Research

Code-switching is a common conversational practice used by multilinguals whereby they switch between their two or more languages within the same conversation. Code-switching is an especially relevant empirical domain because it is one of the few naturalistic manifestations of the combination of two systems within language production, both across and within sententential boundaries. 

Ban on word-internal code-switching

Focusing on illicit code-switches within the DP, I contribute to our understanding of the ban on word-internal code-switching by reporting on the acceptability of code-switching within prosodic constituents that are simultaneously syntactically complex as well as low on the prosodic hierarchy. This study tests the acceptability of code-switching within varying locations in the Egyptian Arabic construct state, a syntactic noun phrase found in Semitic languages. The construct state is particularly relevant for discussions on the syntax-phonology interface, especially within code-switching, because it is a complex syntactic unit that is mapped to a single prosodic word. Ultimately, I report results on the baseline acceptability judgments of code-switched sentences within various locations of the construct state, arguing that said ban should be extended to this syntactic domain (Sedarous, 2020, 2022).

Code-switching and lexical cognates

In this study I explored psycholinguistic factors which have been argued to play a role in code-switching, by investigating the effects of cognates on the acceptability of grammatical and ungrammatical code-switches. Cognates have been argued to have a facilitatory effect on code-switch location, as code-switched corpora has consistently shown an above chance tendency for code-switch locations to occur near a cognate. In Sedarous (2020, 2021, 2022), I tested whether the facilitatory effect of cognates which had been found in already produced codes-switched corpora translated into bilinguals’ acceptability of code-switched sentences. Ultimately, I concluded that bilingual speakers were not sensitive to cognates in offline processing tasks concerning code-switched sentences. We are yet to determine whether trigger words would also have no effects in online processing of code-switched sentences. 

The processing of structurally (in)congruent mixed-language expressions

In an ongoing collaboration with Sarah Phillips, we contribute to our understanding of language processing by testing bilinguals’ sensitivity to various single- and mixed-language expressions. Assuming that bilinguals possess a single parser that processes both types of expressions, we investigate which element(s) bilinguals favor when building (in)congruent mixed-language expressions. We conducted a picture identification task where three bilingual groups – Arabic/English, Korean/English, and Spanish/English – were presented with three-word nominal expressions that varied by word order (Det-Mod-N, Det-N-Mod) and location of switch (no switch, after word 1, after word 2, double switch). Our results showed that participants elicited faster processing when the word order of the mixed-language expression matched the preferences of the language of the noun. Based on these results, we propose that bilinguals are biased by distributional preferences of heads of extended projections (e.g nouns, verbs) during online comprehension processing, even if they can generate alternative linearizations (see Phillips and Sedarous, 2022, Phillips and Sedarous, under review). 

In various work, I look broadly at structures that fully or partially overlap in their surface syntax across Egyptian Arabic and English. I probe into these questions by investigating the processing of various long-distance dependencies (LDDs). 

Structural Priming 

In my analysis of an English-to-Egyptian Arabic bilingual corpus (the MADAR corpus, Boumar et al., 2018), I observed that the wh-structures used in English triggered repetition of the same construction in the switch to Egyptian Arabic (Sedarous and Baptista, 2022). Focusing on wh-questions that are canonically in the object position, my corpus analysis revealed the above chance presence of gap strategy wh-questions in Egyptian Arabic questions when translated from English. The presence of structural repetition suggests that some structures, specifically the gap strategy found with non-nominal wh-constituents, are not language specific.

Resumptive Pronoun Sensitivity in Restrictive Relative Clauses

In an acceptability judgement task involving code-switched restrictive relative clauses, I show that speakers’ dominance of their two or more languages affects their sensitivity to illicit sentences in partially convergent structures (Sedarous, 2021, contact me for slides). I tested speakers’ sensitivity to resumptive pronouns in code switched RRCs and found that heritage speakers were more sensitive to the illicit presence of a resumptive pronoun in English than they were to the absence of resumptives in Egyptian Arabic. These results indicate that speakers’ dominance of their two languages has an effect on the syntactic representations of their partially convergent structures. 

Overlapping surface order, different derivations: Wh-resumptive structures in Egyptian Arabic and English

In this paper I investigate heritage bilinguals’ sensitivity to two structures that display overlapping word orders across their two languages but are argued to have different derivational properties in their formation. Using an acceptability judgment task to compare island effects in unilingual and code-switched A-bar dependencies (wh-questions) with cliticized resumptive pronouns, I report evidence of island sensitivity in unilingual English, code-switched Egyptian Arabic to English, and code-switched English to Egyptian Arabic wh-questions, but not in unilingual Egyptian Arabic wh-questions. I analyze these results under a framework which proposes that the distribution of resumptive pronouns satisfies independent well-formedness conditions of both the syntax and the (morpho)phonology, following Rasin (2017) for Hebrew (see Sedarous 2022, 2023). The results of this experiment also suggest that this population of bilinguals may treat Egyptian Arabic fronted wh-questions as instances of moved wh-phrases sensitive to syntactic islands.

Optimization strategies and the acquisition of seemingly congruent syntactic phrase

In order to explain how and why some features are favored over others for different populations of bilinguals, in collaboration with Marlyse Baptista, we emphasize the role of congruence (Baptista, 2020) in heritage language acquisition. Here, we propose that the similarities (the congruent features) that speakers perceive between the languages in contact are favored to participate in the emergence of new languages or new language varieties. We suggest that the presence of this novel derivational strategy in heritage Egyptian Arabic/English dominant speakers may have resulted from optimization strategies whereby these speakers are operationalizing L1/L2 syntactic mappings whenever possible. This then resulted in a structure that is ultimately common to both English and Egyptian Arabic, at least for this population of bilinguals (Sedarous and Baptista, 2024)

Although long distance dependencies are unconstrained with respect to length between the filler and the gap, they are said to be constrained by islands, i.e constructions in which wh-phrases cannot ‘escape,’ and so filler-gap dependencies cannot be formed (Ross 1967). My interest in the processing/syntax interface has naturally led to a deep interest in investigating the processing and storage of syntactic islands.

