Blair C. Armstrong

Domain generality versus modality specificity: the paradox of statistical learning

Statistical learning (SL) is typically considered to be a domain-general mechanism by which cognitive systems discover the underlying distributional properties of the input. However, recent studies examining whether there are commonalities in the learning of distributional information across different domains or modalities consistently reveal modality and stimulus specificity. Therefore, important questions are how and why a hypothesized domain-general learning mechanism systematically produces such effects. Here, we offer a theoretical framework according to which SL is not a unitary mechanism, but a set of domain-general computational principles that operate in different modalities and, therefore, are subject to the specific constraints characteristic of their respective brain regions. This framework offers testable predictions and we discuss its computational and neurobiological plausibility.

The what, when, where, and how of visual word recognition

A long-standing debate in reading research is whether printed words are perceived in a feedforward manner on the basis of orthographic information, with other representations such as semantics and phonology activated subsequently, or whether the system is fully interactive and feedback from these representations shapes early visual word recognition. We review recent evidence from behavioral, functional magnetic resonance imaging, electroencephalography, magnetoencephalography, and biologically plausible connectionist modeling approaches, focusing on how each approach provides insight into the temporal flow of information in the lexical system. We conclude that, consistent with interactive accounts, higher-order linguistic representations modulate early orthographic processing. We also discuss how biologically plausible interactive frameworks and coordinated empirical and computational work can advance theories of visual word recognition and other domains (e.g., object recognition).

Brainprint: Assessing the uniqueness, collectability, and permanence of a novel method for ERP biometrics

Abstract The human brain continually generates electrical potentials representing neural communication. These potentials can be measured at the scalp, and constitute the electroencephalogram (EEG). When the EEG is time-locked to stimulation – such as the presentation of a word – and averaged over many such presentations, the Event-Related Potential (ERP) is obtained. The functional characteristics of components of the ERP are well understood, and some components represent processing that may differ uniquely from individual to individual—such as the N400 component, which represents access to the semantic network. We applied several pattern classifiers to ERPs representing the response of individuals to a stream of text designed to be idiosyncratically familiar to different individuals. Results indicate that there are robustly identifiable features of the ERP that enable labeling of ERPs as belonging to individuals with accuracy reliably above chance (in the range of 82–97%). Further, these features are stable over time, as indicated by continued accurate identification of individuals from ERPs after a lag of up to six months. Even better, the high degree of labeling accuracy achieved in all cases was achieved with the use of only 3 electrodes on the scalp—the minimal possible number that can acquire clean data.

SOS! An algorithm and software for the stochastic optimization of stimuli.

The characteristics of the stimuli used in an experiment critically determine the theoretical questions the experiment can address. Yet there is relatively little methodological support for selecting optimal sets of items, and most researchers still carry out this process by hand. In this research, we present SOS, an algorithm and software package for the stochastic optimization of stimuli. SOS takes its inspiration from a simple manual stimulus selection heuristic that has been formalized and refined as a stochastic relaxation search. The algorithm rapidly and reliably selects a subset of possible stimuli that optimally satisfy the constraints imposed by an experimenter. This allows the experimenter to focus on selecting an optimization problem that suits his or her theoretical question and to avoid the tedious task of manually selecting stimuli. We detail how this optimization algorithm, combined with a vocabulary of constraints that define optimal sets, allows for the quick and rigorous assessment and maximization of the internal and external validity of experimental items. In doing so, the algorithm facilitates research using factorial, multiple/mixed-effects regression, and other experimental designs. We demonstrate the use of SOS with a case study and discuss other research situations that could benefit from this tool. Support for the generality of the algorithm is demonstrated through Monte Carlo simulations on a range of optimization problems faced by psychologists. The software implementation of SOS and a user manual are provided free of charge for academic purposes as precompiled binaries and MATLAB source files at http://sos.cnbc.cmu.edu.

PSPs and ERPs: Applying the dynamics of post-synaptic potentials to individual units in simulation of temporally extended Event-Related Potential reading data.

The Parallel Distributed Processing (PDP) framework is built on neural-style computation, and is thus well-suited for simulating the neural implementation of cognition. However, relatively little cognitive modeling work has concerned neural measures, instead focusing on behavior. Here, we extend a PDP model of reading-related components in the Event-Related Potential (ERP) to simulation of the N400 repetition effect. We accomplish this by incorporating the dynamics of cortical post-synaptic potentials--the source of the ERP signal--into the model. Simulations demonstrate that application of these dynamics is critical for model elicitation of repetition effects in the time and frequency domains. We conclude that by advancing a neurocomputational understanding of repetition effects, we are able to posit an interpretation of their source that is both explicitly specified and mechanistically different from the well-accepted cognitive one.

