50107 research outputs found
Sort by
Informing Home: Natural Materials' Potential to Reinvigorate Residential Architecture
No full textIn the United States, residential housing is largely understood as a commodity that is dependent on industrial processes. This leads to houses that lack character, have high environmental impact, and often include synthetic materials and toxins that are detrimental for the occupants. Natural and bio-based materials offer alternative construction techniques that have the potential to improve occupant health and wellbeing, reduce environmental impacts, and reintroduce craft and local connection to construction. This thesis explores the impact of natural material-based systems on spatial organization, formal expression, and construction. These ideas are tested through the design of a modest, single family home in rural New England to be constructed by self-builders.Master of Architecture (MArch
Local Changes in Sleep Following Declarative Learning
No full textTraditionally, sleep has been viewed as a uniform, global state. This idea is contrasted by local sleep, which is characterized by transient periods of brain activity within specific brain regions, or even at the level of individual neurons during sleep. Previous studies have shown increased slow wave activity (SWA) in the parietal cortex during sleep following motor learning, indicating a local increase in sleep depth in response to motor learning. Here we ask a simple question: does this local use-dependent sleep response also occur during non-motor/declarative memory tasks?
During one wake and one overnight session, participants completed a word pair learning task, consisting of an encoding phase, followed by immediate and delayed recall sessions. As a control, participants completed an overnight non-learning task, where they passively viewed sham word pairs. There was an increase in SWA following the declarative learning task compared to the control task, particularly in the fronto-central and left temporal regions, suggesting that localized SWA increases occur following non-motor learning. These localized changes in SWA were significantly correlated with memory recall performance when controlling for subjective sleepiness. These findings provide novel evidence for localized sleep responses following declarative memory tasks and suggest that SWA may be region-specific to task demands, though further investigation is needed to understand the relationship between SWA and memory consolidation in non-motor tasks.Master of Science (M.S.)2025-11-1
THE PLURIVERSAL PREFIGURATION OF KURDISH ACTIVISTS IN HAMBURG AND LONDON & PRACTICING SOCIAL ECOLOGY: FROM BOOKCHIN TO ROJAVA AND BEYOND
No full textWhat is at stake in radically democratic politics in the 21st century? How do we account for the continuities and discontinuities in the conception and practice of prefigurative politics and radical democracy in grassroots social movements? How do activists from different corners of the world interpret “real democracy”? A decade ago, a period of intense, global grassroots mobilization called the Movements of the Squares brought prefigurative politics (the linkage of present means and future ends) to the center of public and scholarly attention, foregrounding the practice of directly democratic deliberation in characteristic forms such as encampment “general assemblies.” Today, a new generation of prefigurative activism, anchored in the notion of “social ecology,”, including the municipalist movement of Barcelona and democratic confederalism in the Rojava Revolution of North and East Syria, is embodying post-capitalism in ways that engage local knowledge, foster place-based identities, and speak to a multiplicity of democratic ideals. The first section, titled “The Pluriversal Prefiguration of Kurdish Activists in Hamburg and London,” mobilizes ethnographic fieldwork among Kurdish diasporic activists and their transnational allies in Germany and the UK between 2016 and 2020 to explore what anthropologist Arturo Escobar calls “pluriversal prefiguration,” the embodiment of post-capitalism in ways that engage local knowledge, foster place-based identities, and speak to a multiplicity of democratic ideals. In this section, I argue that Kurdish liberation activism challenges the primacy of deliberation and decision-making in our understanding of prefigurative politics by centering, rather, horizontal interpersonal relationships, popular consciousness-raising, and political community. To do so, I focus on the emic notions of hevalti (friendship), tekmil (self-criticism), and perxwede (education). The second section, “Practicing Social Ecology,” weaves ethnographic research and personal experience to show how activists in democratic ecology movements are trying to harness society's potential to change the trajectory of the climate crisis. From the revolution in Rojava to Barcelona's municipalist movement and beyond, activists are developing assemblies, confederations, study groups, and permaculture projects. Also looking to history, this section maps out how social ecologists, such as Murray Bookchin, have led inspirational struggles around climate and energy, agriculture and biotechnology, globalization and economic inequality.UMass Amherst Graduate SchoolDoctor of Philosophy (Ph.D.)2030-05-1
EXPERIMENTAL INVESTIGATION OF INTERMOLECULAR INTERACTIONS AND VIBRATIONAL COUPLING EFFECTS IN THERMAL AND ELECTRICAL TRANSPORT OF POLYMER-BASED MATERIALS
No full textPolymers with high thermal conductivity are highly sought after for heat dissipation applications due to their unique properties, yet common polymers are thermal insulators. While incorporating thermally conductive fillers is a common strategy for enhancing thermal conductivity of the polymer-based materials, the thermal conductivity of polymer composite is far from the optimized simulation value due to the existence of high interfacial thermal resistance and low thermal conductivity of polymer matrices. To address the gap, this thesis designs three systematic experiments to investigate the thermal and electrical transport on two pivotal factors: interfacial thermal transport in bulk polymer composites and interchain thermal transport in polymer matrix.
