• hero-1.jpg
  • hero-2.jpg
  • hero-3.jpg
  • hero-4.jpg
  • hero-5.jpg
  • hero-7.jpg
  • newhero-1.jpg
  • newhero-5.jpg
  • newhero-6.jpg
  • newhero-7.jpg
  • CLS_Hero_1_Fa16.jpg
  • CLS_Hero_2_Fa16.jpg
  • CLS_Hero_3_Fa16.jpg
  • CLS_Hero_4_Fa16.jpg
  • CLS_Hero_5_Fa16.jpg

Fall 2016

Sept 9, 2016 - Isabel Deibel and John Lipski (Penn State University) 

Title: Licensing adpositions in Media Lengua: Quichua or Spanish?

AbstractThis study analyzes the distribution of adpositions in Media Lengua, a mixed language found in the northern Ecuadorian region of Imbabura and composed of mainly Quichua grammar and Spanish vocabulary (Muysken 1981). In terms of their linguistic profile, Quichua and Spanish differ fundamentally: While Spanish is a synthetic language employing head-initial prepositional phrases, Quichua is an agglutinating, postpositional language. Little consensus exists in the literature on the exact nature and realization of adpositions in the world’s languages; some have suggested that they straddle the boundary between lexicon and grammar. Since models of generative grammar describe syntactic structures as projections from the lexicon, the neat separation between lexicon and grammar found in Media Lengua can offer interesting insights into the licensing of adpositions in the context of language contact and into their status as a linguistic category. While an earlier study had described Spanish prepositions as alternating with Quichua suffixes or occurring in double constructions (Dikker 2008), the results of the current study underscore the robustness of Quichua morphosyntax and stand in direct contrast to the results found in Dikker 2008. 

A group of participants, trilingual in Quichua, Spanish and Media Lengua, participated in a video description task in Media Lengua and a translation task from Spanish or Quichua into Media Lengua. The study was conducted in the villages of Casco Valenzuela, Angla and Pijal, Imbabura, Ecuador, rendering approximately 20 minutes of recorded speech per participant. Results indicate that most adpositional phrases were headed by Quichua postpositions in fulfillment of Quichua structural requirements – even in the contexts of priming in Spanish to Media Lengua translations. Very few Spanish prepositions are found incorporated in Media Lengua in either task and they mostly occurred in frozen expressions or borrowed collocations. Among the few Spanish tokens found, most were embedded with their respective Quichua counterpart postposition. In line with Muysken’s relexification hypothesis (1981), complex adpositional phrases appeared with the Spanish preposition occupying the spot of a Quichua lexical item and hence seem to carry lexical features. However, no simple Spanish prepositions were found incorporated as postpositions, indicating that they carry grammatical instead of lexical features. 

Sept 16, 2016 - Clara Cohen (Penn State University) 

Title: Durational clues to number morphology and word structure: What are they, and when do we hear them?

Abstract: In suffixing languages, segmental information about word structure does not arrive until the end of the word, yet from the very beginning of the word, the stems contain subsegmental, durational clues about the presence of following suffixes or additional syllables. A canny listener might therefore do well to draw on those clues during speech perception, in order to make predictions and speed their understanding of what they hear. This talk will discuss ongoing work to determine what sorts of durational cues listeners draw on during online speech perception, and the role of sentence context and language structure in affecting listener sensitivity.

Sept 23, 2016 - Meredith Tamminga (University of Pennsylvania)

Title: Architectural implications of the dynamics of variation

Abstract: In this talk I discuss the proposal from Tamminga, MacKenzie and Embick (forthcoming) of a framework that recognizes three types of factors conditioning linguistic variation: sociostylistic (s-), internal linguistic (i-), and psychophysiological (p-). I elaborate on the point that p-conditioning and i-conditioning are distinct in their mental implementation, and discuss what implications this point has for a) understanding the locality of the factors conditioning alternations, b) the universality and language-specificity of variation, and c) the general question of whether grammar and language use are distinct. 

