About

Elaine J. Francis is a Professor of English and Linguistics at Purdue University, where she also serves as the Associate Head in the Department of English and directs the Experimental Linguistics Lab. She teaches linguistics courses at both the graduate and undergraduate levels. Her research employs experimental methods to investigate the syntactic, semantic, discourse-pragmatic, and cognitive factors that influence the grammar and usage of complex sentence structures. Her publications have appeared in numerous journals including Language and Cognition, Glossa Psycholinguistics, Linguistics, Lingua, Cognitive Linguistics, Natural Language and Linguistic Theory, Journal of Linguistics, and Journal of Psycholinguistic Research. Her recent book, 'Gradient Acceptability and Linguistic Theory' (Oxford University Press, 2022), explores the interpretation of data from acceptability judgment tasks in relation to theoretical questions in syntax. She is actively involved in the wider linguistics community, teaching short courses at the Linguistic Society of America Linguistic Institutes, serving on the LSA Ethics Committee, and being a member of the editorial board of Glossa Psycholinguistics. Dr. Francis completed her Ph.D. in Linguistics at the University of Chicago in 1999 and was an Assistant Professor at the University of Hong Kong from 1999 to 2002.

Research topics

• Artificial Intelligence
• Computer Science
• Linguistics
• Philosophy
• Mathematics
• Psychology
• Natural Language Processing
• Statistics

Selected publications

• Beyond Binary Animacy: A Multi-Method Investigation of LMs’ Sensitivity in English Object Relative Clauses

  2025-01-01

  articleOpen accessSenior author

  Animacy is a well-documented factor affecting language production, but its influence on Language Models (LMs) in complex structures like Object Relative Clauses (ORCs) remains underexplored.This study examines LMs' sensitivity to animacy in English ORC structure choice (passive vs. active) using surprisalbased and prompting-based analyses, alongside human baselines.In surprisal-based analysis, DistilGPT-2 best mirrored human preferences, while GPT-Neo and BERT-base showed rigid biases, diverging from human patterns.Prompting-based analysis expanded testing to GPT-4o-mini, Gemini models, and DeepSeek-R1, revealing GPT-4o-mini's stronger human alignment but limited animacy sensitivity in Gemini models and DeepSeek-R1.Some LMs exhibited inconsistencies between analyses, reinforcing that prompting alone is unreliable for assessing linguistic competence.Corpus analysis confirmed that training data alone cannot fully explain animacy sensitivity, suggesting emergent animacy-aware representations.These findings underscore the interaction between training data, model architecture, and linguistic generalization, highlighting the need for integrating structured linguistic knowledge into LMs to enhance their alignment with human sentence processing mechanisms.

  PublisherOA PDFDOI

• Truth‐Value Judgment Tasks in Second Language Research

  Language and Linguistics Compass · 2025-08-19 · 2 citations

  articleOpen accessSenior author

  ABSTRACT This paper provides a focused review of truth‐value judgment tasks (TVJTs) as a method for eliciting interpretations in adult second language learners. We present the historical perspectives, the rationale for their use, the nature of the knowledge they target, and critical design considerations. Additionally, we discuss their effectiveness in uncovering how second language learners access and compute meaning, as well as emerging directions for research and pedagogy using this method. We advocate for refining TVJTs to more accurately capture linguistic competence by empirically validating relevant crucial design features. Moreover, we highlight some of the advantages of incorporating web‐based TVJT experiments, which enhance transparency, facilitate replication, and accommodate a diverse learner population.