The long road of statistical learning research: past, present and future

Almost all types of learning involve, to some degree, the ability to encode regularities across time and space. Although statistical learning (SL) research initially focused on offering a viable alternative to rule-based grammars and specialized mechanisms for word learning (e.g. [[1][1],[2][2]]),

eDom: Norming software and relative meaning frequencies for 544 English homonyms

Words that are homonyms—that is, for which a single written and spoken form is associated with multiple, unrelated interpretations, such as COMPOUND, which can denote an < enclosure > or a < composite > meaning—are an invaluable class of items for studying word and discourse comprehension. When using homonyms as stimuli, it is critical to control for the relative frequencies of each interpretation, because this variable can drastically alter the empirical effects of homonymy. Currently, the standard method for estimating these frequencies is based on the classification of free associates generated for a homonym, but this approach is both assumption-laden and resource-demanding. Here, we outline an alternative norming methodology based on explicit ratings of the relative meaning frequencies of dictionary definitions. To evaluate this method, we collected and analyzed data in a norming study involving 544 English homonyms, using the eDom norming software that we developed for this purpose. Dictionary definitions were generally sufficient to exhaustively cover word meanings, and the methods converged on stable norms with fewer data and less effort on the part of the experimenter. The predictive validity of the norms was demonstrated in analyses of lexical decision data from the English Lexicon Project (Balota et al., Behavior Research Methods, 39, 445–459, 2007), and from Armstrong and Plaut (Proceedings of the 33rd Annual Meeting of the Cognitive Science Society, 2223–2228, 2011). On the basis of these results, our norming method obviates relying on the unsubstantiated assumptions involved in estimating relative meaning frequencies on the basis of classification of free associates. Additional details of the norming procedure, the meaning frequency norms, and the source code, standalone binaries, and user manual for the software are available at http://edom.cnbc.cmu.edu.

Chronset: An automated tool for detecting speech onset

The analysis of speech onset times has a longstanding tradition in experimental psychology as a measure of how a stimulus influences a spoken response. Yet the lack of accurate automatic methods to measure such effects forces researchers to rely on time-intensive manual or semiautomatic techniques. Here we present Chronset, a fully automated tool that estimates speech onset on the basis of multiple acoustic features extracted via multitaper spectral analysis. Using statistical optimization techniques, we show that the present approach generalizes across different languages and speaker populations, and that it extracts speech onset latencies that agree closely with those from human observations. Finally, we show how the present approach can be integrated with previous work (Jansen & Watter Behavior Research Methods, 40:744–751, 2008) to further improve the precision of onset detection. Chronset is publicly available online at www.bcbl.eu/databases/chronset.

Ultrastructural synaptic changes associated with neurofibromatosis type 1: a quantitative analysis of hippocampal region CA1 in a Nf1(+/-) mouse model.

Neurofibromatosis type 1 (NF1) is one of the most frequently diagnosed autosomal dominant inherited disorders resulting in neurological dysfunction, including an assortment of learning disabilities and cognitive deficits. To elucidate the neural mechanisms underlying the disorder, we employed a mouse model (Nf1+/−) to conduct a quantitative analysis of ultrastructural changes associated with the NF1 disorder. Using both serial light and electron microscopy, we examined reconstructions of the CA1 region of the hippocampus, which is known to play a central role in many of the dysfunctions associated with NF1. In general, the morphology of synapses in both the Nf1+/− and wild‐type groups of animals were similar. No differences were observed in synapse per neuron density, pre‐ and postsynaptic areas, or lengths. However, concave synapses were found to show a lower degree of curvature in the Nf1+/− mutant than in the wild type. These results indicate that the synaptic ultrastructure of Nf1+/− mice appears relatively normal with the exception of the degree of synaptic curvature in concave synapses, adding further support to the importance of synaptic curvature in synaptic plasticity, learning, and memory. Synapse, 2012.

Generalization from newly learned words reveals structural properties of the human reading system.

Connectionist accounts of quasiregular domains, such as spelling–sound correspondences in English, represent exception words (e.g., pint) amid regular words (e.g., mint) via a graded “warping” mechanism. Warping allows the model to extend the dominant pronunciation to nonwords (regularization) with minimal interference (spillover) from the exceptions. We tested for a behavioral marker of warping by investigating the degree to which participants generalized from newly learned made-up words, which ranged from sharing the dominant pronunciation (regulars), a subordinate pronunciation (ambiguous), or a previously nonexistent (exception) pronunciation. The new words were learned over 2 days, and generalization was assessed 48 hr later using nonword neighbors of the new words in a tempo naming task. The frequency of regularization (a measure of generalization) was directly related to degree of warping required to learn the pronunciation of the new word. Simulations using the Plaut, McClelland, Seidenberg, and Patterson (1996) model further support a warping interpretation. These findings highlight the need to develop theories of representation that are integrally tied to how those representations are learned and generalized.