Mauricio Navarro '14 discusses his 2013 linguistics and cognitive science Summer Undergraduate Research Project, which he undertook under the mentorship of Professor David Divita. He conducted in-depth interviews with 7 transfeminine-identified people in Guadalajara, Mexico, as a basis for a fuller sociolinguistic analysis.

In the Linguistics and Cognitive Science Department, students are encouraged to undertake research projects, making new explorations into this field. Below are completed Summer Undergraduate Research Program (SURP) projects from our students.

Pomona Passions. "There are plenty of research opportunities that I've had here...because I've been encouraged to carry out [my] ideas and projects."


Disfluent windows into bilingual speech planning: Evidence from spontaneous codeswitching

Mary Paster and student Rodrigo Ranero work to Reclaim the Rumsen Language. 

Orren Arad-Neeman ’16; Mentors: Deborah Burke, Melinda Fricke (Pennsylvania State University), and Judith Kroll (Pennsylvania State University

Previous research on speech disfluency largely associates disfluencies with difficulties in speech planning (Bortfeld et al. 2001, Arnold et al. 2007, inter alia). Further, different types of disfluencies may signal different types of planning difficulties on the part of the speaker (MacGregor et al., 2009). We previously found that codeswitched utterances were significantly more likely to be disfluent than unilingual utterances in the Bangor Miami Corpus of Spontaneous Codeswitching (BMC; Deuchar et al., 2014). However, we did not examine the nature of these disfluencies. Therefore, this project asks whether the specific form and location of disfluencies in the speech of balanced bilinguals can reveal differences in the mechanisms involved in planning codeswitched versus non-codeswitched speech. We address this question by classifying the speech disfluencies present in conversations between balanced English-Spanish bilinguals in the BMC. We tagged each codeswitched utterance and matching unilingual utterance for disfluencies. We are currently in the process of coding the disfluencies on a finer-grained scale. Following previous work, disfluencies will be classified as a filled pause, repetition, articulation error, restart, elongation, laugh, insertion, other, or any combination thereof. We will also code for their location to determine whether the timing of planning difficulty differs significantly between codeswitched and unilingual utterances.
Funding provided by: Craddock

Effects of Expectancy on Retrospective Duration Judgments and Memory for Music

Shaina Brady ’16; Mentor: Laura Johnson

The framework of music can be broken into two overarching fundamental structures, rhythm and melody. These aspects have been demonstrated to create in the listener a sense of expectancy as to what will logically occur next in a musical work. The purpose of this present study is to examine the effects that these rhythmic and melodic expectancies have on a listener’s memory and accuracy in making a retrospective judgment of the duration of a segment of music. Participants were given a set of four short songs to listen to that were manipulated to be either rhythmic or arrhythmic and tonal or atonal. After each piece, participants filled out a short survey about the aesthetics of the music. For the retrospective timing task, they were then presented with the first four seconds of each piece that they heard, and were asked to imagine continuing the song in their head and to press a button when they reached the end. Finally, they were given a memory test to see how well they could recognize the particular music clips. The results from the retrospective timing task demonstrated an interaction between rhythm and melody, with longer judgments for tonal than atonal in the arrhythmic condition, and rhythmic than arrhythmic in the atonal condition. In addition, recall was better for tonal than atonal music, but there was no effect of rhythm. These findings allow for a better understanding of the effects of rhythm and melody on the recall of music duration and underlying structures.
Funding provided by: Pomona Unrestricted

Effects of motor expertise on body part priming, imitation and mental rotation processes in expert acrobats.

