UMBC Language, Literacy, and Culture Department

Permanent URI for this collectionhttp://hdl.handle.net/11603/42

Browse

Now showing 1 - 20 of 183

Is the FTX cryptocurrency case just an ole fashioned swindle?
(wbur, 2023-02-03) Tong, Scott; Froide, Amy
Financial fraud cases go back to the birth of capitalism. So is the FTX cryptocurrency case something new? Or should we have learned lessons on how to regulate these massive frauds?
The Long History of Financial Fraud
(Inside Higher Ed, 2023-09-21) Pasquerella, Lynn; Froide, Amy
Financial fraud may be in the news, but it’s also old news. In today’s Academic Minute, part of University of Maryland Baltimore County Week, Amy Froide examines past wrongdoings. Froide is professor of history and chair of the department of history at UMBC. A transcript of this podcast can be found here.
Single, Independent Women, in the 16th Century
(That Shakespeare Life, 2024-09-02) Cash, Cassidy; Froide, Amy
You may be surprised to learn that marriage in the 16th century was not required, nor a foregone conclusion, for all women of this time period. In addition to spinsters, who were older women that had never been married, there were widows that lost their husband, women who were divorced or separated from their husbands, and still some women who our guest this week calls “never married” women. A “never-married” woman chose never to get married at all, and provided for themselves financially. While Shakespeare doesn’t use the phrase “never married” he does talk about spinsters, widows, prostitutes, and even divorce, reflecting the society of his time period. While all versions of single women in Shakespeare’s lifetime operated outside of what we generally expect for the 16th century, the reality is that being single, and even women who were independently made, occupied a much larger section of society than you may have assumed. Here today to tell us about all the single ladies of the 16th century, and what life was like for a women who never married, is our guest, Amy Froide.
Blue Sky: Expert-in-the-Loop Representation Learning Framework for Audio Anti-Spoofing: Multimodal, Multilingual, Multi-speaker, Multi-attack (4M) Scenarios
(SIAM International Conference on Data Mining, 2025) Khanjani, Zahra; Janeja, Vandana; Mallinson, Christine; Purushotham, Sanjay
Audio spoofing has surged with the rise of generative artificial intelligence, posing a serious threat to online communication. Recent studies have shown promising avenues in detecting spoofed audio specifically those that use human expert knowledge in representation learning, but more work is needed to evaluate performance across various realistic scenarios that tend to pose challenges in spoofed audio detection. In this paper, we introduce a comprehensive framework for expert-in-the-loop representation learning for audio anti-spoofing that is robust enough to address four specific challenging scenarios. Multimodal, Multilingual, Multi-speaker, and Multi-attack (4M). Preliminary results demonstrate the framework’s potential effectiveness in audio anti-spoofing.
Disparate Pathways: Understanding Racial Disparities in Teaching
(Annenberg Institute, 2024-04) Blazar, David; Anthenelli, Max; Gao, Wenjing; Goings, Ramon; Gershenson, Seth
Mounting evidence supporting the advantages of a diverse teacher workforce prompts policymakers to scrutinize existing recruitment pathways. Following four cohorts of Maryland public high-school students over 12 years reveals several insights. Early barriers require timely interventions, aiding students of color in achieving educational milestones that are prerequisites for teacher candidacy (high school graduation, college enrollment). While alternative pathways that bypass traditional undergraduate teacher preparation may help, current approaches still show persistent racial disparities. Data simulations underscore the need for race-conscious policies specifically targeting or differentially benefiting students of color, as race-neutral strategies have minimal impact. Ultimately, multiple raceconscious policy solutions addressing various educational milestones must demonstrate significant effects—approximately 30% increases—to reshape the teacher workforce to align with student body demographics.
Chimerica: Disorienting Politics
(Positions, 2025-01-26) Phillips, Delores B.; Razzano, Kathalene; Yang, Fan; Tseng-Putterman, Mark; Breen, Marcus
Delores Phillips and Cultural Studies Association’s Globalization and Culture Working Group Co-Host Kathalene Razzano discuss Disorienting Politics: Chimerican Media and Transpacific Entanglements (University of Michigan Press, 2024) with author Fan Yang, along with writer and historian Mark Tseng-Putterman. This podcast is accompanied by a scholarly commentary by Marcus Breen.