Sept 30, 2016 - Frances Blanchette (Penn State University) 

Title: Linguistic variation in English negation: Structure, meaning, and sound

Abstract: Linguistic negation is a fundamental aspect of human language and thought. In English, there exists rich variation in how negative meanings are expressed. For example, a sentence like ‘Ididn’t eat nothing’ can mean either that I ate nothing, or that it is not the case that I ate nothing.Inthistalk I discuss a series of studies that examine micro- variation in the structure, meaning, and sounds of English negative sentences. The results demonstrate how the expression and interpretation of negative sentences andwords is shaped by a complex interaction and interdependence between: (i) syntactic structure; (ii) sentence-internal (semantic) meaning; (iii) pragmatic context; (iv) prosody; and (v) social or prescriptive norms. I discuss what these combined results implicate for grammatical theories of negation and linguistic variation, as well as for the role of linguistic complexity in sentence processing, and I also discuss how the results inform our understanding of the relationship between grammar and usage. 

Oct 14, 2016 - Scott Schwenter (Ohio State University)

Title: Would you just die already? Priming and Obsolescence in Grammar

Abstract: For 30 years, priming—thetendencytore-use a given linguistic structure after its prior, typically recent, use in discourse—has been an important topic in psycholinguistics and language processing. In the last10+years it has also (under the name “persistence”) come to be a key element in corpus research. The implications of priming, however, have mainly been drawn outside the linguistic system and have been related to general cognitive principles associated with memory and processing. In this talk, I show that priming/persistence also has important implications for the linguistic system and especially for variation and change. First of all, when two or more grammatical variants are in competition, priming/persistence effects are always stronger on the obsolescing variants. Second, priming/persistence actually keeps obsolescing grammatical forms alive much longer than they would survive otherwise. Third, beyond the well-known syntagmatic effects of priming/persistence in discourse, there are also important paradigmatic results, and priming/persistence can actually lead to partial resuscitation of grammatical forms that are otherwise dying. Data from both Spanish and Portuguese will be used to illustrate these points.

Oct 14, 2016 - Gretchen Sunderman (Florida State University)

Title: The Bilingual Lexicon: Connecting Words in the Mind 

Abstract: Learning a second language (L2) necessarily entails building a new lexicon in that language. But how do L2 learners accomplish this feat given that they begin with a well-developed lexicon in their first language (L1)?  In this talk, I first describe one of the most well-known developmental models of the lexicon, Kroll and Stewart’s (1994) Revised Hierarchical Model (RHM) and its predictions related to cross-language lexical and conceptual connections, focusing specifically on the notion of concept mediation within the RHM. I then describe two psycholinguistic production studies that have investigated conceptual mediation. The first study uses the Deese-Roediger-McDermott (DRM) false memory paradigm (Deese, 1959; Roediger & McDermott, 1995) to test L2 semantic associative links. The second study investigates the nature of learners’ errors as they named pictures in the L2 under blocked and mixed presentation. The results of the combined studies suggest 1)that conceptual mediation is related to more skilled performance, but comes at a cost to the L1 and 2) that control of spoken production may be affected by proficiency as well as individual differences in the ability to allocate cognitive resources.  Finally, applications of the RHM for learning L2 vocabulary are presented. 

Oct 21, 2016 - Patricia Schempp (Penn State University)

Title: L2 Learners’ Processing of Grammatical Gender Varies According to Cognate Status and Proficiency: An ERP study

Abstract: Previous studies highlight the difficulty of mastering L2 gender (e.g., Hopp 2013), however recent ERP evidence suggests that after training, L1 English speakers can exhibit native-like ERPs to gender violations (e.g., Morgan-Short et al., 2012).  This study extends Morgan-Short et al.'s (2012) work on artificial language learning to natural language, investigating whether classroom-based L2 German learners are sensitive to gender violations in L2 German, before and after training, and whether advanced L2 German learners are sensitive to these same violations.  Additionally, it investigates L2 noun-gender mappings for cognates versus non-cognates, broadening research showing that L2 speakers process cognates faster than non-cognates (see Schwartz & van Hell, 2012).  In an ERP task, twenty-three intermediate L2 German learners and 19 advanced L2 German learners read target questions and then answered the question by choosing the appropriate picture onscreen. The target questions were grammatically correct or incorrect, and contained cognate or noncognate nouns. Next, intermediate participants were trained offline to high accuracy with the nouns and their gender using picture naming. One week later they were tested on their vocabulary and gender accuracy offline, retrained, and then repeated the ERP task. ERP analyses timelocked to the critical noun phrase revealed that before training, intermediate learners exhibited no significant ERP responses to grammatical gender violations.  After training, intermediate participants exhibited an N400 effect for gender violations with cognates only. At 500-900ms there was a frontal positivity for gender violations with cognates, and at 900-1100ms there was a frontal positivity for gender violations with cognates and non-cognates. Analysis of the picture naming test prior to retraining reveals that cognates were easier to recall, but gender accuracy did not vary with cognate status. These results suggest that for intermediate learners cognate effects extend to the processing of L2 morphosyntactic features, whereby cognates facilitate the online processing and retrieval of nouns and associated grammatical information. In contrast, advanced learners exhibited sensitivity to gender violations for noncognates only, suggesting that there is a fundamental difference in the processing of gender for cognates and noncognates among L2 learners, that may vary as a function of proficiency level.