Harrison Goodall ’16; Mentors: Deborah Burke and Emily Cross '01 (University of Bangor); Collaborator: Nicola Smith (University of Bangor)

The mirror neuron system has been a widely investigated field of cognitive science over the past decade. Perhaps one of the most intriguing findings from recent research is evidence for automatic imitation and body part priming. When a subject’s attention is drawn to another person’s body part, the subject is able to move their corresponding body part faster than if their attention is directed elsewhere or to a different body part. These findings support the theory of embodiment, which states perception and mimicking the motion of others is accomplished by mapping their movement onto one’s own body schema. However this theory has called into question how humans are able to process movements they are not capable of embodying. This four-part study investigated the effects of inversion on perceptual process of circus acrobats compared to controls. Participants completed a mindfulness questionnaire, body part priming tasks and two mental rotation tasks, one using shapes the other using bodies. All the tasks used stimuli in upright and inverted configurations. We hypothesize that, compared to controls, the acrobats’ motor and inversion expertise will allow them to respond to the inverted images faster, as well as complete the body mental rotation task faster than controls. We also expect higher levels of expertise will correlate with higher reports of mindfulness, as expertise requires intention and attention, attributes directly correlated with reports of mindful behavior.
Funding provided by: Cion Estate

Language Processing and its Effect On Time Perception

Evan Hamaguchi ’16; Mentor: Laura Johnson

Two experiments examined whether language processing demands have an effect on subjects’ ability to simultaneously memorize sentences and estimate time intervals. The first study varied syntactic complexity through word order, and the second varied sentence length. Subjects were shown sentences of varying complexity for different durations, then were asked to reproduce the duration by pressing a button to mark the beginning and end of the interval. After each sentence and interval estimation, they were asked to re-type the sentence in question to the best of their ability. For the duration reproduction task, the results revealed statistically marginal effects of syntactic complexity and sentence length, with longer duration estimates given for less complex sentences and shorter sentences. In the recall task, subjects’ recall was less accurate for longer sentences. Results also revealed that subjects gave longer estimates for incorrectly-recalled sentences than for correctly-recalled sentences. These results suggest that time perception is influenced by language processing through competition for attentional resources. This conclusion expands applications of the attentional gate model of time perception to include sentence processing, building on similar findings regarding lexical processing and multitasking.
Funding provided by: Pomona Unrestricted

The Role of Gesture in Early Verb Learning

Emily Kubota ’17; Mentor: Deborah Burke; Collaborator: Liz Wakefield (University of Chicago), Casey Hall (University of Chicago)

Previous research has shown that action enhances verb learning in four and five olds. When children are taught novel verbs in the context of novel objects they are better able to remember the words if they perform a relevant action. The results implicate the role of embodiment in verb learning, wherein a mental representation of a physical action facilitates understanding of the meaning of a verb. Unfortunately, learning verbs in the context of specific objects can prevent generalization (for example: if a child learns “twisting” through twisting a jar, it can be difficult to understand that “twisting” can also apply to twisting a doorknob). This study examined whether gesture can be a learning tool that allows children to benefit from the embodiment of the verb, without binding the verb to a specific object. Children were taught novel verbs (ratching, tiffing, leaming, yocking) by either performing an action on an object, or performing the gesture of the action near the object. Then they were shown videos of actions being performed on objects and asked to choose the video depicting the learned verb. In some cases the correct video depicted the learned verb performed on the original object, in other cases the correct video depicted the learned verb performed on a different object (trials requiring generalization). Preliminary results indicate that children who learned using gesture were better able to generalize verbs to new contexts than those who learned through action.
Funding provided by: Craddock

Documenting the Logoori Language

Rebecca Shipan '17 and Cliff Mountjoy-Venning '16; Mentors: Michael Diercks and Meredith Landman

Logoori, sometimes referred to as Maragoli, is a Bantu language in the Luyia subgroup spoken by nearly 620,000 people in Western Kenya and Uganda. Our team is part of a collaborative National Science Foundation project including 12 researchers from 8 different institutions who are working to document four Luyia languages. This summer, the Pomona College team began an investigation into the syntactic properties of Logoori, with the goal of producing a broad overview of the grammatical structure of the language as well as more detailed descriptions of selected topics.  We worked with a native speaker of Logoori to gather linguistic data on a wide range of topics. We reviewed the existing literature to identify areas of inquiry and to prepare interviews with our speaker that were designed to evaluate which constructions are acceptable in the language. After gathering the data, we developed analyses of the patterns we observed, drafting our conclusions into 139 pages on twelve major aspects of Logoori grammar. These chapters comprise the foundation of the syntax portions of the Logoori reference grammar being published as part of the NSF project. This work contributes to the documentation of a minority language (Logoori) in an understudied language family (Luyia); this research serves these Kenyan communities by preserving aspects of their language and culture, as well as contributing to our knowledge of what kinds of linguistic structures the human mind can generate. 
Funding provided by: Cion Estate


