About

Jennifer Larson is an Associate Professor in the Department of English and Comparative Literature at the University of North Carolina at Chapel Hill. She has a strong focus on online education and digital learning, with a particular interest in effective online teaching methods. Her work includes creating instructional videos and resource guides to assist faculty in transitioning to online formats. Larson's extensive experience in remote teaching has motivated her to serve as a peer mentor, helping faculty adapt to the challenges of online education, especially in response to the COVID-19 pandemic. She is passionate about online learning as a research area and sees opportunities for innovation in digital education.

Research topics

• Computer Science
• Data Mining
• Sociology
• Data science
• Epistemology
• Social Science
• History
• Philosophy
• Geography
• Geology
• Archaeology
• Paleontology

Selected publications

• Gods and Monsters

  Oxford University Press eBooks · 2024-11-19

  book-chapter1st authorCorresponding

  Abstract A common component in definitions of the monster is its challenge to categories. Yet not every category violation yields a monster. Deviation from the essential norm of a living kind (or perfect adherence to an essential norm) is what allows animals to function symbolically. Like gods and ‘perfect’ animals, monsters perform symbolic work. Human cognition helps to shape what we find monstrous: innate attentional biases sharpen our awareness of missing or misshapen body parts, while large predators of all kinds are impossible to ignore. Pascal Boyer’s theory of minimally counterintuitive concepts explains why monster body plans follow predictable patterns, and why their deviations tend to be limited in number and degree, simple enough to transmit accurately. Finally, the work of Dan Sperber on the ‘epidemiology of representations’ can help to illuminate the transmission of monster representations from the ancient Near East to the Greek world.

  PublisherDOI

• Indexes

  2023

  • Computer Science
  • Computer Science

  Her research foci include religion and magic in the Greek and Roman worlds, digital humanities, maritime mobility, archaeometallurgy, and anthropological and comparative approaches to the ancient Mediterranean.Her current project explores the pragmatic realization of the promises of safe sailing associated with initiation into the mysteries of the Great Gods of Samothrace, positioning epigraphic, legendary, and literary data in social networks and ancient geospaces.Megan Daniels is assistant professor of ancient Greek material culture at the University of British Columbia.Her interests revolve around several areas, including data science and social sciences approaches to ancient Mediterranean religions, the study of migration and mobility in antiquity, and the shared ideologies of divine kingship between Greece and the Near East.She is currently working on a monograph on this latter topic, and has edited a volume of papers on interdisciplinary approaches to ancient migration, which came out in 2022.

  DOI

• Are religious rituals always causally opaque?

  Religion Brain & Behavior · 2023-06-22 · 1 citations

  article1st authorCorresponding

  Click to increase image sizeClick to decrease image size Disclosure statementThe author is a member of the board of Seshat: Global History Databank.

  PublisherDOI

• Harnessing the Gods:

  2023-05-01

  book-chapter1st authorCorresponding
  PublisherDOI

• Synthetic Replacements for Human Survey Data? The Perils of Large Language Models

  2023-05-04 · 36 citations

  preprintOpen accessSenior author

  Large Language Models (LLMs) offer new research possibilities for social scientists, but their potential as "synthetic data" is still largely unknown. In this paper, we investigate how accurately the popular closed-source LLM ChatGPT can recover public opinion, prompting the LLM to adopt different "personas" and then provide feeling thermometer scores for 11 sociopolitical groups. The average scores generated by ChatGPT correspond closely to the averages in our baseline survey, the 2016–2020 American National Election Study. Nevertheless, sampling by ChatGPT is not reliable for statistical inference: there is less variation in responses than in the real surveys, and regression coefficients often differ significantly from equivalent estimates obtained using ANES data. We also document how the distribution of synthetic responses varies with minor changes in prompt wording, and we show how the same prompt yields significantly different results over a three-month period. Altogether, our findings raise serious concerns about the quality, reliability, and reproducibility of synthetic survey data generated by LLMs.

  PublisherOA PDFDOI

• “All Spooked Out”: Topdog/Underdog ’s Ghosts

  2023-01-01

  other1st authorCorresponding
  PublisherDOI

• Cognitive Science of Religion and the Work of Henk Versnel

  2023-05-15

  book-chapter1st authorCorresponding
  PublisherDOI

• Moralizing Supernatural Punishment and Reward

  Journal of Cognitive Historiography · 2023-10-19

  article1st authorCorresponding

  In this article we respond to three critiques of our 2019 article ‘Complex Societies Precede Moralizing Gods throughout World History.’ We clarify that our research does not, as our critics suppose, support the claim that moralizing gods played a decisive role in the development of complex societies. Indeed our goal was to test this claim and we found it wanting. Our methods ‘reduce’ neither religion or social complexity in the ways claimed, while our tentative conclusions about the relationship between frequent, routinized ritual and social cohesion are supported by much research beyond the paper under discussion. In the Roman Empire, many forms of collective ritual contributed to the propagation of Romanitas. We have never claimed that this depended on absolute uniformity of belief. Other misconceptions about our supposedly ‘inattentive’ qualitative analysis result from misreadings of information in our open-access database, which functions as an evolving set of information relevant to specific research questions rather than a general encyclopedia. Despite these disagreements, we continue to maintain that neither qualitative historical methods nor quantitative analytic approaches alone can produce satisfying answers to causal questions about world history. The best approach, we argue, is to integrate the insights from humanities with ‘Big Data’ analyses from social science, and we welcome continued engagement and collaboration across traditional disciplinary boundaries.

  PublisherDOI

• Introducing a special issue on the role of moralizing gods in the evolution of socio-political complexity

  Religion Brain & Behavior · 2023-04-03

  articleOpen access
  PublisherOA PDFDOI

• Causal Opacity or Causal Translucence?

  Journal of Cognitive Historiography · 2023-10-19

  article1st authorCorresponding

  According to longstanding interpretations in the social and cognitive sciences, rituals are said to be characterized by arbitrary action and the lack of a causal connection between action and desired outcome. The observer who assigns a physical-causal connection has taken the instrumental stance, while one who accepts a group convention is said to take the ritual stance. I argue that in religious rituals at least, including those with magical elements, the gap is bridged and causal intuitions are present, if limited. For example, we rely on a mental heuristic called Representativeness in order to make many causal judgments, and Representativeness tells us that effects usually resemble their causes. This heuristic, studied by Daniel Kahnemann and Amos Tversky, corresponds to J. G. Frazer’s so-called “Law of Similarity” in magic. Representativeness and other forms of magical thinking appear to yield weaker causal inferences than our intuitions about physical processes or the agency of other people. Accordingly, religious rituals are often employed in situations where a goal cannot be achieved in more obvious ways, but some lesser intuition of causal efficacy can still be generated. Illustrative examples are drawn from ancient Greek rituals of offering, oath-taking, and purification.

  PublisherDOI

Frequent coauthors

• Peter Turchin

  53 shared

• Pieter François

  42 shared

• Harvey Whitehouse

  University of Oxford

  41 shared

• Patrick E. Savage

  University of Auckland

  40 shared

• Marc Plutarch

  Emory University

  36 shared

• Apollonius Rhodius

  American Society of Overseas Research

  36 shared

• Megan Daniels

  36 shared

• Enrico Cioni

  The Alan Turing Institute

  36 shared