Sheila E. Blumstein - US grants

A theory of acoustic invariance claims that the universal set of phonetic features used in natural language can be characterized by invariant properties derived directly from the acoustic signal, that the perceptual mechanism is specifically tuned to extract these properties, and that they form the primary perceptual attributes used in ongoing speech perception. It is the goal of the proposed research to explore these issues. To this end, we plan to continue to investigate place of articulation in stop consonants by exploring the role of spectral shape as a perceptual cue to the labial and alveolar place of articulation and by studying the acoustic properties of another place of articulation, dentals in stop consonants. Our methodology will consist of acoustic analysis of natural speech from English and othe languages as well as perceptual studies utilizing both natural and synthetic speech stimuli. In addition, we plan to explore the feature contrast, stop-glide, in relation to a theory of acoustic invariance. Our methodology will consist of attempting to isolate the invariant property after detailed acoustic anslyses of natural speech, providing an operational definition of that property, testing its effectiveness on a new set of natural speech data, and then exploring the perceptual consequences of this property using both synthetic and natural speech stimuli for the identification of the particular property. Our ultimate aim is to characterize the inventory of acoustic properties corresponding to the finite set of speech sounds used in natural language. Such study should help characterize the nature of the speech processing system and its underlying mechanisms, as well as provide an essential base-line for delineating the developmental processes for speech in the young child and the effects of brain-damge on such processing abilities.

The broad goal of this research is to provide a window into those aspects of the cognitive architecture of the lexical processing system that are impaired and those that are spared in aphasia and to examine the neural systems underlying such processing. Specific aims are to examine lexical processing and the effects of competition on mapping from sound to meaning in auditory word recognition and on mapping from meaning to sound in spoken word production to test three hypotheses. Hypothesis 1: Access to lexical form recruits parietal areas including the supramarginal gyrus (SMG), whereas access to lexical meaning recruits frontal areas including the inferior frontal gyrus (IFG). It is hypothesized that under conditions of lexical competition Conduction aphasics with parietal damage including the SMG will exhibit deficits when accessing phonological form under conditions of lexical competition and Broca's aphasics with frontal lesions including the IFG will exhibit deficits when accessing lexical meaning under conditions of lexical competition. Hypothesis 2: Aphasic patients have normal interactive processes. This hypothesis will be tested in two ways. First, based on normal results in spoken word production showing the influence of phonological lexical competition on processes downstream from it, i.e. on acoustic phonetic output, it is hypothesized that both Wernicke's aphasics with lesions including the superior temporal gyrus (STG) and Conduction aphasics with lesions including the SMG who show impairments in accessing lexical form under conditions of lexical competition will show smaller or no influence of lexical competition on their acoustic-phonetic output. In contrast, Broca's aphasics with IFG lesions who show normal patterns of performance under such conditions will show lexical competition effects on their acoustic-phonetic output. Second, it is hypothesized that information from higher-level sources including lexical-semantics and sentence context will override lexical deficits that emerge under conditions of phonological competition. Hypothesis 3: Lexical deficits in aphasia are not independently affected in auditory word recognition and spoken word production, implicating a central deficit. Thus, Conduction aphasic patients with SMG lesions who show deficits in auditory word recognition under conditions of phonological lexical competition should show concomitant impairments in spoken word production, and Broca's aphasics with lesions involving the IFG who show normal patterns of performance in auditory word recognition under conditions of phonological lexical competition should show normal patterns of performance in spoken word production. To test these hypotheses, we will use a variety of experimental paradigms including eye-tracking, same-different judgment, lexical decision, semantic priming, and word reading and picture naming followed by acoustic analysis of the spoken output. Relevance: Understanding the basis of these impairments is one step towards to development of appropriate rehabilitation strategies.

Six undergraduate Courses are being developed which unite Science and the Humanities.

9453676 Blumstein Nationally, many fewer women than men study science as undergraduates and ultimately pursue careers as scientists. This disparity is not the result of any inherent difference in ability between men and women, but instead grows out of the culture of undergraduate science: the climate of science classrooms, and the way in which science classes tends to be taught, are often such that women who have an initial interest in science choose to leave in favor of academic fields that they find more congenial. The first year of college is an especially critical time, as it is during this period that students are first exposed to college-level science and form their opinions about whether or not they wish to become scientists. With previous support from the National Science Foundation, Brown University initiated the Women in Science and Engineering (WISE) program, which was designed to enhance the environment experienced by all undergraduate women science students. While successful, it did not focus specifically on the concerns of first-year students, an area which student researchers from the program itself have identified as a particularly crucial area. Thus, the development and inauguration of a new program, to be called WISE Beginnings, is proposed. Building on the successes of the original WISE program, the goal of this new initiative will be to increase the participation and retention of women in Brown's undergraduate science programs by focusing on and enhancing the experience of first- year women in introductory science courses, the first major challenge that these students face. The program will develop a new paradigm for introductory science courses which brings together many recently developed approaches which have been shown to make science education more inclusive. WISE Beginnings will feature three components: 1) an academic program centered on study groups, facilitated by upperclass women science students, that will be attached to each of the introductory science classes at Brown, will provide a feedback mechanism to faculty about the nature of the problems and challenges the students face, and are anticipated to result in permanent changes in introductory science courses and curricula; 2) an informational and advising program to run during Brown's Orientation Week; and 3) a co-curricular program of academic, advising, and other support activities that will complement students' in class experience. ***