Oct 28, 2016 - Federica Bulgarelli (Penn State University)

Title: Double trouble: Statistical learning of multiple structures

Abstract: How do naïve learners come to identify the number of languages they are learning?  This question is central to the study of language acquisition, but to date our understanding is far from complete. One means of approaching this problem is through the study of statistical learning, the process by which learners track rudimentary distributional information from their sensory input. For the past two decades, research has established that statistical learning is particularly critical for early language acquisition, allowing learners of all languages to gain a foothold into acquisition from which language-specific properties emerge. Research in our lab aims at broadening the scope of statistical learning tasks to understand how statistical learning might operate when learners are exposed to multiple underlying structures, arguably more closely approximating bilingual language acquisition. During this talk, I will discuss previous and ongoing studies that have focused on how learners across the lifespan and from different linguistic backgrounds detect shifts in patterns of speech streams as well as contend with multiple statistical regularities and rules.

Nov 4, 2016 - Matt Carlson and Alex McAllister (Penn State University) 

Title: Phonological repair of initial /s/-consonant sequences in speech perception and production

Abstract: Spanish phonotactics prohibits word-initial /s/-consonant (#sC) clusters, repairing them as needed by prepending an initial /e/, e.g. in loanwords such as esnob ‘snob’. This process is easily described in any of the available phonological frameworks, but in an age where much of grammar is turning out to be gradient to some degree, it is not yet clear what contributes to such apparently stable and consistent patterns as this. One possible source of stability comes from speech perception: recent evidence suggests that [e] is so likelytoprecedeansC sequence, that Spanish speakers tend to hear it even when it is not there (and that they do not tend to hear other vowels in this context) (Cuetos, Hallé, Domínguez, & Segui, 2011; Hallé, Dominguez, Cuetos, & Segui, 2008; cf. Dupoux, Kakehi, Hirose, Pallier, & Mehler, 1999 on Japanese). We show using nonword discrimination and lexical decision data that this is indeed the case, but that under certain circumstances, listeners can respond as if other vowels are present. Our results support an abstract process whereby [e] isinsertedbeforesC sequences in Spanish, but they also suggest that phonetic details in the signal can shape this process, details that may be related to reduction processes in speech production (e.g. Davidson, 2006; Munson, 2001; Van Son & Pols, 2003).We therefore pursue two hypotheses in a speech production task. First, we ask whether word-initial [e]precedingsC is in fact so predictable that speakers leave it out (such that Spanish speakers may produce and hear examples of "illicit" #sC clusters in natural Spanish speech), and that the reduction of articulatory gestures may nonetheless preserve sufficient acoustic detail to support identification of certain vowels in this position. 

Nov 7, 2016 - John McWhorter (Columbia University)

Title: The missing Spanish creoles are still missing: revisiting afrogenesis and its implications for a coherent theory of creole genesis

Abstract: Theories that plantation creoles were all born as pidgins at West African coast slave castles, including that proposed in McWhorter (2000), have not fared well among creolists, amidst a preference for supposing that creoles are born, or not, according to factors local to a given context.In this paper I review some of the responses to McWhorter (2000) and spell out why, especially in light of research since, the “Afrogenesis” paradigm is still worth serious consideration. A key fact is the following. Many creolists argue that a creole did not appear when there was extensive black-white contact and many slaves were locally-born, a scenario most often associated with the Spanish Caribbean and Reunion and now proposed for South American colonies by Sessarego (2014) and Díaz-Campos & Clements (2008). However, conditions were of just this kind in early St. Kitts and Barbados, where most scholars now locate the birth of English-based and French-based plantation creoles. The disparity in outcomes between these locations means that after fifty years, there is no coherent theory of how or why creoles come to be. I argue that only Afrogenesis shows the way out of this conundrum.