Pragmatic Contexts

In collaboration with several researchers from Michigan and Princeton, we tested the hypothesis that the island status of clausal adjuncts, as determined by judgments on wh-questions, are predicted by the degree of “backgroundedness” of the adjuncts, as determined by a separate negation task. Results of two experiments support the hypothesis that acceptability of extraction from adjuncts in wh-questions is inversely correlated with the degree to which the adjunct is backgrounded in discourse. Taken together, results show that temporal clausal adjuncts (headed by before, after, while) are stronger islands than adjuncts that are causal (headed by to or by). This demonstrates that adjuncts differ in degree of island status, depending on their meaning, despite parallel syntactic structure. See Namboodiripad et al., (2022).

Syntactic Category

In collaboration with Andrew McInnerney, we test to what extent island strength is affected by the syntactic category of the extracted constituent - specifically, PP vs. NP. We focus on adjunct islands, which  have been claimed to show selective opacity (at least marginally) to NP extraction but not PP extraction. We, however, have found that this preference is not specific to island contexts. In on going experimental work, we find that PP extraction is rated as being consistently less acceptable than NP extraction, both within and outside of island contexts. Further work is needed to understand the general preference for extraction of NP over PP. See McInnerney and Sedarous (2022), for some of our early findings.

Prosodic factors 

Based on observations of how resumptive pronouns are produced in naturalistic contexts, in collaboration with Savithry Namboodiripad and Felicia Bisnath, we test whether cliticizing a resumptive pronoun (by phonologically reducing it and attaching it prosodically to a prejacent) increases acceptability compared to morphophonologically independent “full” resumptive pronouns and gaps.  Using audio stimuli, we tested subject- and that-trace islands. Our results indicate that although cliticizing did not lead to higher acceptability for RPs, means for all RP constructions were comparable to grammatical-but-syntactically-complex fillers and significantly higher than ungrammatical fillers. This indicates that prosody may reconcile the apparent disconnect between acceptability and production of English RPs. See Sedarous, Bisnath, and Namboodiripad (2023) for some of our initial results. Feel free to contact me for any additional materials or our future plans.

As an Egyptian-American of Coptic heritage, my identity naturally informs my research, my commitment to diversity, equity, and inclusion is present throughout my own scholarship. My research focuses mainly on Egyptian Arabic-American English bilinguals belonging to immigrant communities in the United States, an understudied population within the field of Linguistics. For this reason, much of the established methodologies in psycholinguistics could not be easily transferred onto my work, and some of the baseline terminology used to in describing the linguistic profiles of language users did not adequately represent the linguistic knowledge of this community.

Methodology: audio stimuli and portable eye-tracking/pupilometry

Methodologically, the fields of theoretical and experimental (psycho)linguistics overrepresents Western, educated, industrialized, rich participants in research, and the methods involved in collecting data, e.g acceptability judgement tasks using written stimuli, self-paced reading, etc., have been catered to these participants at the exclusion of others. I collaborated with Savithry Namboodiripad on a methodology paper in which we strongly emphasize the favorable outcomes associated with moving towards utilizing audio stimuli in linguistic experiments, and away from using written stimuli, which systematically exclude behavioral investigation of languages without writing systems, nonliterate populations, and nonstandard varieties of languages (see Sedarous and Namboodiripad, 2020). In addition to advocating for the use of audio stimuli, I am currently collaborating with Savithry Namboodiripad and Felicia Bisnath on developing a portable pupillometry setup that is accessible for fieldwork, and sensitive to draker eyes. Access to this equipment will make it more possible for researchers to gather processing data from speakers who do not live within close proximity to universities, and so are not represented in these kinds of studies (see Tobin et al., 2020, for some of our group's initial findings).

Terminology: Problematizing the "native speaker"

Both the concept of a "native speaker" and the term itself has been critiqued and problematized across fields such as linguistic anthropology, second language acquisition, and English language teaching. However, within less peripheralized disciplines of linguistics this term is often used uncritically. As part of a large collaboration, we traced the history of the term “native speaker” and showed how historically the term has excluded racialized/ethnicized and disabled individuals and communities. Descriptions of the ‘native speaker' are typically vague, with undertones of monolingualism and linguistic homogeneity that do not reflect the linguistic reality of a majority of the world. We propose that researchers engaged in linguistic research specifically describe their participants’ language use and learning as contextually relevant. We have presented this work at LSA2022 (see Birkeland et al., 2022) and have published our findings as a Commentary piece in the journal Language (see Birkeland et al., 2024). 

Revitalizing Attitudes Towards Creole Languages

Inspired by Léglise and Migge's (2006), the CCLE lab at the University of Michigan used a community centered approach to explore conceptions and assumptions about Cabo Verdean Kriolu (spoken in the Cabo Verde islands), Kwéyòl Donmnik (spoken on the island of Dominica), and Trinidadian English Creole (spoken in Trinidad and Tobago). We conducted two surveys, one with language experts (in-group community members) and the other with linguists, and a follow-up workshop to further our understanding of these two populations’ representations and ideologies about Creole languages. Based on these results we make several recommendations regarding the presentation of Creoles in linguistics classrooms, such as integrating Creoles throughout a course rather than structurally exceptionalizing them and relegating them to a single unit. This paper is published to the Charity-Hudley, Mallinson, and Bucholtz (eds.) Decolonizing Linguistics volume. Our group has also been invited to submit an additional followup to our study that is set to appear in the Teaching Linguistics section of Language.