The Educational and Labor Market Impacts of Maryland’s Grow-Your-Own Teacher Recruitment Program
(Urban Institute, 2024-07-12) Blazar, David; Gershenson, Seth; Goings, Ramon
Pathways into the CTE Teaching Profession: A Descriptive Analysis of Degrees, Licenses, and Race in Maryland
(Annenberg Institute, 2024-06) Blazar, David; Song, Danett; Goings, Ramon; Plasman, Jay; Gottfried, Michael
Despite substantial interest in Career and Technical Education (CTE) courses in U.S. high schools— and associated scholarship on this topic—very little is known about characteristics of CTE teachers who are a critical resource for program implementation and expansion. Using eight years of statewide data from Maryland, we document key facts about the CTE teacher workforce and pathways into the profession. First, a sizable share (17%) of CTE teachers enter the profession with a high school diploma or associate’s degree, aligned to state policy that allows Professional and Technical Educationcertified teachers to substitute years of professional experience for higher degrees. Relatedly, CTE teachers are substantially more likely than non-CTE teachers to enter the profession through an “alternative” path that bypasses traditional undergraduate teacher education (54% versus 30%). Finally, there is a larger share of Black teachers in CTE versus out of CTE (25% versus 16%), leading to greater opportunities for teacher-student race matching. We hypothesize that these patterns are related: decreased barriers to entry into the CTE teaching profession may support more Black individuals to become CTE teachers.
Queering 'Common Sense': Using Critical Discourse Analysis of LGBTQ Book Bans in Florida Public Schools to Unmask Heteronormativity
(University of Nebraska, 2025-02-18) Zito, Talia; Morse, Nicole
Historically, censorship of LGBTQ+ content in education has been rooted in socio-political efforts to marginalize non-heteronormative identities. However, even those who oppose censorship tend to frame their arguments around concepts like "age-appropriateness" and "protecting children." Our Critical Discourse Analysis examines formal challenges to And Tango Makes Three in one Florida school district in 2022 to show how these concepts are deployed on both sides of the debate. The book, depicting a same-sex penguin couple raising a chick, faced objections under Florida's HB 1557, a law designed to restrict discussions of sexual orientation and gender identity in schools. Ultimately, the picture book was removed from the shelves despite proponents advocating for its educational value. Our analysis examines how both opponents and proponents of LGBTQ+ children's literature rely on "common sense" discourse and arguments about "age appropriateness" to support their claims. We argue that such discursive moves, while seemingly neutral, perpetuate discriminatory ideologies and reinforce normative power structures. Given that both sides' appeals to common sense and age-appropriateness fail to challenge and often reinforce exclusionary beliefs, we argue that advocates should look beyond common sense and instead draw on queer theory as a resource in these struggles against censorship. Effective advocacy must confront how heteronormative assumptions are embedded in educational policies and practices. Queer theory offers a powerful framework to reimagine advocacy strategies, challenging the notion of political neutrality by exposing how what appears to be unbiased or "common sense" is often deeply embedded in normative assumptions and power dynamics. By reframing the discourse, advocates can challenge and dismantle the underlying heteronormative ideologies. Embracing queer theoretical insights can lead to a more inclusive and equitable environment, where LGBTQ+ literature is celebrated for its role in promoting diversity, challenging normative constructs, and reimagining educational environments and social worlds.
Art-Style is the Future: The Spider-Verse as a Model for Conceptualizing and Visualizing Graphic Narratives in Visual Media
(2024-01-01) Wildy, Caron; Saper, Craig; Pasandi, Haniyeh B.; Language, Literacy & Culture; Language Literacy and Culture
Spider-Man: Into the Spider-Verse was a watershed moment in American animation. A stylized animated adaptation of Marvel Comics’ Spider-Man comics that drew directly on the aesthetics and visual language of comics, Into the Spider-Verse signaled a stylistic shift toward stylized aesthetics and imagery that aligned narrative and visuals and diverged from photorealistic aesthetics by drawing from diverse visual languages and aesthetics. The influence of Into the Spider-Verse, and its 2023 sequel Across the Spider-Verse, on the visual style of American animation serves as context for an exploration of how the films’ synthesis of aesthetics, visual languages and narratives from multiple visual mediums can inspire creators to reframe how they conceive of graphic narratives, visualize narrative, and adapt graphic narratives across visual mediums. The intersection between aesthetics and narrative in the Spider-Verse films challenges an emphasis on realism in the aesthetics of visual narratives in mainstream American visual narrative media that flatten visual signatures. The Spider-Verse films contest this emphasis on a mainstream stage and arrive in a landscape where audiences have greater access to a diverse visual storytelling aesthetics modes and mediums. This exploration uses analysis aesthetics and the synthesis of visual languages in the Spider-Verse films to demonstrate how the films offer a model for conceptualizing and visualizing narratives and explore how the convergence of aesthetics and visual language in the depiction of characters, environments, and scenes in the Spider-Verse films inspire creators to reframe the thinking that informs the conceptualizing, writing, and visualizing stories for scripted media.