Classification of an Event-Related Potential that follows kinematic errors in a motor reaching task

Dylan Pollack (2015); Additional Collaborator(s): Paul DiZio (Brandeis University), Alberto Pierobon (Brandeis University); Mentor(s): Deborah Burke

Abstract: Electroencephalography (EEG) is a brain recording technique in which cortical electrical activity is measured using scalp electrodes. Event- related potentials (ERP)—signals traditionally found by averaging multiple time-locked EEG recordings— are thought to reflect neural correlates of cognitive processes. Previous research has identified a negative ERP that occurs 200 ms following performance errors on a perceptual discrimination task. The goals of this study were 1) to investigate the possibility that an analogous signal follows errors in a motor reaching task and 2) to identify this ERP in individual trials. Reaching errors were produced by having subjects move a robotic manipulandum that occasionally delivered lateral forces, causing the arm to deviate away from its intended target. EEG signals were recorded during 64 unexpected force-induced motor errors (“catch” condition) and compared to recordings from 64 trials in which reaches were accurate (“baseline” condition). Machine learning techniques were used to construct a support vector machine (SVM) that classified EEG recordings as either “catch” or “baseline”. On average, the SVM labeled 93% of trials correctly. Future experiments may be needed to reduce artifacts that artificially inflate classification accuracy. Successful identification of an ERP related to motor error detection could be used to help rehabilitate physical therapy patients.
Funding provided by: Pomona College SURP

Neural Substrates of Imagining the Future: Multivoxel Pattern Analysis

Karin Denton (2015); Additional Collaborator(s): Karl Szpunar (Harvard University), Daniel Schacter (Harvard University); Mentor(s): Deborah Burke

Abstract: Recent research in cognitive science has investigated how people use information from their memory to imagine future events in their lives (Schacter et al., 2008). The present study used multivoxel pattern analysis (MVPA) of fMRI data in order to investigate the neural substrates of imagining the future, using the data from Szpunar et al. (2014). Participants in the study were told to imagine a future event involving a specified set of person, place, and object. Trials used either a novel or repeated set in order to elicit repetition-related reductions in neural activity. MVPA involves feeding data from each scan into a classification algorithm that is trained to recognize the trial type of each scan. Traditional analysis looks for overall differences in neural activation, whereas MVPA can recognize different patterns of neural activity. In this investigation, we used a back propogation neural network to distinguish novel and repeat trials. During training, the classifier adjusts the weights of each voxel according to its importance in determining trial type. These weights showed the importance of a frontal-parietal-temporal network in distinguishing repeated and novel future event simulations, with specific regions of the network being tied to the imagination of people, places, and objects.
Funding provided by: National Endowment for the Humanities

Putting the Testing Effect to the Test on TOTs: Do Retrieval Attempts Reduce Word Finding Failures?

Emma Gardner (2016); Student Collaborator(s): Adrian Yuratovac (2016); Mentor(s): Deborah Burke

Abstract: Long-term retention of new learning is improved more by repeated testing than studying, an effect attributed to the retrieval attempts during testing. We tested this effect on tip-of-the-tongue experiences (TOTs), in which known words temporarily cannot be retrieved. Participants were older adults in Control, Study, and Test groups. In Trial 1, all groups were presented with TOT-inducing questions that required production of a famous name or a word. In Trials 2 and 3, the Test group had to answer the questions again without feedback, while the Study group was given the questions with answers to study. The Control group received only Trial 1. All groups were tested again after a 1-week and a 3-month interval. After one week, correct responses increased from the initial test in all groups, but most for Study and least for Control. TOTs decreased from the first to the 1-week test in all groups, and this effect did not vary with the condition. The proportion of TOTs in Trial 1 that were resolved at the 1-week test was significantly higher for Study than for Test, and for Test than Control. Three-month data is currently incomplete. Our results so far suggest that repeated testing and studying are both effective in preventing TOTs for a target word, even after three months. Repeated testing, which is superior to Study when learning new information, affects TOTs differently, implying that retrieval of new learning involves a different pathway than that for old knowledge.
Funding provided by: Evelyn B. Craddock-McVicar Memorial Fund (EG); Fletcher Jones Foundation (AY)

