Rebecca L. Gómez

Variability and Detection of Invariant Structure

Two experiments investigated learning of nonadjacent dependencies by adults and 18-month-olds. Each learner was exposed to three-element strings (e.g., pel-kicey-jic) produced by one of two artificial languages. Both languages contained the same adjacent dependencies, so learners could distinguish the languages only by acquiring dependencies between the first and third elements (the nonadjacent dependencies). The size of the pool from which the middle elements were drawn was systematically varied to investigate whether increasing variability (in the form of decreasing predictability between adjacent elements) would lead to better detection of nonadjacent dependencies. Infants and adults acquired nonadjacent dependencies only when adjacent dependencies were least predictable. The results point to conditions that might lead learners to focus on nonadjacent versus adjacent dependencies and are important for suggesting how learning might be dynamically guided by statistical structure.

Artificial Grammar Learning by 1-Year-Olds Leads to Specific and Abstract Knowledge.

Four experiments used the head-turn preference procedure to assess whether infants could extract and remember information from auditory strings produced by a miniature artificial grammar. In all four experiments, infants generalized to new structure by discriminating new grammatical strings from ungrammatical ones after less than 2 min exposure to the grammar. Infants acquired specific information about the grammar as demonstrated by the ability to discriminate new grammatical strings from those with illegal endpoints (Experiment 1). Infants also discriminated new grammatical strings from those with string-internal pairwise violations (Experiments 2 and 3). Infants in Experiment 4 abstracted beyond specific word order as demonstrated by the ability to discriminate new strings produced by their training grammar from strings produced by another grammar despite a change in vocabulary between training and test. We discuss the implications of these findings for the study of language acquisition.

Infant artificial language learning and language acquisition

The rapidity with which children acquire language is one of the mysteries of human cognition. A view held widely for the past 30 years is that children master language by means of a language-specific learning device. An earlier proposal, which has generated renewed interest, is that children make use of domain-general, associative learning mechanisms. However, our current lack of knowledge of the actual learning mechanisms involved during infancy makes it difficult to determine the relative contributions of innate and acquired knowledge. A recent approach to studying this problem exposes infants to artificial languages and assesses the resulting learning. In this article, we review studies using this paradigm that have led to a number of exciting discoveries regarding the learning mechanisms available during infancy. These studies raise important issues with respect to whether such mechanisms are general or specific to language, the extent to which they reflect statistical learning versus symbol manipulation, and the extent to which such mechanisms change with development. The fine-grained characterizations of infant learning mechanisms that this approach permits should result in a better understanding of the relative contributions of, and the dynamic between, innate and learned factors in language acquisition.

Naps Promote Abstraction in Language-Learning Infants

Infants engage in an extraordinary amount of learning during their waking hours even though much of their day is consumed by sleep. What role does sleep play in infant learning? Fifteen-month-olds were familiarized with an artificial language 4 hr prior to a lab visit. Learning the language involved relating initial and final words in auditory strings by remembering the exact word dependencies or by remembering an abstract relation between initial and final words. One group napped during the interval between familiarization and test. Another group did not nap. Infants who napped appeared to remember a more abstract relation, one they could apply to stimuli that were similar but not identical to those from familiarization. Infants who did not nap showed a memory effect. Naps appear to promote a qualitative change in memory, one involving greater flexibility in learning.

The dynamics of memory: Context-dependent updating

Understanding the dynamics of memory change is one of the current challenges facing cognitive neuroscience. Recent animal work on memory reconsolidation shows that memories can be altered long after acquisition. When reactivated, memories can be modified and require a restabilization (reconsolidation) process. We recently extended this finding to human episodic memory by showing that memory reactivation mediates the incorporation of new information into existing memory. Here we show that the spatial context plays a unique role for this type of memory updating: Being in the same spatial context during original and new learning is both necessary and sufficient for the incorporation of new information into existing episodic memories. Memories are automatically reactivated when subjects return to an original learning context, where updating by incorporating new contents can occur. However, when in a novel context, updating of existing memories does not occur, and a new episodic memory is created instead.

