underspecification, the idea that task representations may be imprecise or vague, especially in contexts where errors occur, and (2) information-sharing, the idea that closely related operations rely on common sets of internal representations.

It is generally recognized that today’s frontier of HCI research lies beyond the traditional desktop computers whose GUI interfaces were built on the foundation of display—pointing device—full keyboard. Many interface challenges arise without such a physical UI foundation. Text writing — ranging from entering URLs and search queries, filling forms, typing commands, to taking notes and writing emails and chat messages — is one of the hard problems awaiting solutions in off-desktop computing. This paper summarizes and synthesizes a research program on this topic at the IBM Almaden Research Center. It analyzes various dimensions that constitute a good text input interface; briefly reviews related literature; discusses the evaluation methodology issues of text input; presents the major ideas and results of two systems, ATOMIK and SHARK; and points out current and future directions in the area from our current vantage point.

People produce long sequences such as speech and music with incremental planning: mental preparation of a subset of sequence events. The authors model in music performance the sequence events that can be retrieved and prepared during production. Events are encoded in terms of their serial order and timing relative to other events in a planning increment, a contextually determined distribution of event activations. Planning is facilitated by events ’ metrical similarity and serial/temporal proximity and by developmental changes in short-term memory. The model’s predictions of larger planning increments as production rate decreases and as producers ’ age–experience increases are confirmed in serial-ordering errors produced by adults and children. Incremental planning is considered as a general retrieval constraint in serially ordered behaviors. When people produce long, complex sequences such as speech and music, they must plan what event to produce next (the serialorder problem) and when to produce it (the timing problem). Bernstein (1967) and Lashley (1951) both pointed to music as a quintessential example of serial-ordering abilities because of its complexity, length, and temporal properties. Although musical

This paper reviews recent approaches to human action planning and the cognitive representation of intentional actions. Evidence suggests that action planning takes place in terms of anticipated features of the intended goal, that is, in terms of action effects. These effects are acquired from early infancy on by registering contingencies between movements and perceptual movement outcomes. Co-occurrence of movements and effects leads to the creation of bidirectional associations between the underlying internal codes, thus establishing distributed perception-action networks subserving both perceiving external events and intentionally producing them. Action plans determine only the general, goal-relevant features of intended actions, while the fine-tuning is left to on-line sensory-motor processing. Action plans emerge from competition for action control between several factors: overlearned habits, perceptual events, and emotional influences, among others. Accordingly, action control represents a balance between personal intentions and wishes on the one hand and environmental affordances and demands on the other. KEYWORDS: action planning, intentional action, goal, perception and action, feedback, action effects, action control, will, priming, imitation, mirror neurons, emotion and action DOMAINS: behavioral psychology, cognition, development, learning and memory, motor processes, sensation and perception, neuroscience, behavior PLANNING AN ACTION Humans perform actions to reach goals, that is, to create or modify some event or state of affairs according to their intentions --- otherwise we would talk of movement but not action. Logically, then, intentional, goal-directed action presupposes some sort of (conscious or unconscious) anticipation of the intended goal event,...

In this paper the traditional paradigm for learning and training of operators in complex systems is discussed and criticised. There is a strong influence (the doctrine of 'mental logic') coming from research carried out in artificial intelligence (AI). The most well known arguments against the AI-approach are presented and discussed in relation to expertise, intuition and implicit knowledge. The importance of faults and errors are discussed in the context of a new metaphor for cognitive structures to describe expertise. Keywords: fault, error, learning, training, cognitive structure, expertise, intuition 1 Introduction "I learned more from my defeats than from my victories" (Napoleon, ca. 1800) Why is this statement sometimes (or always) true? To answer this question we need a new understanding of human errors, inefficient behaviour, and expertise. In this paper we will discuss the importance of learning from unsuccessful behaviour. What percentage of unanticipated events (e.g., a...

We extend the in-situ Hebbian-linked SOMs network by Miikkulainen to come up with two unsupervised neural networks that learn the mapping between the individual modes of a multimodal dataset. The first network, the single-pass Hebbian linked SOMs network, extends the in-situ Hebbian-linked SOMs network by enabling the Hebbian link weights to be computed through one-shot learning. The second network, a modified counterpropagation network, extends the unsupervised learning of crossmodal mappings by making it possible for only one self-organising map to implement the crossmodal mapping. The two proposed networks each have a smaller computation time and achieve lower crossmodal mean squared errors than the in-situ Hebbian-linked SOMs network when assessed on two bimodal datasets, an audio-acoustic speech utterance dataset and a phonological-semantics child utterance dataset. Of the three network architectures, the modified counterpropagation network achieves the highest percentage of correct classifications comparable to that of the LVQ-2 algorithm by Kohonen and the neural network for category learning by de Sa and Ballard in classification tasks using the audio-acoustic speech utterance dataset.