How much does the reward matter? That depends on who's winning it: exploring culture and self- construal in EEG responses to a gambling task

Emily Wasserman (2015); Additional Collaborator(s): Shinobu Kitayama (University of Michigan), James Glazer (University of Michigan); Mentor(s): Deborah Burke

Abstract: Previous work in cultural psychology suggests that European-Americans are more likely than Asian-Americans to construe themselves as individuals, rather than group members. I investigated this question by examining electrocortical responses to reward/penalty feedback in a gambling task using a previously collected dataset. 70 subjects (35 Euro-/35 Asian- American) had performed a gambling task, winning/losing 50/150 points per trial. Replicating previous work, we found a negative deflection of the electrical potential at 220ms post-feedback (marking the valence of the outcome), followed by a positive peak at 300ms. This positive component, P3, is assumed to reflect subjective salience of reward/penalty feedback, with larger P3 typically accompanying either larger magnitudes or positive valence of reward/penalty feedback. However, previous research did not sufficiently control for the influence of the preceding negative peak when measuring P3, making these effects difficult to interpret. As expected, when measured in a way that reduces this influence, P3 was found to be larger for European-Americans than Asian- Americans overall. This suggests that all feedback, regardless of valence or magnitude, was interpreted by European-Americans to be self-relevant. In addition, loss trials elicited larger P3 than wins across all subjects, an unexpected result which may indicate the usefulness of this new measurement technique in more accurately assessing P3 and other components.
Funding provided by: National Endowment for the Humanities

Filling Two Verbs with One Tone: Tense and Aspect in Maragoli

Alex Samuels (2015); Mentor(s): Mary Paster, Michael Diercks

Abstract: Maragoli, a tonal language in the Luyia group of Bantu languages spoken in Western Kenya, has eight different tenses which surface on its verbs. Each of these has its own tonal pattern, or “melody,” which along with various prefixes and suffixes allows speakers to distinguish the tenses. This project looked at five of these tenses across a selection of verbs in order to determine the melodies used in each. The verbs used varied in number and length of syllables, and were found to fall into two lexical types: high-toned and non-high- toned. High-toned verbs have a high tone on the initial syllable of the verb root, while non-high- toned verbs do not. The tense melody adds more high tones which can then interact with the initial high tone to change the phonological output. Both lexical and melodic high tones were found to undergo a complex set of regular phonological processes that may remove or spread them from their original locations. Any syllable that does not receive a high tone from any of these sources is produced with a low tone. Research was focused on differentiating the lexical high tone from the melody and determining how the two interacted in each of the five tenses analyzed.
Funding provided by: National Endowment for the Humanities

Switch costs in spontaneous bilingual codeswitching: Evidence from disfluencies and speech rate

Orren Arad-Neeman (2016); Additional Collaborator(s): Melinda Fricke (Pennsylvania State University), Judith Kroll (Pennsylvania State University); Mentor(s): Michael Diercks

Orren Arad-Neeman '16 discusses her Summer Undergraduate Research Project

Abstract: Studies employing the language switching paradigm have consistently provided evidence for a “switch cost”: when bilinguals are required to switch languages during picture naming, errors and naming latencies tend to increase (Meuter & Allport, 1999). This project asks whether there is evidence for a switch cost during spontaneous codeswitching. We address this question by examining speech disfluency and speech rate in conversations between balanced English-Spanish bilinguals in the Bangor Miami Corpus of Spontaneous Codeswitching (Deuchar et al., in preparation). For each codeswitched utterance, a set of unilingual utterances matched for several factors were identified. The critical portion of each utterance (the portion preceding the switch point or matched non-switch point) was hand-labeled using Praat (Boersma & Weenink, 2014), and speech rate was calculated by dividing the duration by the number of syllables in that portion. Utterances containing disfluencies were hand-tagged and excluded from the speech rate analysis. It was found that codeswitched utterances were significantly more likely to be disfluent than unilingual utterances. Additionally, not only were mean syllable durations significantly shorter in Spanish than in English, but also were significantly shorter in unilingual than in codeswitched utterances. These results provide initial converging evidence for a processing cost associated with switching languages in spontaneous speech.
Funding provided by: National Endowment for the Humanities


