Broadly speaking, my work addresses the intersection of language processing and linguistic form. I am interested in the ways in which processing-related factors relate to variation in linguistic form (i.e. predominantly speaker-oriented processes), and also in the ways in which linguistic variation impacts language processing (i.e. predominantly listener-oriented processes). Most of my work focuses on phonetic variation, but I have also published on lexical and even syntactic variation. If it varies, I will study it!
Much of my work examines processing in bilingual speakers, but importantly, I do not take bilinguals to be a "special" population, per se; my work is motivated by the assumption that we are all operating on the same hardware, so differences across speaker populations all have something interesting to say about the basic processes that enable language acquisition, speech planning and comprehension, and ultimately, successful communication.
Below, I have grouped my major published works into three main themes, with links for downloading these papers. You can also check out my Google Scholar page for a complete listing of published work, and/or send me an email if any of the links below don't seem to be working.
Linking speech planning processes to variation in linguistic form
My work on this topic looks at the ways in which factors that make production planning more or less difficult are ultimately reflected in the linguistic form of produced speech.
Fricke, Baese-Berk, & Goldrick, 2016 (html, pdf)
The involvement of the lexicon during speech planning: Phonological neighborhood density predicts word-initial voice onset time better than minimal pair status.
Jacobs, Fricke, & Kroll, 2016 (html, pdf)
Language immersion impacts cross-language phonetic influence during cognate production in second language learners.
Fricke & Kootstra, 2016 (html, pdf)
The influence of the non-target language on the target language can be seen during spontaneous codeswitching in the form of statistically robust phonetic, lexical, and syntactic patterns. Moreover, bilingual listeners are highly attuned to these patterns!
BerrĂos, Swain, & Fricke, 2023 (html, pdf)
Insights from corpus psycholinguistics can be used to train more intelligent machine learning models. In ongoing work as part of my collaborative NSF grant (NSF CISE Grant #IIS-2007656) with Yulia Tsvetkov and Shuly Wintner, we are developing new methods for training machine learning models, building bilingual corpora, and improving technologies for bilingual language generation.
Relating language experience, language control, and language processing
A second major theme in my work is the idea that speakers' experiences using language necessarily impact their processing strategies. I have argued that this relationship constitutes a fundamental aspect of language processing, and that taking it seriously in all of its forms is essential for moving the field forward.
Fricke, Zirnstein, Navarro-Torres, & Kroll, 2019 (html, pdf)
In this review, we argue that studying the relationship between language use and language processing can help us better understand the basic principles that govern how languages are learned and represented.
This paper delves into some of the ways in which aspects of language experience help predict bilingual speech perception in noise. Contrary to what has been suggested in the literature, I do not find that English speech perception is particularly disrupted for U.S.-based Spanish heritage listeners, and I do not find evidence that background noise leads to increased activation of the non-target language.
Fricke & Zirnstein, 2022 (html, pdf)
This paper looks at speech perception in noise for the non-dominant language. We find some evidence that coping with informational masking may draw on the same cognitive resources as making predictions about upcoming words, and we also find that listeners with greater inhibitory control experience less disruption from the dominant language.
Psycholinguistic origins of diachronic sound changes
I am also interested in how processing-related factors can lead to biases that are ultimately reflected in the ways in which languages change over time (i.e. on a historical time scale). This is an area in which I hope to grow my research further!
Katz & Fricke, 2018 (html, pdf)
We show that the cross-linguistically common pattern of word-initial stops and word-medial taps is associated with better performance in a word segmentation task, as compared to the cross-linguistically unattested pattern of word-initial taps and word-medial stops. So lenition patterns may have a perceptual basis!
Fricke, Bjorndahl, Heath, & Wiener, in revision (manuscript available on request)
We test an empirical assumption underpinning several computational models of sound change (namely, that phonetically ambiguous words may be less likely to be encoded in long-term memory), and we find no evidence for it, indicating that models relying on this assumption may require some revision.