Publication Type: Conference Paper/Unpublished Manuscript
Abstract: In a word learning situation, a child may face many possible meanings for a given novel word. It has been argued (Gleitman, 1990) that the syntactic structures in which new words occur may help constrain their possible meanings. But can infants use syntax early enough that lexical acquisition will be significantly improved? We recently observed that 23 month-olds use the syntactic context in which novel words occur to decide whether they should be mapped towards an object or an action (Bernal & al, 2005). However, the question remains as to whether infants of this age have only developed a way of categorizing the input based on local co-occurrences or whether they already compute some syntactic analysis online.
In this experiment we developed a novel way of asking whether infants develop syntactic expectancies during online processing. In French, the word “la” can either be used as an article (e.g. “je prends la poire”, I take the apple) or as a reference pronoun (e.g. “je la mange”, I eat it). Depending on the beginning of the sentence, a native speaker of French will thus expect either a noun or a verb to follow the function word “la”. We exposed our subjects to short auditory stories in French and recorded their EEG activity. Sentences were either Grammatical (e.g. je prends la poire or je la mange) or Ungrammatical (e.g. je prends la mange or je la poire). Note that none of the local co-occurrence between the function and content words (e.g. la poire) was in itself ungrammatical; rather, some endings were in contradiction with the expectancies built after processing the beginning of the sentence. We then compared the EEG response evoked by these two conditions. We predicted that adults should show a classical syntactic mismatch response.
In adults (N=12), the presence of a grammatical category mistake gave rise to both an early frontal mismatch response (100-200 ms post the onset of the critical word), disrupting the normal N1-P1 complex as well as to a classical syntactic positive shift (or P600) effect. We are currently running the 2 year-olds infant experiment (N=8 at present), in order to see whether they will also exhibit specific responses to the syntactic mismatches.

Publication Type: Conference Paper/Unpublished Manuscript
Review Method: Peer Reviewed
Abstract: Using a series of event-related potential analyses, a technique of electroencephalography, the present study examines how brains of perceivers react differently depending on whether the visual stimuli seem appropriate or not. The present study especially considers the effects of social contexts—whether the combinations of shirts and ties seem appropriate for a job interview and for a party. The behavioral data revealed that the reaction times were faster when the stimuli were judged as appropriate for the party condition. For the job interview condition, however, reaction times were faster with the inappropriate judgments. The electroencephalography analyses detected N270 component, known to be associated with cognitive conflicts. The results imply that electroencephalography analyses would be useful for examining perceived appropriateness (or conflicts) of combinations of a set of texts and texts (brand names slogans), texts and images (images of items and headlines in a visual ad), photos and captions, to name a few.

Publication Type: Conference Paper/Unpublished Manuscript
Abstract: This paper develops a sensitizing concept, “fit-gap work,” to integrate institutional and adaptation perspectives on the implementation of innovations in organizations. Two cases are examined with interview and archival data on the implementation of Enterprise Resource Planning (ERP) systems and the dual-consequences, which stem from this process, for both organizational form and the form of the innovation. Results suggest that early adopters customize ERP systems in strategic ways anticipating that their local innovations will become part of the vendor’s “standard” product, while later adopters use implementation as a strategic opportunity to remove unwanted elements of their organizational form and partake in what I call “organizational forgetting.”

Publication Type: Conference Paper/Unpublished Manuscript
Abstract: Enterprise Resource Planning (or ERP) systems are very expensive and can significantly disrupt the workplace. The extent of this disruption, however, will depend on the goals of implementation. In this respect we can identify two different goals or rationales for ERP implementations: 1) upgrading independent legacy systems in order to better support existing workplace practices; or 2) using a new technological system to fundamentally transform the organization processes. If an organization has the first goal, technology is seen as facilitating work by providing better tools to employees. In the second rationale, technology is seen as (re)structuring work by providing a framework through which work gets done. This paper critically explores the role of consultants in supporting ERP implementations. This paper explores the divergence between how ERP systems are designed and implemented based on workplace process versus actual workplace practice. This paper argues that the adopted methodology used by consultants only gives them an “idealized” of business processes. The paper proposes instead that consultants adopt a practice-based approach rooted in the detail of daily workplace activities. It is this view that can better inform system design, as well as business, organizational, and managerial decisions.

Publication Type: Individual Poster
Abstract: Background and Aims: One of the cues available for language learners to segment continuous auditory stream into distinctive words is the transition probability between vocal tokens. Segmentation of speech stream is a prerequisite for language acquisition, especially for infants. Behavioral experiments utilizing head-turn response and habituation paradigms have shown that 8-month-old infants could detect word boundaries using statistical cues. Here we ask whether this ability could be considered as “innate” by testing newborns utilizing event-related potentials (ERPs). Since we consider that the ability for statistical learning based on transitional probabilities is not specific to language, but more general cognitive competence, we used tones instead of phonemes in our study.

Methods: In this study, we examine whether neonates can learn transitional probabilities and statistically segment words from a continuous sound sequence. We measured ERPs for 11 neonates (0～5days after birth). 7 electrodes were placed according to the International 10-20 Electrode System at F3, Fz, F4, T3, Cz, T4 and Pz. During recordings of the electroencephalography, neonates were presented with tone sequences. This sound sequences consisted twelve pure tones of the same octave. Three of these tones were randomly combined into one “tone word” (cf. Saffran 1999). Four “tone words” were presented in random order without intervals for 5 min. Transitional probabilities between tones within a “words” were 1.00, but transitional probabilities between the word boundaries were 0.33.

Key Results: If neonates can segment tone-words from these sound sequences, the first tone of each word should evoke a conspicuous response compared with the second or third tones. As a result, only the first tone of each word evoked a significant positive component in the frontal area. It suggests that a brain correlate of boundary detection. Since this potential is not evident during the first session, this is likely to be due to statistical learning. A similar our study in adults suggested that the polarity of the ERP correlated with perceptual segmentation was negative. The opposite polarity found in infants may represent immaturity of their brain.

Conclusion: The ability to distinguish words in a speech stream based on statistical information is innately prepared in humans, as our newborn subject showed boundary specific event related responses.