Putting the Testing Effect to the Test: Studying Versus Testing to Reduce Word Retrieval Failures

Emma Gardner (2016); Mentor(s): Deborah Burke

Abstract: Previous research has shown that long-term retention of new learning benefits more from repeated testing than from repeated studying. This has been attributed to the more frequent self-generated retrieval attempts involved in testing than in studying. In the present study, the repeated study versus testing paradigm is applied to tip-of-the-tongue experiences (TOTs) to see if the “testing effect” holds true for previously-learned information as well. A TOT is the temporary inability to retrieve the phonology of a well-known word or name despite access to its meaning or other features. In the Study condition, half of the participants were repeatedly given the answers to TOT-inducing questions to study; in the Test condition, the other half were tested repeatedly on the same questions with no feedback. On the final test one week later, correct responses increased from the initial test, with a greater effect for participants in the Study condition than in the Test condition because Don’t Know responses decreased most in the Study condition. TOTs decreased from the first to the final test in both groups, and this effect did not vary with the condition. These preliminary results suggest that repeated retrieval attempts with no feedback and repeated studying are equally effective in preventing future TOTs for a target word. The testing effect thus does not apply in the same way to TOTs, implying that retrieval of new learning involves a different pathway than that for old learning.
Funding provided by: Fletcher Jones Foundation; Hirsch Grant

Los Gorilas in Our Midst: Bilingualism and Inattentional Blindness

Emily Wasserman (2015); Mentor(s): Deborah Burke

Abstract: Following previous results which indicate advantages for bilinguals on various non-verbal executive function tasks, including measures of attention and control, the present research aimed to investigate whether bilinguals would be less likely than monolinguals to notice the unexpected stimulus in an inattentional-blindness task. If bilinguals do possess superior attentional skills in general, it might be expected that they would show an advantage in focusing on a target stimulus, thus failing to notice the unexpected introduction of another stimulus. Each of 50 college-age subjects (29 bilingual) completed a series of executive tasks, including two measures of inattentional blindness, in one of two orders. Results showed no significant differences in noticing rates between the bilingual and monolingual groups on either task. Though conclusions are difficult to draw given the lack of between-group differences in the data, this is possibly due to a ‘ceiling effect’ in the executive performance of young adults, which would mitigate any performance differences between bilingual and monolingual groups. Future extensions of this research would therefore seek to explore any differences on these tasks between bilinguals and monolinguals within different age strata.
Funding Provided by: Fletcher Jones Foundation

Non-linguistic communication in rhesus macaque monkeys and bottlenose dolphins

Sasha Winkler (2014); Additional Collaborator(s): Brianne Beisner (UC Davis); Eliza Bliss-Moreau (UC Davis); Mentor(s): Deborah Burke

Abstract: My SURP research used behavioral biology and bioacoustics techniques to understand the meaning of specific behavioral signals in rhesus macaques and bottlenose dolphins. These species are of particular interest because of their complex social structures and communication repertoires. The goal of the first study was to understand the behavioral contexts in which silent bared teeth (SBT) displays occur in rhesus macaques and whether they truly are “fear grimaces,” as suggested in previous literature. I observed spontaneous behaviors of members of a socially-housed group of 135 rhesus macaques, using focal and event sampling. Preliminary results show that SBTs often occur in peaceful contexts and are generally given by subordinate monkeys to dominant monkeys. This suggests that SBTs may be a complex signal used in social interactions to communicate subordination, unrelated to fear. The goal of the second study was to find evidence of communication of affective states in bottlenose dolphin vocalizations. I analyzed recordings of dolphin vocalizations for differences in various acoustical parameters across high and low arousal states and positive and negative valence states. Parameters associated with emotion or arousal in humans and other mammals were examined in the dolphin recordings to look for interspecies commonalities. We are still processing the data associated with this study, but finding differences may have important consequences for captive dolphin management and welfare.
Funding provided by: Paul K. Richter and Evelyn E. Cook Richter Memorial Fund