Memory formation, consolidation and transformation

Memory formation is a highly dynamic process. In this review we discuss traditional views of memory and offer some ideas about the nature of memory formation and transformation. We argue that memory traces are transformed over time in a number of ways, but that understanding these transformations requires careful analysis of the various representations and linkages that result from an experience. These transformations can involve: (1) the selective strengthening of only some, but not all, traces as a function of synaptic rescaling, or some other process that can result in selective survival of some traces; (2) the integration (or assimilation) of new information into existing knowledge stores; (3) the establishment of new linkages within existing knowledge stores; and (4) the up-dating of an existing episodic memory. We relate these ideas to our own work on reconsolidation to provide some grounding to our speculations that we hope will spark some new thinking in an area that is in need of transformation.

Transfer and Complexity in Artificial Grammar Learning

Implicit and explicit learning are sensitive to various degrees of complexity and abstractness, ranging from knowledge of first-order dependencies and specific surface structure to second-order dependencies and transfer. Three experiments addressed whether implicit learning is sensitive to this entire range of information or whether explicit knowledge becomes an important factor in cases of more complex learning. Experiment 1 used recognition and prediction to assess deliberate access to knowledge of letter patterns in an artificial grammar learning paradigm. Experiment 2 manipulated stimulus presentation and response in a sequence-based grammar learning paradigm. Learning can occur without awareness in cases of lesser complexity (such as learning first-order dependencies). However, more complex learning, such as that involved in learning second-order dependencies or in transfer to stimuli with the same underlying syntax but new surface features is linked to explicit knowledge. In contrast to Experiments 1 and 2 which assessed deliberate access to knowledge of the acquisition stimuli, Experiment 3 assessed deliberate access to knowledge of the transfer stimuli. Knowledge of initial trigrams in the transfer stimuli appears to play an important role in transfer. These findings are evaluated in terms of postulated implicit learning mechanisms.

Episodic memory reconsolidation: Updating or source confusion?

Reactivation of apparently stable, long-term memory can render it fragile, and dependent on a re-stabilisation process referred to as “reconsolidation”. Recently we provided the first demonstration of reconsolidation effects in human episodic memory (Hupbach, Gomez, Hardt, & Nadel, 2007; Hupbach, Hardt, Gomez, & Nadel, 2008). Memory for a set of objects was modified by the presentation of a new set, if and only if participants were reminded of the first learning episode before learning the new set. The present study asks whether this effect can be interpreted as a source discrimination problem; i.e., participants have difficulties remembering which objects were presented during which session, and do not actually incorporate new objects into the reactivated memory. The present study used a recognition test and asked participants directly about the source of their memories. Participants in the no-reminder group showed very few source errors. Participants in the reminder group misattributed the source of objects from the second set as being from the first set but not vice versa, thus demonstrating updating of the original memory. This finding is informative with respect to the misinformation paradigm, and reconsolidation is discussed as a possible mechanism underlying our results and the misinformation effect.

What Is Learned From Artificial Grammars? Transfer Tests of Simple Association

Ss were trained on letter pairs or letter strings in an artificial grammar learning paradigm to determine the extent to which implicit learning is driven by simple associative knowledge. Learning on strings resulted in sensitivity to violations of grammaticality and in transfer to a changed letter set. Learning on letter pairs resulted in less sensitivity and no transfer. Discrepancies in performance were later reduced, but not eliminated, by equating the task demands of the conditions during learning. A direct test of associative knowledge showed that training on letter pairs resulted in knowledge of legal bigrams, but this knowledge was only weakly related to violation sensitivity. The experiments demonstrate that knowledge of isolated associations is sufficient to support some learning, but this knowledge cannot explain the more abstract knowledge that results from learning on complete exemplars.

Sensitivity to word order cues by normal and language/learning disabled adults.

UNLABELLED Sixteen adults with language/learning disabilities (L/LD) and 16 adults who lacked a personal or familial history of L/LD participated in a study designed to test sensitivity to word order cues that signaled grammatical versus ungrammatical word strings belonging to an artificial grammar. In an exposure phase, participants heard word strings constructed of novel CVC words for a period of 5 min. In a test phase, participants were asked to judge new sentences as either obeying or violating the rules of the grammar they heard. L/LD participants performed significantly below the comparison group on this task. The results suggest that this skill, which emerges early in life for normal children, is problematic for adults with L/LD. LEARNING OUTCOMES The reader will become familiar with a paradigm that allows assessment of rapid learning of word order rules and how this learning differs for normal and language/learning disabled adults.