First, two polymer composite systems (polyethylene/graphite composites and polyaniline/graphite composites) were designed and synthesized to experimentally investigate the fundamental role of noncovalent intermolecular interactions in governing thermal transport in polymers. The thermal conductivity and electrical conductivity of the synthesize composites with weak van der Waals and/or strong π-π stacking interactions were systematically investigated. We established a systematic framework to experimentally probe interfacial thermal resistance in bulk polymer composites. This parameter is critical yet poorly understood. Compared to van der Waals forces, stronger noncovalent interactions, such as π-π stacking, more effectively reduce interfacial thermal resistance from ~〖6×10〗^(-8) m^2 K W^(-1) (polyethylene/graphite composites) to ~〖1×10〗^(-8) m^2 K W^(-1) (polyaniline/graphite composites). These interactions enhance the effective thermal transport of polymer composites. The electron contributions to thermal conductivity are both negligible in polyethylene/graphite composites (~3.20×10^(-15) W m^(-1) K^(-1)) and polyaniline/graphite composites 2.69×10^(-12) W m^(-1) K^(-1)).
Second, two systematic polymer composite systems (polyvinyl alcohol/“perfect” filler composites and polyvinyl alcohol/defective filler composites) were designed and synthesized to experimentally investigate the fundamental role of vibrational coupling between polymers and fillers in governing thermal transport properties in polymer composites, including interfacial thermal resistance between polymers and fillers. Defective fillers with tuned vibrational modes were synthesized. Although defects increase surface roughness and reduce the intrinsic thermal conductivity of the filler from ~292.55±25.72 W m^(-1) K^(-1) to ~66.29±4.64 W m^(-1) K^(-1), polymer composites with defective fillers exhibited higher effective thermal conductivity (~1.38±0.22 W m^(-1) K^(-1)) than those with pristine (“perfect”) fillers (~0.86 ±0.21 W m^(-1) K^(-1)). This improvement is primarily due to the decreased interfacial thermal resistance (〖~2×10〗^(-7) m^2 K W^(-1)) between defective fillers and polymers, which is lower than the high thermal resistance (〖~1×10〗^(-6) m^2 K W^(-1)) of “perfect” filler polymer composites, enhanced by stronger vibrational coupling between the polymer and the defective filler.
Third, the focus shifts to interchain thermal transport within the polymer matrix because thermal transport within the polymer matrix is critical. Two systematic polymer systems (undoped and doped poly(3-hexylthiophene) thin film) were designed and synthesized to experimentally investigate the fundamental role of noncovalent intermolecular interactions (e.g., π-π stacking interactions) and charge carriers in governing thermal transport between polymers chains. Oriented face-on poly(3-hexylthiophene) thin films were prepared, and the thermal conductivity in the cross-plane direction—influenced by thermal transport across chain through π-π stacking interactions—was systematically investigated. The poly(3-hexylthiophene) thin film doped at 1.3 V showed an increased cross-plane thermal conductivity of 0.26±0.02 W m^(-1) K^(-1) compared to the undoped thin film (0.19±0.02 W m^(-1) K^(-1)). The enhancement in cross-plane thermal conductivity could be primarily due to the reduced π-π stacking distance from 3.76 Å for undoped poly(3-hexylthiophene) to 3.56 Å for poly(3-hexylthiophene) doped at 1.3 V. This reduced distance confirmed by the X-ray scattering indicated a stronger π-π stacking interactions between polymer chains in the doped samples. The poly(3-hexylthiophene) thin film doped at 1.3 V showed an increment in cross-plane electrical conductivity of 6.68×10^(-8) S cm^(-1) compared to the undoped thin film (2.87×10^(-15)±1.94×10^(-15) S cm^(-1)). But the electron contribution to poly(3-hexylthiophene) thin film doped at 1.3 V is negligible (4.86×10^(-11) W m^(-1) K^(-1)).
