Current Research
My research interests are diverse and constantly evolving. What I have written here represents only what of my research I feel comfortable summarizing, not an exhaustive list of my ongoing research projects. Feel free to contact me for more information.

Morphological processing in Semitic

Much of my research (and why I went to grad school) focuses on the role of root-and-pattern morphology in lexical storage and access in Semitic languages. I've conducted research on visual and auditory word recognition in Maltese and Hebrew, using both visual (Forster and Davis, 1984) and auditory (Kouider and Dupoux, 2005) masked priming with lexical decision. Maltese is especially interesting to me: though it descends from Siculo-Arabic, the language has developed in contact with Sicilian, Italian, French, and English over the last millennium. Today, the Maltese lexicon and Maltese morpho-syntax are largely non-Semitic (e.g. Bovingdon and Dalli (2006) estimate that only 41% of the lexicon is of Arabic-origin), leading some to question whether Maltese has lost its Semitic character. However, our research has obtained results consistent with an ongoing role for nonconcatenative morphology in lexical access for Semitic Maltese words (and crucially not for non-Semitic words) (Geary and Ussishkin, 2018). I have also looked at contact-induced morphological change in Maltese (see below).

Representative publications

    Geary, J. (2020). "MaltLex: A database of visual lexical decision responses to 11,000 Maltese words". The 33rd Annual CUNY Conference on Human Sentence Processing (CUNY 2020), Amherst, MA, March 20, 2020. [Abstract] [Poster]
    Geary, J., and Ussishkin, A. (2019). Morphological priming without semantic relationship in Hebrew spoken word recognition. Proceedings of the Linguistics Society of America 4. [https://doi.org/10.3765/plsa.v4i1.4509]


The role of phoneme order in constraining auditory lexical access

Visual word recognition largely tolerates changes to the order of letters in words, as has been demontsrated with various experimetnal paradigms including masked transposed-letter priming (e.g. Forster et al., 1987; Perea et al., 2008) and rapid serial visual presentation (RSVP) (e.g. Perea et al., 2012; Velan and Frost, 2007), as well as by the fact that readers often fail to even notice letter-transpositoins (like the several in this sentence). Since, unlike the visual signal, the acoustic signal unfolds across time with parts of the signal being heard and so processed earlier than other parts, one might expect that the order of elements in the acoustic signal (e.g. phonemes) would play a more essential role in constraining auditory lexical access. This is reflected in many models of spoken word recognition, such as the cohort model (e.g. Marslen-Wilson and Welsh, 1978), in which the order of phonemes in the acoustic signal determines (1) the lexical candidates which comprise the initial cohort and (2) the order in which said candidates are deactivated. Currently, I am working to adapt the visual masked transposed-letter priming paradigm for use in the auditory modality, by using the auditory masked priming paradigm (Kouider and Dupoux, 2005) to test for priming by non-words formed via phoneme-transpositions (Geary, 2019).

Representative publications

    Geary, J. (2019). "Testing the role of phoneme order in lexical access using transposed-phoneme priming". The 32nd Annual CUNY Conference on Human Sentence Processing (CUNY 2019), Boulder, CO, March 29-31, 2019. [Poster]


The role of consonant letters in constraining visual lexical access

Readers of various Indo-European languages are faster to recognize a word when primed subliminally by a string of letters comprising the target's constituent consonants (e.g. csn priming CASINO) (Anderson and Geary, 2019; Duñabeitia and Carreiras, 2011; Grainger et al., 2006; Peressotti and Grainger, 1999), but not when primed by its constituent vowels (e.g. aio priming CASINO), reflecting that consonant letters more tightly constrain lexical identity than do vowel letters (Duñabeitia and Carreiras, 2011). In the Semitic language Maltese, which has a lexicon that is split roughly in half between words of Semitic and non-Semitic origin, consonant-letter primes facilitate the recognition of Semitic words for which such strings comprise the root morpheme, but not non-Semitic words for which such strings are non-morphemic (Geary and Ussishkin, 2018). This suggests that language-specific morphological characteristics may modulate the role of consonant-letters in constraining lexical activation (Anderson and Geary, 2019), and in ongoing research we continue to explore the lack of "subset priming" effects in the non-Semitic stratum of the Maltese lexicon.