Making the Future with the Nonhuman:Shenzhen, the Greater Bay, and “Made in China Intelligently”
(OpenEdition, 2023-12-01) Yang, Fan
This essay examines two interconnected human-made nonhuman entities stemming from Shenzhen, China’s first special economic zone, that have become dominant figures in mapping the city’s – and by extension, China’s – future: the robot and the drone. I bring an interdisciplinary, cultural studies approach to the multiple meaning-making practices that engage with these two objects; both participate in enacting the vision for the Guangdong-Hong Kong-Macao Greater Bay Area as an extension of the success of Shenzhen. These practices simultaneously normalise aspirations for a future fuelled by the power of nonhuman technological agents while offering glimpses into the uneven power relations between different humans that underpin such future making. At the same time, they also point to the emergent possibilities of meaning-making that conjoin the human and the nonhuman.
Listening for Expert Identified Linguistic Features: Assessment of Audio Deepfake Discernment among Undergraduate Students
(2024-11-21) Bhalli, Noshaba Nasir; Naqvi, Nehal; Evered, Chloe; Mallinson, Christine; Janeja, Vandana
This paper evaluates the impact of training undergraduate students to improve their audio deepfake discernment ability by listening for expert-defined linguistic features. Such features have been shown to improve performance of AI algorithms; here, we ascertain whether this improvement in AI algorithms also translates to improvement of the perceptual awareness and discernment ability of listeners. With humans as the weakest link in any cybersecurity solution, we propose that listener discernment is a key factor for improving trustworthiness of audio content. In this study we determine whether training that familiarizes listeners with English language variation can improve their abilities to discern audio deepfakes. We focus on undergraduate students, as this demographic group is constantly exposed to social media and the potential for deception and misinformation online. To the best of our knowledge, our work is the first study to uniquely address English audio deepfake discernment through such techniques. Our research goes beyond informational training by introducing targeted linguistic cues to listeners as a deepfake discernment mechanism, via a training module. In a pre-/post- experimental design, we evaluated the impact of the training across 264 students as a representative cross section of all students at the University of Maryland, Baltimore County, and across experimental and control sections. Findings show that the experimental group showed a statistically significant decrease in their unsurety when evaluating audio clips and an improvement in their ability to correctly identify clips they were initially unsure about. While results are promising, future research will explore more robust and comprehensive trainings for greater impact.
Commentary: Bringing people and technology together to combat the threat of deepfakes
(Maryland Matters, 2024-03-25) Mallinson, Christine; Janeja, Vandana
UMBC team is creating and testing short training sessions to help listeners spot common 'tells' of real and fake speech.
ALDAS: Audio-Linguistic Data Augmentation for Spoofed Audio Detection
(2024-10-21) Khanjani, Zahra; Mallinson, Christine; Foulds, James; Janeja, Vandana
Spoofed audio, i.e. audio that is manipulated or AI-generated deepfake audio, is difficult to detect when only using acoustic features. Some recent innovative work involving AI-spoofed audio detection models augmented with phonetic and phonological features of spoken English, manually annotated by experts, led to improved model performance. While this augmented model produced substantial improvements over traditional acoustic features based models, a scalability challenge motivates inquiry into auto labeling of features. In this paper we propose an AI framework, Audio-Linguistic Data Augmentation for Spoofed audio detection (ALDAS), for auto labeling linguistic features. ALDAS is trained on linguistic features selected and extracted by sociolinguistics experts; these auto labeled features are used to evaluate the quality of ALDAS predictions. Findings indicate that while the detection enhancement is not as substantial as when involving the pure ground truth linguistic features, there is improvement in performance while achieving auto labeling. Labels generated by ALDAS are also validated by the sociolinguistics experts.
Toward Transdisciplinary Approaches to Audio Deepfake Discernment
(2024-11-08) Janeja, Vandana; Mallinson, Christine
This perspective calls for scholars across disciplines to address the challenge of audio deepfake detection and discernment through an interdisciplinary lens across Artificial Intelligence methods and linguistics. With an avalanche of tools for the generation of realistic-sounding fake speech on one side, the detection of deepfakes is lagging on the other. Particularly hindering audio deepfake detection is the fact that current AI models lack a full understanding of the inherent variability of language and the complexities and uniqueness of human speech. We see the promising potential in recent transdisciplinary work that incorporates linguistic knowledge into AI approaches to provide pathways for expert-in-the-loop and to move beyond expert agnostic AI-based methods for more robust and comprehensive deepfake detection.