Overall, these results from systematic experimental design reveal the role of intermolecular interactions and vibrational coupling effects on thermal and electrical transport in polymer-based materials, highlighting the rational design for enhancing thermal conductivity in polymer-based materials.Doctor of Philosophy (Ph.D.)2030-05-1
NOROVIRUS INTERACTIONS WITH BACTERIA AND THEIR UTILIZATION FOR CONCENTRATION PRIOR TO VIRAL DETECTION
No full textHuman norovirus is one of the leading causes of foodborne illnesses, causing various symptoms such as nausea, vomiting and stomach cramps. To prevent norovirus outbreaks, researchers have been working to develop effective norovirus detection methods. However, there are many limitations associated with current detection techniques. Since noroviruses are often present in low concentration in food and environmental samples, a concentration step is essential prior to detection. One promising strategy is to exploit the natural interaction between viruses and bacterial surfaces.
This study aimed to evaluate whether Lactobacillus species can be used to capture viruses as a method for concentration, to quantify their virus capture efficiency, and to examine how bacterial growth conditions affect capture performance. Two species, Lactobacillus plantarum and Lactobacillus gasseri, were tested. The four viruses used in this study were murine norovirus, Tulane virus, GI.1 human norovirus, and GII.2 human norovirus. Following incubation with the bacterial species, capture efficiency was measured.
The results showed that Lactobacillus species captured murine norovirus more efficiently than Tulane virus. Among the human noroviruses, GII.2 was captured more effectively than GI.1. Additionally, nutrient-limited media, particularly half-strength MRS, enhanced virus capture. Across all experiments, L. plantarum consistently demonstrated greater virus capture efficiency compared to L. gasseri.Master of Science (M.S.)2026-05-1
Beyond the Hopfield Memory Theory: Dynamic Energy Landscapes and Traveling Waves in RNNs
No full textRecurrent Neural Networks (RNNs) are central to artificial intelligence, excelling in sequence processing tasks across domains—from natural language processing to protein folding in biology. However, fundamental questions about how RNNs store and process information over time remain unanswered, limiting our ability to design and build capable AI models. Current theories, such as the Hopfield memory theory, primarily focus on static memory storage and associative retrieval, lacking explanations for the dynamic and adaptive memory processes in real-world applications. In this thesis, I formulate two theoretical frameworks addressing the challenge: the Dynamic Energy Theory, elucidating the RNN long-term memory processes through synaptic interactions over time, and the Wave Theory, describing the dynamic storage of inputs as transient working memories in neural activity. By building mathematical models from these theories, studying their properties, capacities, and limitations, I derive new RNNs with improved capabilities.
The Dynamic Energy Theory (DET) generalizes Hopfield memory by allowing the energy function to evolve over time, representing sequences as dynamic trajectories on the network's changing energy landscape. Using the DET, I develop a class of continuous-time RNNs with slow-fast timescale dynamics—where some neurons update rapidly while others change slowly—and analyze their escape times, the durations spent in each memory state, and memory capacity. The escape time analysis shows the conditions necessary for state transitions and a phase transition from static to dynamic memory when inter-memory interaction strength increases. The memory capacity shows increase with the strength of inter-memory interactions, enabling deriving networks with long-sequence storage capacities that exponentially outperform the linear capacity of current sequence networks. Next, I show how local biologically plausible learning rules are derived from the energy function adapting existing synaptic memories based on the input provided to the network. These new networks could potentially transform tasks requiring adaptive long-term memory retention.
The Wave Theory conceptualizes the binding of input and task-relevant variables in RNNs as propagating neural activity waves. Building upon the DET, it elucidates how synaptic interactions support wave propagation in the neural activity through local interactions. I demonstrate that practical sequence models like the Autoregressive Moving Average models, Elman RNNs, State Space Models (SSMs), and transformers can be derived from the wave theory, suggesting that it serves as a canonical framework for understanding existing sequence processing models. Using the theory, I show hidden traveling waves in Elman RNNs that store memories and mitigate the vanishing gradient problem. Building on the wave theory, I introduce a new class of adaptive and unitary SSMs with rapid and adaptive synaptic strength changes enabling high accuracy with relatively low number of parameters.
