How to balance Balanced Reporting and Reliable Reporting

Philosophical Studies - The paper draws on philosophy of science to help resolve a tension between two central journalistic ideals: That of resenting diverse viewpoints (Balanced Reporting) and...     

Decisions, decisions: infant language learning when multiple generalizations are possible

Two experiments presented infants with artificial language input in which at least two generalizations were logically possible. The results demonstrate that infants made one of the two generalizations tested, the one that was most statistically consistent with the particular subset of the data they received. The experiments shed light on how learners might go about solving the induction problem for human language.     

Infants use rational decision criteria for choosing among models of their input

Previous work demonstrated that 9-month-olds who were familiarized with 3-syllable strings consistent with both a broader (AAB or ABA) and narrower (AAdi or AdiA) generalization made only the latter. Because the narrower generalization is a subset of the broader one, any example that is consistent with the broader generalization but not the narrower one should allow a rational learner to select the broader generalization. The current experiment asked whether infants show evidence of being such learners. Infants who heard the stimuli that previously led to the narrower generalization plus three counterexamples mixed into the last five stimuli made the broader generalization at test. A control condition ruled out the possibility that infants based their generalization on the last five familiarization stimuli. The new findings suggest that infants effectively consider multiple competing models for their input and use rational decision criteria for selecting among these models.     

Assessment of preschoolers at-risk with the WPPSI--R and the Stanford-Binet L-M.

The purpose of this study was to provide data on the validity of using the WPPSI--R with 16 preschoolers who had been referred for psychological assessment. A comparison of the preschoolers' performance on the WPPSI--R and the Stanford-Binet L-M was made. Children ranged in age from 3-0 to 6-0 years, with a mean age of 4-5 yr. The correlations between WPPSI--R and Stanford-Binet L-M scores were strong; WPPSI--R Verbal IQ and Stanford-Binet L-M IQ, r = .85; WPPSI--R Performance IQ and Stanford-Binet L-M IQ, r = .75, and WPPSI--R Full Scale IQ and Stanford-Binet L-M IQ, r = .82. For this very small sample of children whose mean IQ was significantly below average, the WPPSI--R appeared to assess validly their current intellectual functioning. Attention should be given concurrent and predictive validity as well as the efficiency of using the WPPSI--R with other and larger samples of referred children.     

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.     

Infants avoid 'labouring in vain' by attending more to learnable than unlearnable linguistic patterns

Every environment contains infinite potential features and correlations among features, or patterns. Detecting valid and learnable patterns in one environment is beneficial for learners because doing so lends predictability to new environments where the same or analogous patterns recur. However, some apparent correlations among features reflect spurious patterns, and attempting to learn the latter costs time and resources with no advantage to the learner. Thus, an efficient learner in a complex environment needs to devote more attention to input that reflects a real and learnable pattern than to input that reflects a spurious or ultimately unlearnable pattern. However, in order to achieve such efficiency in the absence of external feedback, learners need to have an implicit metric of their own learning progress. Do human infants have such a metric? Data from two experiments demonstrate that 17‐month‐olds attend longer to learnable vs. unlearnable linguistic grammars, taking more trials to habituate and more overall time to habituate for grammars in which a valid generalization over input stimuli can be made. These data provide the first evidence that infants have an implicit metric of their own learning progress and preferentially direct their attention to learnable aspects of their environment.     

The basis of transfer in artificial grammar learning

In two experiments, we examined the extent to which knowledge of sequential dependencies and/or patterns of repeating elements is used during transfer in artificial grammar learning. According to one view of transfer, learners abstract the grammar's sequential dependencies and then learn a mapping to new vocabulary at test (Dienes, Altmann, & Gao, 1999). Elements that are repeated have no special status on this view, and so a logical prediction is that learners should transfer as well after exposure to a grammar without repetitions as after exposure to a grammar with them. On another view, repetition structure is the very basis of transfer (Brooks & Vokey, 1991; Mathews & Roussel, 1997). Learners were trained on grammars with or without repeating elements to test these competing views. Learners demonstrated considerable knowledge of sequential dependencies in their training vocabulary but did not use such knowledge to transfer to a new vocabulary. Transfer only occurred in the presence of repetition structure, demonstrating this to be the basis of transfer.     

Linguistic Intuitions are the Result of Interactions Between Perceptual Processes and Linguistic Universals

We found a direct relationship between variation in informants' grammaticality intuitions about pronoun coreference and variation in the same informants' use of a clause segmentation strategy during sentence perception. It has been proproposed that ‘c-command’, a structural principle defined in terms of constituent dominance relations, constrains within-sentence coreference between pronouns and noun antecedents. The relative height of the pronoun and the noun in the phrase structure hierarchy determines whether the c-command constraint blocks coreference: Coreference is allowed only when the complement structure containing the noun is attached higher than the pronoun. We collected informants' judgments on pronoun-noun coreference in which the noun antecedent was contained in a complement structure dominated by either the Sentence-node (S-node) (higher than the pronoun) or the Verb-phrase-node (VP-node) (not higher than the pronoun). We also assessed each informant's perceptual clause-closure tendency using an auditory word-monitor paradigm. Informants who strongly segmented clauses in the perceptual task did not differentiate between an S- and VP-attachment of sentence complements, as revealed in their coreference judgments, but rather appeared to attach all sentence complements to the S-node. Informants with relatively weak perceptual segmentation differentiated their coreference judgments according to the node attachment of the complement structure. These results indicate that the linguistic universal controlling within-sentence coreference applies to the perceptually available structure for a sequence, not to its pure linguistic structure. Hence, linguistic intuitions result from the interaction of three independent faculties: language-specific knowledge, perceptual processes, and linguistic universals.     

Surprise! Infants consider possible bases of generalization for a single input example

Infants have been shown to generalize from a small number of input examples. However, existing studies allow two possible means of generalization. One is via a process of noting similarities shared by several examples. Alternatively, generalization may reflect an implicit desire to explain the input. The latter view suggests that generalization might occur when even a single input example is surprising, given the learner's current model of the domain. To test the possibility that infants are able to generalize based on a single example, we familiarized 9‐month‐olds with a single three‐syllable input example that contained either one surprising feature (syllable repetition, Experiment 1) or two features (repetition and a rare syllable, Experiment 2). In both experiments, infants generalized only to new strings that maintained all of the surprising features from familiarization. This research suggests that surprise can promote very rapid generalization.