Dec 2, 2016 - Laurel Brehm (Penn State University)

Title: Distinguishing Discrete and Gradient Category Structure in Language

Abstract: Work in cognitive psychology underscores the probabilistic (gradient) nature of mental classes, but traditional linguistic analysis rests upon the discrete separation of classes. I present work that uses memory errors to examine the mental representation of verb-particle constructions (VPCs, e.g., *make up* the story, *cut up* the meat). VPCs are diverse in terms of their semantic and syntactic properties; an outstanding question is how this variability connects with the class structure in the mental representation of VPCs. To experimentally examine this question, I present a novel paradigm that elicits illusory conjunctions of sentence elements-- memory errors that are sensitive to linguistic structure. Applying piecewise regression on these error data demonstrates that illusory conjunctions of verbs and particles follow a graded cline rather than discrete classes, supporting the presence of gradience in the mind's representation of linguistic elements.

Dec 9, 2016 - Chaleece Sandberg (Penn State University)

Title: Development of a Theoretically-Based Culturally-Relevant Therapy for Anomia in Bilingual Aphasia

Abstract: Training abstract word retrieval improves generative naming for trained items, and promotes generalization to generative naming of concrete words in the same context category. However, this therapy has not yet been adapted for bilingual persons with aphasia (PWA). This study aimed first to develop culturally and linguistically relevant stimuli for the extension of this therapy to bilingual PWA, and second to conduct this therapy with a test case. Pertinent differences regarding cultural relevance and linguistic content of several context-categories across three languages will be discussed as will the results of the test case.

Sept 30, 2016 - Frances Blanchette (Penn State University) 

Title: Linguistic variation in English negation: Structure, meaning, and sound

Abstract: Linguistic negation is a fundamental aspect of human language and thought. In English, there exists rich variation in how negative meanings are expressed. For example, a sentence like ‘Ididn’t eat nothing’ can mean either that I ate nothing, or that it is not the case that I ate nothing.Inthistalk I discuss a series of studies that examine micro- variation in the structure, meaning, and sounds of English negative sentences. The results demonstrate how the expression and interpretation of negative sentences andwords is shaped by a complex interaction and interdependence between: (i) syntactic structure; (ii) sentence-internal (semantic) meaning; (iii) pragmatic context; (iv) prosody; and (v) social or prescriptive norms. I discuss what these combined results implicate for grammatical theories of negation and linguistic variation, as well as for the role of linguistic complexity in sentence processing, and I also discuss how the results inform our understanding of the relationship between grammar and usage. 

Oct 14, 2016 - Scott Schwenter (Ohio State University)

Title: Would you just die already? Priming and Obsolescence in Grammar

Abstract: For 30 years, priming—thetendencytore-use a given linguistic structure after its prior, typically recent, use in discourse—has been an important topic in psycholinguistics and language processing. In the last10+years it has also (under the name “persistence”) come to be a key element in corpus research. The implications of priming, however, have mainly been drawn outside the linguistic system and have been related to general cognitive principles associated with memory and processing. In this talk, I show that priming/persistence also has important implications for the linguistic system and especially for variation and change. First of all, when two or more grammatical variants are in competition, priming/persistence effects are always stronger on the obsolescing variants. Second, priming/persistence actually keeps obsolescing grammatical forms alive much longer than they would survive otherwise. Third, beyond the well-known syntagmatic effects of priming/persistence in discourse, there are also important paradigmatic results, and priming/persistence can actually lead to partial resuscitation of grammatical forms that are otherwise dying. Data from both Spanish and Portuguese will be used to illustrate these points.

Oct 14, 2016 - Gretchen Sunderman (Florida State University)

Title: The Bilingual Lexicon: Connecting Words in the Mind 

Abstract: Learning a second language (L2) necessarily entails building a new lexicon in that language. But how do L2 learners accomplish this feat given that they begin with a well-developed lexicon in their first language (L1)?  In this talk, I first describe one of the most well-known developmental models of the lexicon, Kroll and Stewart’s (1994) Revised Hierarchical Model (RHM) and its predictions related to cross-language lexical and conceptual connections, focusing specifically on the notion of concept mediation within the RHM. I then describe two psycholinguistic production studies that have investigated conceptual mediation. The first study uses the Deese-Roediger-McDermott (DRM) false memory paradigm (Deese, 1959; Roediger & McDermott, 1995) to test L2 semantic associative links. The second study investigates the nature of learners’ errors as they named pictures in the L2 under blocked and mixed presentation. The results of the combined studies suggest 1)that conceptual mediation is related to more skilled performance, but comes at a cost to the L1 and 2) that control of spoken production may be affected by proficiency as well as individual differences in the ability to allocate cognitive resources.  Finally, applications of the RHM for learning L2 vocabulary are presented. 