Transfeminine Identity in Guadalajara: An Analysis of Sociolinguistic Practices

Mauricio Navarro (2014); Mentor(s): David Divita

Abstract: Transfeminine* speakers, like others of non¬heteronormative gender identities and sexualities, often employ particular sociolinguistic practices to construct and negotiate their social identities. The grammatical structure of Romance languages, such as Spanish, allows speakers unique opportunities to navigate the complex contexts presented by gender transfeminine identity and expression. Drawing on research in the emerging field of Queer Linguistics, I conducted in-depth interviews with 7 transfeminine-identified people in Guadalajara, Mexico. The subjects interviewed include transsexual women, travestis, and gender non-conforming people on the transfeminine spectrum. Most of the practices documented in the interviews revolve around the strategically fluid use of masculine and feminine grammatical gender markers, including multiple personal names, pronouns, and descriptive adjectives. I will use the data collected from these interviews to complete a fuller sociolinguistic analysis, perhaps in the form of my senior thesis.
*‘Transfeminine’ can describe a range of feminine gender identities and expressions with which people that were male-assigned at birth may align themselves.
Funding provided by: Aubrey H. and Eileen J. Seed Student Research Fund

Pupillometry in Attentional Blink

Karin Denton (2015); Mentor(s): Jesse Harris

Abstract: A growing body of research indicates that increased pupil dilation corresponds to increased cognitive load, including target identification. For instance, when subjects are instructed to detect a specific sensory signal, their pupils dilate significantly upon presentation of the target signal. We used pupillometry as a tool to investigate the attentional blink, a well-documented effect in which subjects are less likely to identify a second target (T2) if it appears 200-500ms after a previous target (T1) among several stimuli presented for 100ms each in a rapid serial visual presentation (RSVP) paradigm. Subjects (N=20) viewed RSVP streams in which number targets were dispersed among letter distractors, while the eye tracker measured pupil diameter. Half the trials were button trials, in which subjects pressed a button when they saw a target; the others were writing trials, in which subjects wrote down which targets were presented at the end of each trial. We found diminished T2 accuracy when T2 was in the attentional blink range, as expected. Unexpectedly, participants performed better on the writing task if they had completed the button task first. Over 2 million pupil measurements were taken, which are still being analyzed. We are interested in looking for pupillary dilation in response to unreported targets, which would indicate that the missed targets were identified unconsciously, and whether the more time-sensitive button task results in larger pupil dilations than the writing task.
Funding provided by: Paul K. Richter and Evelyn E. Cook Richter Memorial Fund

Making Polite Inferences in Japanese: The “N Desu” Phrase

Will Hunt (2014); Additional Collaborator(s): Kyoko Kurita; Mentor(s): Jesse Harris

Abstract: Many English speakers find Japanese difficult to study for various reasons. Writing system aside, Japanese language includes sociolinguistic elements such as politeness, which can be difficult to grasp for speakers of the majority of the world’s languages, which don’t have politeness and other usually implied language directly coded within the language. One such elusive phrase is the “N Desu” phrase, which Japanese speakers themselves often have trouble explaining to Japanese learners, who often struggle with when and how to use the phrase. It has been described as an “explanatory” sentence ending, but has many uses. This summer I have begun investigating this phrase in greater detail from a linguistic perspective in hopes to get at the heart of the phrase’s meaning and usage. Studying previous research on Japanese language and politeness and conducting a few pilot studies has lead me in several different directions, unraveling more uses of the phrase and possibilities as I continue to investigate. Drawing from Japanese cultural values placed on discourse, specifically emphasizing the importance of politeness, indirectness and excuse making, my working hypothesis is that the “N Desu” phrase is used to politely allow listeners to make appropriate inferences. In other words, the phrase’s purpose is to guide listeners to relevant information in a polite and culturally appropriate way.
Funding provided by: Pomona College SURP