Together, the two theories aim to fill a critical gap in understanding how RNNs store and process information over time. With the insights from the theory, we are able to design parameter efficient and performant models for sequential data paving the way for the next generation of AI architectures.Doctor of Philosophy (Ph.D.
Strategies used to Increase Implementation of Family-School Partnerships by School Mental Health Practitioners
No full textSupporting the mental health of youth and fostering caregiver collaboration through family-school partnerships (FSP) are critical functions of schools. Both school-based mental health services and FSP have demonstrated a myriad of positive outcomes for students. Including FSP within school-based mental health services often produces enhanced and generalizable outcomes. While research has demonstrated that the presence of school-based mental health services and family-school partnership (FSP) practices have increased in recent years, little is known regarding the discrete practices, referred to as implementation strategies, which best predict school mental health practitioners’ (SMHP) ability to effectively implement FSP. Using multiple linear regression, this project used SMHPs’ perspectives on their use of an established compilation of school-based implementation strategies (i.e., School Implementation Strategies, Translating ERIC Resources (SISTER) project; Cook et al., 2019; Lyon et al., 2020), as well as their perspective on the importance of each implementation strategy to predict their self-reported use of FSP practices within a multi-tiered system of support (MTSS). Results demonstrated that greater rates of SMHP self-reported use of implementation strategies were highly predictive of SMHP use of FSP practices. The results further indicated that greater rates of self-perceived importance of the implementation strategies moderately predicted SMHPs use of FSP practices. The survey developed within this project represents a novel method for measuring the applied use of implementation strategies in schools. This project provides implications related to increasing use of effective FSP in schools by SMHPs. Further research on this topic is required to continue bridging the research to practice gap in the area of effective implementation of FSP by SMHPs.Mary Margaret Whittaker-Webster Memorial AwardDoctor of Philosophy (Ph.D.
BETWEEN REST AND RESISTANCE: EXPLORING THE MENTAL HEALTH IMPACT OF UNDOCUMENTED ACTIVISM
No full textThe 11 million undocumented immigrants living in the U.S. navigate constant uncertainty under policies that limit their rights, exclude them from social protections, and leave them vulnerable to sudden legal and economic shifts. These conditions persist despite global drivers of migration and many families’ deep-rooted histories in U.S. communities. Activism can be a source of purpose and solidarity, but it also carries a significant mental health cost, contributing to burnout, exhaustion, depression, anxiety, and other comorbidities. Informed by three years of field work at an immigrant rights nonprofit, 17 in-depth life history interviews, and a digital storytelling workshop with seven activists, this research critically examines how undocumented activists experience and conceptualize burnout, revealing how movement structures, storytelling practices, and internalized expectations of self-sacrifice contribute to an unsustainable advocacy culture. Findings indicate that movement sustainability is undermined by the expectation that activists prioritize collective progress over personal well-being, leading to cycles of activist burnout and withdrawal. Activists described the pressure to publicly share personal trauma for advocacy gains, put their personal growth on hold for movement-related work that is often unpaid, and to prioritize activism activities over personal health and care. Participants also identified radical joy as a sustaining aspect of movement culture, and one that could provide a necessary shift in movement structures towards an approach that recognizes personal fulfillment, family life, and long-term stability as integral to movement sustainability. These insights highlight the need for advocacy models that not only acknowledge the emotional and physical toll of activism but actively incorporate well-being and care as core principles of organizing. Findings contribute to research on activist mental health, social movement sustainability, and narrative-based advocacy, with implications for research, policy, and practice.This dissertation was generously supported by the National Geographic Society, Define American, the Center for Research on Families, and the UMass Graduate School Office of Professional Development.Doctor of Philosophy (Ph.D.