Sustainability of a dual language program during and beyond COVID-19 challenges
(Taylor & Francis, 2024-10-28) Mata-McMahon, Jennifer; Williams, Sabrina; Daramola, Adebola; Kruse, Lance; Hossain, Shahin
This study evaluates the Dual Language Program (DLP) implemented at a Title I public school in Baltimore City during the 2019–2020 and 2020–2021 school years. Building on previous research, the DLP's implementation, sustainability, and effects on students’ learning outcomes were examined. Utilizing a mixed-methods approach, the study’s second phase included participants from the school’s mainstream, English-only program (MP), enriching the overall understanding of the school community’s perception of the program. Data were collected through surveys, classroom observations, and standardized assessments—the DIBELS and the iReady Diagnostic assessments. Findings showed challenges with curriculum standardization and prevalent misconceptions about bilingualism. Nevertheless, the program's successes were evident in DLP students’ enhanced engagement, parental involvement, and community support. Despite the challenges of the COVID-19 pandemic, the DLP demonstrated its capacity for scaleability and sustainability. During the 2020–2021 school year, regardless of virtual learning, DLP students not only maintained but, in the case of Cohort 1, showed greater growth in reading skills compared to MP students, with Cohort 3 DLP students improving to reach statistically similar performance levels to their MP peers. Findings indicate that the DLP has the potential to serve as a sustainable educational program, fostering both dual language proficiency and academic outcomes.
ALDAS: Audio-Linguistic Data Augmentation for Spoofed Audio Detection
(2024-10-21) Khanjani, Zahra; Mallinson, Christine; Foulds, James; Janeja, Vandana
Spoofed audio, i.e. audio that is manipulated or AI-generated deepfake audio, is difficult to detect when only using acoustic features. Some recent innovative work involving AI-spoofed audio detection models augmented with phonetic and phonological features of spoken English, manually annotated by experts, led to improved model performance. While this augmented model produced substantial improvements over traditional acoustic features based models, a scalability challenge motivates inquiry into auto labeling of features. In this paper we propose an AI framework, Audio-Linguistic Data Augmentation for Spoofed audio detection (ALDAS), for auto labeling linguistic features. ALDAS is trained on linguistic features selected and extracted by sociolinguistics experts; these auto labeled features are used to evaluate the quality of ALDAS predictions. Findings indicate that while the detection enhancement is not as substantial as when involving the pure ground truth linguistic features, there is improvement in performance while achieving auto labeling. Labels generated by ALDAS are also validated by the sociolinguistics experts.
Commentary: Bringing people and technology together to combat the threat of deepfakes
(Maryland Matters, 2024-03-25) Mallinson, Christine; Janeja, Vandana
UMBC team is creating and testing short training sessions to help listeners spot common 'tells' of real and fake speech.
The Influence of the Sherman STEM Teacher Scholars Program on Persistence in Science, Technology, Engineering, and Mathematics: A Mixed-Methods Study
(MDPI, 2024-10-02) Goings, Ramon; Boyd, Brittany
This sequential explanatory mixed-methods study investigated the differences in persistence between students from the Sherman STEM Teacher Scholars Program (STEP), a STEM teacher scholarship and career preparation program, and STEM majors not in the program. Quantitative results indicated that STEP participants had higher levels of academic integration, women scored higher on persistence factors than men, and White students had a higher degree commitment than students of color. Qualitative findings indicated that STEP provided a family atmosphere and connected their coursework to their career aspirations. Women of color felt stereotyped by White classmates in STEM courses, which impacted their degree of commitment, and students of color in STEP relied on the program as a counterspace to racially insensitive STEM classrooms.
Investigating Causal Cues: Strengthening Spoofed Audio Detection with Human-Discernible Linguistic Features
(2024-09-09) Khanjani, Zahra; Ale, Tolulope; Wang, Jianwu; Davis, Lavon; Mallinson, Christine; Janeja, Vandana
Several types of spoofed audio, such as mimicry, replay attacks, and deepfakes, have created societal challenges to information integrity. Recently, researchers have worked with sociolinguistics experts to label spoofed audio samples with Expert Defined Linguistic Features (EDLFs) that can be discerned by the human ear: pitch, pause, word-initial and word-final release bursts of consonant stops, audible intake or outtake of breath, and overall audio quality. It is established that there is an improvement in several deepfake detection algorithms when they augmented the traditional and common features of audio data with these EDLFs. In this paper, using a hybrid dataset comprised of multiple types of spoofed audio augmented with sociolinguistic annotations, we investigate causal discovery and inferences between the discernible linguistic features and the label in the audio clips, comparing the findings of the causal models with the expert ground truth validation labeling process. Our findings suggest that the causal models indicate the utility of incorporating linguistic features to help discern spoofed audio, as well as the overall need and opportunity to incorporate human knowledge into models and techniques for strengthening AI models. The causal discovery and inference can be used as a foundation of training humans to discern spoofed audio as well as automating EDLFs labeling for the purpose of performance improvement of the common AI-based spoofed audio detectors.

Browse

Recent Submissions