  PublisherOA PDFDOI

• Long-term effects of repeated exposure to Subject Island constructions: evidence for syntactic adaptation

  Glossa Psycholinguistics · 2024-08-05

  articleOpen accessSenior author

  Repeated exposure to Subject Island violations can lead to increased acceptability ratings and faster reading times (Chaves & Dery, 2014, 2019; Clausen, 2011; Francom, 2009; Hiramatsu, 2000; Lu et al. 2021; Lu et al., 2022). However, it remains unclear what the nature of this effect and the driving mechanism is. The present paper describes a longitudinal investigation to test whether the effect of repeated exposure to Subject Island constructions is short-lived or whether it can spread over three weeks’ time, as measured by offline measures (Likert acceptability ratings) and online measures (self-paced reading). Using more observations and more sensitive methodologies, our work builds and improves on the only previous longitudinal study on such islands, Snyder (2022). We uncover evidence suggestive of gradual and strategic (by-construction and by-region) adaptation to Subject Island violations, indicated by faster response times, as well as higher acceptability ratings following repeated exposure, most consistent with a syntactic adaptation effect over and above task adaptation.

  PublisherOA PDFDOI

• Resumptive pronouns in Hebrew, Cantonese, and English relative clauses

  2021-12-15

  book-chapter1st authorCorresponding

  Abstract Chapter 7 explores grammatical and processing-based constraints that affect acceptability judgments of relative clauses with a resumptive pronoun in Hebrew, English, and Cantonese. For all three languages, it is shown that some patterns of judgments are open to different theoretical interpretations with respect to the division of labor between grammar and processing. For such cases, it is argued that additional data from corpora and from other types of experiments can help determine the most plausible analysis. As in Chapter 6, it is argued that some patterns of data are plausibly interpreted as involving soft constraints in the grammar. Finally, the studies presented in the chapter are discussed in relation to the proposed typological distinction between grammatical resumption (Cantonese, Hebrew) and intrusive resumption (English). It is argued that a more nuanced interpretation of this distinction may be needed to allow for the synchronic representation of partially grammaticalized constraints.

  PublisherDOI

• Relative clause extraposition and PP extraposition in English and German

  Oxford University Press eBooks · 2021 · 1 citations

  1st authorCorresponding
  • Computer Science
  • Natural Language Processing
  • Artificial Intelligence

  Abstract Chapter 6 applies theoretical concepts from earlier chapters to the interpretation of corpus and acceptability judgment data from several studies of relative clause extraposition and PP extraposition in English and German. It is shown that assumptions about form–meaning isomorphism influence how grammatical constraints that have a semantic component are interpreted, while assumptions about gradient grammaticality influence how lower acceptability ratings of structures that naturally occur in corpora are interpreted. It is further shown that by considering corpus data and data from elicited production tasks, it is possible to narrow down the range of plausible explanations for a given contrast in acceptability. Finally, it is argued that certain gradient patterns of acceptability are most plausibly interpreted as involving soft constraints—non-categorical grammatical constraints that encode the conventional preferences of language users.

  PublisherDOI

• On distinguishing formal syntactic constraints from processing constraints

  2021-12-15

  book-chapter1st authorCorresponding

  Abstract Chapter 4 explores ongoing debates in experimental syntax regarding the distinction between syntactic and processing-based effects on acceptability. Several types of arguments have figured prominently in the debates. Effects that have been claimed to favor a syntactic explanation include overgeneration of ungrammatical sentence types, cross-linguistic differences, resistance to repeated exposure effects, and resistance to individual differences in working memory capacity. Effects that have been claimed to favor a processing-based explanation include amelioration effects, isomorphism between acceptability ratings and time-sensitive measures such as reading time, repeated exposure effects, and effects of individual differences in working memory capacity. Each of these types of explanations is considered in turn, drawing on studies of syntactic islands, subject–verb agreement, missing VP illusions, and Superiority effects.

  PublisherDOI

• On distinguishing formal syntactic constraints from other aspects of linguistic knowledge

  2021-12-15

  book-chapter1st authorCorresponding

  Abstract Chapter 3 examines how syntactic effects on acceptability can be distinguished from effects of other aspects of linguistic knowledge, including semantics, pragmatics, and prosody. It is argued that in cases where the acceptability of purportedly ungrammatical sentences can be substantially improved by changing the discourse context or by changing syntactically irrelevant aspects of the sentence, syntactic constraints may not be needed. It is further shown that arguments in favor of a non-syntactic explanation can be strengthened with reference to attested examples and with data from different types of experimental tasks that specifically target semantic, pragmatic, or prosodic features. These points are supported with examples from outbound anaphora and factive islands in English, word order variation in Czech, and split intransitivity in German.