9450319 Wyche Brown University proposes to expand its Summer Research-Early Identification Program (SR-EIP) in the biomedical sciences to a year-round research and career oriented experience for highly motivated Brown underrepresented minority students in the life sciences and engineering. The program, Research Careers for Minority Scholars (RCMS) at Brown University, will involve talented students from the freshman through senior year and expand the career options for those students during the four-year period of the proposal. Students will be matched with faculty sponsors who will provide mentorship, guidance, and a first-rate research environment starting in the freshman year. During the summer following the freshman year, students will participated in a nine- week program comprising a continuation of the research experience, seminars and field trips, weekly research colloquia by Brown faculty and outside speakers, and visits to industrial companies. During the junior and senior year, students will be encouraged to apply to Brown's Ph.D. programs and to other highly selected graduate schools. One measure of the RCMS experience's success at Brown will be the number of recruits to graduate degree programs. The program's goal will be to produce a minimum of 15 pre-doctoral candidates by the year 2000.

DESCRIPTION (provided by applicant): The goal of this research is to understand how the functional architecture of the speech-lexical processing system maps on to the neural systems underlying it using event-related functional magnetic resonance imaging (fMRI). Two hypotheses will be investigated, both exploring the neural systems underlying the computational properties of auditory speech and lexical processing. The first hypothesis explores whether the computational mechanisms of the two hemispheres preferentially process different properties of the speech signal. A series of experiments using both speech discrimination and phonetic categorization tasks will examine a number of acoustic-phonetic parameters underlying both consonant and vowel perception including voice-onset time and formant frequency duration in the perception of consonants, and duration and formant frequency (F l-F2) space in the perception of vowels. It is hypothesized that the acoustic properties corresponding to the phonetic categories of speech will be preferentially processed by the left hemisphere irrespective of their acoustic structure owing to the functional role that sound structure plays in language processing. In contrast, it is hypothesized that activation of the right hemisphere will be modulated by the inherent acoustic properties of the speech signal. The second hypothesis explores how the computational properties of the speech lexical processing system map on to its neural substrate. It is hypothesized that there will be increased left frontal activation under conditions of phonetic-phonological, lexical, and/or semantic competition. Phonetic competition will be investigated by exploring the extent to which acoustic-phonetic structure and the proximity of an exemplar to a contrasting, and hence, competing phonetic category influences frontal activation. Lexical competition will be investigated by exploring the influence of phonetic factors on lexical competition and phonological factors on lexical neighborhood density. Semantic competition will be investigated by exploring the effects of semantically ambiguous words compared to unambiguous words on neural activation patterns.

With support from a National Science Foundation Major Research Instrumentation Award, Brown University will acquire a 3 Tesla magnetic resonance imaging (MRI) system. The MRI system will become housed in a research-dedicated MRI suite within the newly constructed Life Sciences Building at Brown, and it will form the core infrastructure for MRI-related research conducted by more than 100 faculty, research staff, and students in the Brown University community, including its College of Arts and Sciences and Medical School. Researchers at Brown will use the NSF-fund MRI system primarily to investigate fundamentals of brain structure and function. In addition to Brown users, researchers from other nearby institutions, such as the University of Rhode Island, Regina Saliva University, and University of Massachusetts-Dartmouth can have access to the 3 Tesla MRI system.

Non-invasive imaging of the human brain has become a key research tool for life scientists interested in understanding brain mechanisms of sensation, perception, cognition, and voluntary movement. MRI has becoming a cornerstone of such activities since it can provide structure and functional information at previously unobtainable brain locations without the need for invasive measures. Structural MRI can provide sub-millimeter resolution of the cellular and fiber tract regions of the brain. These capabilities now allow precise measurement of local brain volumes and visualization of the source and destination of major axon pathways. Functional MRI can rapidly measure local changes in blood dynamics in volumes as small as 1 cubic mm. Blood dynamics reflect changes in local neural activity, and its exploitation has become a key tool in exploring brain mechanisms of a variety of functions that constitute everyday experience. The NSF funded resource will allow Brown researchers and those from nearby institutions to develop new strategies and knowledge about how the human brain mediates complex behavior.