Representative publications

    Anderson, S. A., and Geary, J. A. (2019). Form priming by discontinuous consonant letter strings in visual masked priming. Proceedings of the 11th International Conference on the Mental Lexicon 1: e060. [https://doi.org/10.7939/r3-pvan-0083]

Collaborators


Documentation, revitalization, and historical development of Piipaash

Together with staff from the Huhugam Heritage Center and the Salt River Pima-Maricopa Indian Community's Cultural Resources Department, I work with Elders from the Gila River and Salt River Indian Communities towards documenting Piipaash (Pee-Posh, aka "Maricopa"). Long-term goals of our research include the development of a dictionary and learning materials for Piipaash. My role has been to help organize and lead elicitation sessions with Piipaash Elders, to process recordings and fieldnotes from these sessions and from archival records, and to assist in the development of resources for the community, such as Piipaash-language signage for the MAR-5 Interpretive Trail. Recently, I led a three-day workshop on phonetics (covering topics in articulatory phonetics, acoustic phonetics and Praat, and ultrasound and other imaging techniques) for O'odham and Piipaash tribal staff and educators. Additionally, I like to work on language apps in Python (e.g. I have written a Piipaash calculator and math game).


Children's acquisition of Quechua morpho-syntax

As an undergraduate, I assisted Dr. Susan Kalt with research on children's acquisition of Cuzco Quechua morpho-syntax. My duties primarily consisted of processing children's narratives for archival at The Archive of the Indigenous Languages of Latin America. I also studied Cuzco Quechua under Sue's guidance in a field methods course at Boston College, and I wrote a paper on kay 'this' (Geary, 2015).

Today, Sue and I continue to work together on children's acquisition of morpho-syntax in Cuzco Quechua and in South Bolivian Quechua. One of our goals is to compare the influence of Spanish on the Quechua of different Quechua-speaking communities, for instance in terms of rates of morphological borrowing and the use of affixes borrowed from Spanish.

Representative publications

    Kalt, S. E., and Geary, J. A. (2020). "Typological shift in bilinguals' L1: word order and case marking in two varieties of child Quechua". The 94th Annual Meeting of the Linguistic Society of America (LSA 2020), New Orleans, LA, January 4, 2020. [Abstract]

Collaborators


Phonetic documentation of Dakar Wolof

I am writing an "Illustrations of the IPA" article on Dakar Wolof for The Journal of the International Phonetic Association (JIPA), under the supervision of Dr. Diana Archangeli. Dakar Wolof is the dialect of Wolof (Niger-Congo: Atlantic-Congo) spoken in major cities in Senegal, including Dakar and Ndar. However, this work is being conducted with the help of a native-speaker member of the Tucson, AZ community.

Once this article is complete, I hope to look at the acoustic correlates of morphology in Dakar Wolof. For instance, in Dakar certain suffixes trigger the gemination of stem-final singleton consonants (e.g. reversive -i /-i/, as in lem [lɛm] 'to fold' vs. lemmi [lɛi] 'to unfold'), while other homophonous suffixes do not (e.g. inchoative -i /-i/, as in lemi [lɛmi] 'to go and fold'), and one can ask whether the length difference between the singleton and morphological geminate is the same as that between singletons and phonological geminates (e.g. geum [gəm] 'to believe' vs. geumm [gə] 'to close one's eyes'). Such a difference could serve as a cue to the identity of the stem and suffix for listeners.


Contact-induced morphological change in Maltese

In addition to having borrowed a large number of lexical items (Bovingdon and Dalli, 2006), Maltese has borrowed a range of grammatical morphemes from Sicilian, Italian, and English, and many of these are now used productively in the language (Gatt and Fabri, 2018). In some of my work, I look at morphological change in Maltese that has been driven less directly by contact with Indo-European languages, particularly as regards the development of the plural suffix -(i)jiet from the native plural suffix -iet.

Representative publications

    Geary, J. (2017). "The historical development of the Maltese plural suffixes -iet and -(i)jiet". The 23rd International Conference on Historical Linguistics (ICHL 23), San Antonio, TX, August 4, 2017. [Slides]

On the Back Burner

Individual differences in masked priming awareness

I use visual and auditory masked priming to explore the processes underlying word recognition (e.g. whether multimorphemic words are decomposed into their constituent morphemes prior to lexical access). In a masked priming experiment, participants are exposed to some stimulus (the prime) subliminally. Exposure to this stimulus can affect the subsequent processing of another stimulus (the target), such as by facilitating its recognition, which can tell us something about how words are processed, how they are stored in the lexicon, etc. Crucial to the successful use of these paradigms is that participants remain consciously unaware of the primes, as prime awareness permits them to adopt conscious strategies for responding to the target which obscure these underlying processes. In some of my research, I explore what factors (e.g. language background) give rise to greater prime awareness, and which researchers using masked priming must thus attend to.