  PublisherDOI

• On the relationship between corpus frequency and acceptability

  Oxford University Press eBooks · 2021 · 1 citations

  1st authorCorresponding
  • Computer Science
  • Artificial Intelligence
  • Psychology

  Abstract Chapter 5 starts by observing that linguists from different theoretical traditions tend to express different views on the relative importance of judgment data and corpus data for developing theories of grammatical knowledge. Similarly, although everyone recognizes frequency of use as a factor to which language users are sensitive, there is disagreement as to how it relates to acceptability. Proponents of usage-based and probabilistic theories have tended to emphasize the close correlations between frequency and acceptability, while proponents of traditional generative theories have tended to emphasize cases in which acceptability judgments differ among low-frequency or zero-frequency structures. This chapter discusses a few studies which have directly examined the relationship between corpus frequency and acceptability and considers their implications for these different theoretical approaches. Alongside more traditional corpus studies, machine learning studies are also considered.

  PublisherDOI

• The problem of gradient acceptability

  2021-12-15

  book-chapter1st authorCorresponding

  Abstract Chapter 1 first provides an overview of the relation between grammatical knowledge and acceptability judgments as it has been understood in the history of generative grammar. It then introduces the concept of gradient acceptability and discusses several of the factors that can cause it. These factors, which include formal syntactic constraints, semantic and pragmatic anomalies, prosodic constraints, and processing effects, are illustrated using several examples from the literature. The introductory chapter concludes by giving a preview of the remaining chapter contents, the organization of the chapters, and the major themes and contentions of the book.

  PublisherDOI

• Gradient Acceptability and Linguistic Theory

  Oxford University Press eBooks · 2021 · 49 citations

  1st authorCorresponding
  • Computer Science
  • Artificial Intelligence
  • Linguistics

  Abstract In Gradient Acceptability and Linguistic Theory, Elaine J. Francis examines a challenging problem at the intersection of theoretical linguistics and the psychology of language: the problem of interpreting gradient judgments of sentence acceptability in relation to theories of grammatical knowledge. This problem is important because acceptability judgments constitute the primary source of data on which such theories have been built, despite being susceptible to various extra-grammatical factors. Through a review of experimental and corpus-based research on a variety of syntactic phenomena and an in-depth examination of two case studies, Francis argues for two main positions. The first is that converging evidence from online comprehension tasks, elicited production tasks, and corpora of naturally occurring discourse can help determine the sources of variation in acceptability judgments and narrow down the range of plausible theoretical interpretations. The second is that the interpretation of judgment data depends crucially on one’s theoretical commitments and assumptions, especially with respect to the nature of the syntax–semantics interface and the choice of either a categorical or a gradient notion of grammaticality. The theoretical frameworks considered in this book include derivational theories (e.g. Minimalism, Principles and Parameters), constraint-based theories (e.g. Sign-Based Construction Grammar, Simpler Syntax), competition-based theories (e.g. Stochastic Optimality Theory, Decathlon Model), and usage-based approaches. While showing that acceptability judgment data are typically compatible with the assumptions of various theoretical frameworks, Francis argues that some gradient phenomena are best captured within frameworks that permit soft constraints—non-categorical grammatical constraints that encode the conventional preferences of language users.

  PublisherDOI

Frequent coauthors

• Sarah E. Duffy

  Northumbria University

  83 shared

• Jeannette Littlemore

  83 shared

• Bodo Winter

  University of Birmingham

  83 shared

• Laura A. Michaelis

  62 shared

• Gary M. Oppenheim

  Bangor University

  58 shared

• Rui P. Chaves

  University at Buffalo, State University of New York

  26 shared

• Judith Holler

  Radboud University Nijmegen

  25 shared

• Panos Athanasopoulos

  Stellenbosch University

  16 shared

Education

• Ph.D., Linguistics

  University of Chicago

  1999

• B.A., Linguistics

  College of William and Mary

  1993