Lexical Decision in Real Time

Lindsey Meyer (2014); Mentor(s): Jesse Harris

Abstract: Several models of lexical access propose a two-step dynamic process involving retrieval of (a) the visual/auditory representation of an item and (b) the semantic representation of the item. Although it is unclear how these two steps interact during lexical retrieval, many low-level factors such as frequency are known to impact ease of access. To what extent other factors impact lexical access is an active area of research. In this experiment, we studied lexical access by focusing on the effects of orthographic neighborhood size (the number of words that can be made from a target word by changing only one letter, e.g., clam is a neighbor of cram) and the presence of a high frequency neighbor on the access of words and pseudowords. Subjects performed a lexical decision task rating 96 words and pseudowords (orthographically licit strings without semantic content, e.g., smark) as real words of English or not by pressing one of two on-screen buttons while their mouse movements were tracked. Results revealed a strong inhibitory effect of neighborhood size on the mouse trajectory of pseudoword responses but few significant effects for other variable combinations. We plan to test additional subjects through the fall semester to better compensate for the small effect with a larger sample size. With further analysis, the mouse tracking data has the potential to elucidate the dynamic processing of words and pseudowords, providing insight into the two-step model of lexical access.
Funding provided by: Pomona College SURP

The Reclamation of Xinka

Rodrigo Ranero (2014); Additional Collaborator(s): COPXIG; Mentor(s): Mary Paster

Abstract: Xinka is a heavily endangered language isolate spoken in southeastern Guatemala. Since the summer of 2012, I have collaborated with the Council of the Xinka People of Guatemala (COPXIG) to improve the teaching of Xinka as a second language in regional schools, in order to ensure the preservation of Xinka for current and future generations. This summer’s work included the following: publication of a second textbook to teach Xinka (Alyamalh Xinka – Wirimalh Xinka, Módulo II: Morfología), teacher workshops in the states of Santa Rosa and Jutiapa, and the completion of the first stages of a long-term community engagement effort intended to raise awareness among Xinka community members of their linguistic rights. Additionally, two COPXIG members and I were invited to present as special guests at the X Congress of Mayan Studies in Guatemala City, which resulted in an increased awareness on the endangerment of Xinka and COPXIG’s efforts to preserve Xinka culture. Other areas of the project currently in progress include the writing of a third textbook focused on syntax, the digitization of all the pedagogical documents, the expansion of the project to the state of Jalapa, and the establishment of alliances with other Xinka organizations to address community participation in the reclamation effort. Given the enthusiastic response received thus far, we predict that this endeavor will have far reaching consequences for the preservation of Xinka.
Funding provided by: Paul K. Richter and Evelyn E. Cook Richter Memorial Fund; Davis Projects for Peace; Strauss Foundation


Emotional Faces and Cognition: the Effects of Ekman’s Emotional Expressions on Memory

Leyla Tarhan (2013); Mentor(s): Deborah Burke

Abstract: Ekman (Ekman, 1992) developed the Directed Facial Action Task (DFAT), which demonstrated that facial expressions can elicit emotions. The present study investigated whether these emotions have cognitive effects, as found when emotions are elicited in other ways. 38 participants performed the DFAT for happy and sad expressions before recalling neutral, positive and negative images. The Mood Congruent Memory hypothesis predicted that, if the DFAT produces sustained affect, participants should recall more mood-congruent than mood-incongruent images. Some participants performed the DFAT while Galvanic Skin Response, an index of emotional response, was recorded. GSR response correlated with reported mood change in the happy condition, while the difference between mood-congruent and –incongruent memory correlated with reported mood change in the sad condition. However, there is no correlation between GSR and memory. These results show that self-reported emotion but not physiological response was linked to congruency effects in memory for emotional images.
Funding provided by: Pomona College SURP

It's on the tip of my tongue: Reducing word retrieval failures through initial syllable priming

Sasha Winkler (2014); Student Collaborator(s): Evan Zahniser (2012); Mentor(s): Deborah Burke

Abstract removed upon request.