The Dvorak keyboard has been claimed to be greatly superior to the standard typewriter keyboard. However, none of the earlier research on the relative merits of the two keyboards provided unconfounded measures, ones permitting attribution of the results solely to the differences between the keyboards. The present research supplied, for the first time, direct measures of speed on the two keyboards by the same persons. Eight experienced standardkeyboard typists, ranging in skill from the median speed of terminal high school trainees to beyond the 97th percentile speed of experienced employees (45-81 words per minute), typed high-frequency digraphs on both keyboards, resulting in a 4.0% superiority for the Dvorak keyboard. The relationship of present digraph findings to performance of realistic tasks is discussed, and research on whether differences in keyboard efficiency vary with the skill level of operators is recommended. Under procedures that produced results entirely attributable to the keyboard designs, free of confounding factors, the Dvorak typewriter keyboard was found to produce speeds of keying 4.0 percent faster than those of the Standard (Qwerty) keyboard. However, employers were reported to be unwilling to bear the costs of the several weeks required to retrain employees on the novel keyboard. KEYWORDS: keyboards, typewriters, Dvorak, QWERTY NOTES "The Standard and Dvorak Keyboards Revisited: Direct Measures of Speed" by Leonard J. West is an unusual paper to appear in the Santa Fe Institute Working Paper Series. The paper was written for researchers in vocational education rather than economists per se. What makes this paper important to the SFI Economics community is its focus on what has become one of the canonical examples of path dependence in economics...

this paper we will summarize our research on written word production and discuss it in the context of other empirical results and models of written language production. Due to methodological reasons, the studies reported focus on typing instead of hand writing or oral spelling. As we are not interested specifically in motor processes but in cognitive and linguistic processes in writing, we chose typing as the most advantageous mode. Certainly, typing is a (discontinuous) motor task and the time course of typing is to a certain amount determined by motor conditions. But the methodology presented below allows for a differentiation between various factors determining the time course. Hand writing, on the other hand, is a continuous and much more complex motor task and hence it is much more difficult to separate linguistic and non-linguistic factors. Finally, oral spelling is on the one hand not determined by graphomotor processes (instead, articulomotorics) but on the other hand it is very slow and certainly not the main mode of writing

Two experiments tested 1 aspect of L. Hasher and R. T. Zacks's (1988) reduced inhibition hypothesis, namely, that old age impairs the ability to suppress information in working memory that is no longer relevant. In Experiment 1, young and older adults were asked to recall lists of letters in the correct order. Half of the lists contained repeated items while half were control lists. Recall of nonadjacent repeated items was worse than that of control items. This Ranschburg effect was larger (i.e., greater response suppression) in older than in young adults. In Experiment 2, young and older adults were required either to recall the list or to report if there was a repeated item. Repetition detection was high and similar in the 2 age groups. When age differences in overall performance were taken into account, there was evidence of increased repetition inhibition with age in both experiments. Thus, contrary to the general reduced inhibition hypothesis, the specific process of response suppression during serial recall is not reduced by aging. In recent years, there has been considerable interest in inhibitory processes in cognition (see Dempster, 1992; Dempster & Brainerd, 1995, for reviews) and, in particular, in their development during childhood (e.g., Harnishfeger, 1995) and decline with normal

This study reports on the results of five discontinuous typing paradigm experiments in which subjects (native English speakers in experiment 1 and native German speakers in experiments 2 to 5) had to type words presented to them in various modes. In experiment 1 the words were presented in visual form. In experiment 2 words were presented orally and the results are compared with typing following visual word presentation. Experiment 3 compares typing following visual word and picture presentation. In experiment 4 subjects were required to type pseudo-words, whilst in the final experiment the typing responses, following oral and visual word presentation, were delayed by an extended preparatory period. In all experiments we found that the increase of inter-keystroke intervals (IKIs) was highly significant at positions that where either exclusively syllable (S) boundaries or combined syllable and morpheme (SM) boundaries. SM type IKIs are significantly larger than S type IKIs and are influenced by word frequencies, indicating lexical dependencies. SM type IKIs were found to be significantly longer for oral than for visual word presentation. This is taken as an indication that additional processes (phonological-graphemic mediation) are involved in the accessing of graphemic word forms when words are presented aurally. The fact that pseudo-words are also written with increased IKIs at syllable borders indicates that at least one major component of the S-type IKIs is produced by bypassing the lexicon, probably at sublexical levels. The fact that augmented SM and S type IKIs are also found in the delayed typing task indicates that input into the motor system is constituted by sub-word units instead by fully specified words. As SM and S type IKIs reflect influences of different ...