Videographic Ecocriticism: Environment and Affect in Iberian Moving Images
No full textAs a deeply affective and ecological medium, moving images—across various forms—can create material-semiotic worlds that not only replicate reality but also shape our emotional responses to it. I am particularly interested in how different aesthetic and narrative devices in fictional moving images can heighten our awareness of our connection to the surroundings and evoke the wide and contradictory range of emotional responses that emerge from the climate crisis. My work is situated within media studies and draws from ecocriticism and affect theory to examine how these emotions are elicited through particular narrative strategies—such as the contemplative rhythms of slow cinema, the haptic qualities of the screen that invite an embodied connection, or the fast-paced editing that opens space for disjunctions where new connections emerge.
Moreover, my work is twofold: I study how moving images provoke environmental affects and I participate in practice-based research by developing a methodology to engage with these images through videographic ecocriticism. I argue that the videoessay, as a form of critical and creative engagement, enables deeper interaction with audiovisual material—increasing the emotional impact and the interconnection between human and nonhuman. In exploring the environmental potential of the medium, I contend that the capacity of videographic criticism of increasing some of the narrative devices mentioned before can enhance some of the emotions that emerge from the relationship with the surroundings and increase the deep interconnections between humans and the nonhuman. To do this, I explore different media—from contemplative cinema to music videos—, a range of phenomena—from the all-encompassing impact of wildfires to the more everyday occurrences of weather—, concrete artifacts —such as the motorcycle—, different speeds—from the sluggishness of a hot summer day to the pleasure of going fast on the road—, and different levels of comfort and tones—from desolation to hedonism—to expand the emotional archive of how we experience and perceive the current crisis. These affective modes, when taken up by art, reorient the viewer’s relation to ecological crisis—not as a call to immediate action, but as an encounter with the uneven, ambivalent, and emotionally complex nature of the Anthropocene.Graduate School Summer 2024 Dissertation Completion Fellowship
Doctoral Scholarship in Catalan Studies from the Institut Ramon Llull and the North American Catalan SocietyDoctor of Philosophy (Ph.D.
THE ROLE OF SOIL MICROBIAL FUNCTIONAL LEGACIES IN DETERMINING ECOSYSTEM CARBON AND NUTRIENT CYCLING DURING AND AFTER FOREST TRANSITION
No full textEastern hemlock (Tsuga canadensis), a long-lived foundation species in eastern North American forests, is experiencing widespread decline due to infestation by the hemlock woolly adelgid (Adelges tsugae). Hemlock loss allows faster growing deciduous species to take over, causing a transition in forest composition that can affect carbon (C) and nutrient cycling and soil microbial function. This dissertation investigates how soil microbial community function, measured through C and nutrient cycling, responds to the speed of hemlock decline across gradients and the extent to which soil properties predict forest vulnerability. Across a latitudinal chronosequence representing gradual northward HWA infestation (spanning the years 1987 – 2010), I found that gradual overstory transition from hemlock to hardwoods did not significantly alter decomposition rates. This suggests that microbial communities can retain a hemlock functional legacy for over three decades post-invasion. In an experimental manipulation simulating rapid hemlock loss, which is more analogous to the sudden mortality observed in the southern part of the species’ range, and natural hemlock decline, local decomposition patterns demonstrated a home-field advantage functional legacy, with accelerated decay when litter type matched the local overstory. However, this hemlock functional legacy was not detected at 17 years following a rapid hemlock loss demonstrating that while microbial resilience can buffer ecosystems under slow transitions, rapid canopy shifts may trigger more immediate and directional changes in decomposition dynamics and nutrient cycling. Finally, I evaluated whether soil physicochemical properties could predict forest condition by comparing two hemlock stands with the same year of infection but contrasting health. The more impacted stand had higher foliar nutrient concentrations but did not show corresponding greater soil nutrient availability, suggesting that physiological stress, mineralogy, or symbiotic associations may contribute more to forest trajectories than bulk soil chemistry alone. Together, these findings underscore the importance of legacy effects, disturbance speed, and belowground complexity in mediating ecosystem responses to biotic stress. This work advances our understanding of how invasive species reshape soil–plant interactions and highlights the need to consider both biotic and abiotic history when predicting the future of forest ecosystems.This research was funded by USDA Forest Service Evaluation Monitoring grant 22-DG-11094200-204. This research was supported by a Lotta M. Crabtree Fellowship to CAP and the Center for Agriculture, Food, and the Environment at the University of Massachusetts Amherst.Doctor of Philosophy (Ph.D.