Well, This is Awkward: Taboo and Sociolinguistic Awkwardness in Arrested Development

Katherine Feller (2013); Mentor(s): Michael Diercks

Abstract: This SURP aims to investigate the phenomena of (socio)linguistic awkwardness in relation to taboo language and subject matter with a specific focus on the comedic television series Arrested Development. Through the lens of Arrested Development the project attempts to explore and define the following: sociocultural and linguistic taboos in modern American society (and their roots where relevant), the concept of awkwardness and awkward situations, sociocultural taboos and their role in creating awkwardness, and how all of these elements come together to create comedic media. The project is advanced in an interdisciplinary manner, relying on theory and literature from linguistics, sociology, anthropology, and cultural studies.
Funding provided by: Faucett Catalyst Fund (HW); Pomona College Department of Linguistics and Cognitive Science (KL)

Awkward Is the New Normal: The Awkwardness Project

Hannah Walhout (2014); Student Collaborator(s): Karen Eisenhauer (2013 PZ); Kaya LeGrand (2015); Chris Leon (2014 PZ); Mentor(s): Michael Diercks; Carmen Fought (PZ)

Abstract: The recent popularity of shows such as Arrested Development, The Office, and Curb Your Enthusiasm encodes a deeper public fascination with the experience of social awkwardness itself. To date, little formal investigation into the experience of awkwardness has been conducted. Drawing on research from disciplines such as linguistic anthropology, pragmatics, social psychology, and more recent interdisciplinary projects such as Enfield and Levinson’s Roots of Human Sociality program, we have developed a working model of awkwardness. Our summer research has produced a framework consisting of several super-categories, such as “face” and “conversation structure,” within which we have classified specific awkwardness triggers – allowing for coding and analysis of individual instances of awkwardness. We have also developed a model of the human experience of awkwardness, which describes how awkwardness plays out in the context of human interaction while taking into account sociocultural frame, interactional norms, and individual cognitive mechanisms.
Funding provided by: Evelyn B. Craddock McVicar Memorial Fund; Pomona College Department of Linguistics and Cognitive Science; Hahn Grant for Teaching with Technology

Chopping Sound: Annotation of Chichewa Texts

Martha Booker Johnson (2013); Additional Collaborator(s): Laura Downing*; Mentor(s): Mary Paster
*Zentrum für Allgemeine Sprachwissenschaft/Göteborgs Universitet

Abstract: Chichewa is a tonal Bantu language spoken by about seven million people in Malawi. I annotated two Chichewa texts that had been read aloud by four different speakers. Annotation allows researchers to attach labels to segments of a recorded sound file, making it possible to locate words or phrases needed for analysis. The most challenging aspect of the work is identifying word and phrase boundaries, as words frequently run together in a file and phrase boundaries can change between speakers. The files will be used to conduct research on the interaction between lexical tone, which is a change in pitch that distinguishes two words from one another, and intonation, which is the varying pitch throughout a phrase or sentence that can alter meaning.
Funding provided by: Oldenborg International Research and Travel Grant

Returning from the Edge: The Revitalization of Xinka

Rodrigo Ranero (2014); Mentor(s): Mary Paster
In collaboration with Consejo del Pueblo Xinka de Guatemala (COPXIG)

Abstract: The language of the non-Mayan Xinka people of Guatemala is nearly extinct. The revitalization of Xinka should be of critical importance in order for this underrepresented minority to reclaim a unique and unjustly lost cultural heritage. In cooperation with the Council of the Xinka People of Guatemala, I carried out a survey of linguistic documents and extracted the most essential aspects of the language’s grammar. With this knowledge, I carried out weekly workshops with Xinka leaders who will train language instructors at select schools in Santa Rosa. We also wrote a beginner’s guide to Xinka focusing on writing and phonology that will be published with support from the OEI. The publishing of two other documents focusing on morphology and syntax has also been funded. The community interest and organizational support for the revitalization of Xinka provides hope for the preservation of valuable linguistic knowledge for future generations.
Funding provided by: Paul K. Richter and Evalyn E. Cook Richter Memorial Funds