Projects currently planned for the 3 Tesla MRI system include research in systems and cognitive neuroscience and biomechanics. A major effort will be to enhance spatial and temporal resolution of structural and functional MR imaging, using special equipment of the new 3 Tesla MRI system. In particular, the infrastructure will facilitate investigating specialization of the myriad brain areas that process visual stimuli, not only across the brain, but also within each area. The new MRI system will allow non-invasive imaging of the input and output processing zones of cortical areas. Several investigators will interrogate the functional MRI signals obtained during single instances of perceptual experience or voluntary movement to predict the conscious experience of the observer or to predict the performed movement(s). While these 'mind-reading' efforts currently occur off-line, the team plans to implement them in real-time, and ultimately at high spatial resolution. The addition of the 3 Tesla MRI resource at Brown will boost ongoing educational and research activities and will stimulate novel interactions between students, faculty and researchers working across life, social, physical and applied science disciplines. The instrumentation will also be used in out-reach programs for under-represented minority high-school students participating in summer programs at Brown.

Understanding how the mind emerges from brain activity is arguably one of the greatest scientific quests of all time. Exciting new work in the domains of neuroscience and cognition is now beginning to bridge areas and methods to ask not just how a particular cognitive process works in isolation, but how it connects to and interacts with other cognitive processes, and not just how a process works at a computational/algorithmic level, but how it is implemented neurally.

The National Science Foundation has long supported fundamental work on neuroscience and cognition and has recognized the central importance of research in this area. With funding from the NSF Dr. Sheila Blumstein of Brown University and her collaborators will hold a workshop on "Mind and Brain: Strategies and Directions for Future Research" at the NSF July 18-19,2006. This workshop will bring together leaders from a broad range of fields at the intersection of neuroscience and cognition to discuss critical emerging issues at this juncture. The strategic goal of the workshop will be to develop new ideas and new directions for research that will help set the agenda for funding future research into cognition and neuroscience.

DESCRIPTION (provided by applicant): One residual clinical feature of chronic aphasics is a speech output disorder. Nonfluent aphasics have limited speech output and articulatory implementation deficits in which the sounds of language produced are often distorted and hard to understand by listeners. Some fluent aphasics select the wrong sounds comprising a word, making successive attempts impacting the aphasic's willingness to speak. The goal of the current project is to apply basic research to the development and preliminary testing of a novel rehabilitation program, 'Therapy Program to Improve Phonetic Production' (TIPP), designed to improve the quality of the acoustic- phonetic output and ultimately the speech intelligibility of aphasic participants. TIPP builds off of basic research on the functional architecture of spoken word production in normals showing that articulatory implementation is influenced by the lexical properties of the stimulus; fMRI results showing the neural system recruited in such processing; and recent findings indicating spared properties of the functional architecture of spoken word production in aphasia. Five nonfluent aphasics who have speech output deficits, and five fluent aphasics who produce phonemic paraphasias, will participate. It is hypothesized that exposure to and modeling speech under different speaking conditions can shape and strengthen the lexical and sound shape representations of words; training with minimal pair words (tart vs. dart) will enhance the distinction between similar sounding words leading to strengthening phonetic and phonological representations and improved quality of the phonetic output; this strengthening will generalize to untrained words that share voicing with the trained words and will also have a cascading effect on accessing lexical representations leading to increased naming ability and spoken output fluency. TIPP is a 10 week, 2-session per week, protocol which includes a two part training phase, one in which the participant is trained to produce a set of alveolar stop consonant minimal pair target words embedded in a carrier phrase spoken with normal intonation but at a slowed pace, and the other in which the participant is trained to intone the same set of target words embedded in the carrier phrase, also at a slowed pace. To assess the program, acoustic analyses will be conducted to determine whether there is a greater distinction between voiced and voiceless stop consonants post-training, whether there are changes in the number of tokens successfully produced, the number of sound errors made, and the number of false starts or stutters, and where there are changes in intelligibility. Generalization and maintenance 2 and 6 monthspost- therapy will be assessed on a set of untrained labial and velar word tokens and on other language variables including naming and speech fluency. The number of adults who have a residual aphasia after a neurological event is in the millions, and the effects on their lives as well as their families are devastating. This research holds promise of establishing a new protocol to improve the quality and intelligibility of the speech output of aphasic participants leading to a substantive improvement in their communication abilities.