Oct 21, 2016 - Patricia Schempp (Penn State University)

Title: L2 Learners’ Processing of Grammatical Gender Varies According to Cognate Status and Proficiency: An ERP study

Abstract: Previous studies highlight the difficulty of mastering L2 gender (e.g., Hopp 2013), however recent ERP evidence suggests that after training, L1 English speakers can exhibit native-like ERPs to gender violations (e.g., Morgan-Short et al., 2012).  This study extends Morgan-Short et al.'s (2012) work on artificial language learning to natural language, investigating whether classroom-based L2 German learners are sensitive to gender violations in L2 German, before and after training, and whether advanced L2 German learners are sensitive to these same violations.  Additionally, it investigates L2 noun-gender mappings for cognates versus non-cognates, broadening research showing that L2 speakers process cognates faster than non-cognates (see Schwartz & van Hell, 2012).  In an ERP task, twenty-three intermediate L2 German learners and 19 advanced L2 German learners read target questions and then answered the question by choosing the appropriate picture onscreen. The target questions were grammatically correct or incorrect, and contained cognate or noncognate nouns. Next, intermediate participants were trained offline to high accuracy with the nouns and their gender using picture naming. One week later they were tested on their vocabulary and gender accuracy offline, retrained, and then repeated the ERP task. ERP analyses timelocked to the critical noun phrase revealed that before training, intermediate learners exhibited no significant ERP responses to grammatical gender violations.  After training, intermediate participants exhibited an N400 effect for gender violations with cognates only. At 500-900ms there was a frontal positivity for gender violations with cognates, and at 900-1100ms there was a frontal positivity for gender violations with cognates and non-cognates. Analysis of the picture naming test prior to retraining reveals that cognates were easier to recall, but gender accuracy did not vary with cognate status. These results suggest that for intermediate learners cognate effects extend to the processing of L2 morphosyntactic features, whereby cognates facilitate the online processing and retrieval of nouns and associated grammatical information. In contrast, advanced learners exhibited sensitivity to gender violations for noncognates only, suggesting that there is a fundamental difference in the processing of gender for cognates and noncognates among L2 learners, that may vary as a function of proficiency level.

Oct 28, 2016 - Federica Bulgarelli (Penn State University)

Title: Double trouble: Statistical learning of multiple structures

Abstract: How do naïve learners come to identify the number of languages they are learning?  This question is central to the study of language acquisition, but to date our understanding is far from complete. One means of approaching this problem is through the study of statistical learning, the process by which learners track rudimentary distributional information from their sensory input. For the past two decades, research has established that statistical learning is particularly critical for early language acquisition, allowing learners of all languages to gain a foothold into acquisition from which language-specific properties emerge. Research in our lab aims at broadening the scope of statistical learning tasks to understand how statistical learning might operate when learners are exposed to multiple underlying structures, arguably more closely approximating bilingual language acquisition. During this talk, I will discuss previous and ongoing studies that have focused on how learners across the lifespan and from different linguistic backgrounds detect shifts in patterns of speech streams as well as contend with multiple statistical regularities and rules.

Nov 4, 2016 - Matt Carlson and Alex McAllister (Penn State University) 

Title: Phonological repair of initial /s/-consonant sequences in speech perception and production

Abstract: Spanish phonotactics prohibits word-initial /s/-consonant (#sC) clusters, repairing them as needed by prepending an initial /e/, e.g. in loanwords such as esnob ‘snob’. This process is easily described in any of the available phonological frameworks, but in an age where much of grammar is turning out to be gradient to some degree, it is not yet clear what contributes to such apparently stable and consistent patterns as this. One possible source of stability comes from speech perception: recent evidence suggests that [e] is so likelytoprecedeansC sequence, that Spanish speakers tend to hear it even when it is not there (and that they do not tend to hear other vowels in this context) (Cuetos, Hallé, Domínguez, & Segui, 2011; Hallé, Dominguez, Cuetos, & Segui, 2008; cf. Dupoux, Kakehi, Hirose, Pallier, & Mehler, 1999 on Japanese). We show using nonword discrimination and lexical decision data that this is indeed the case, but that under certain circumstances, listeners can respond as if other vowels are present. Our results support an abstract process whereby [e] isinsertedbeforesC sequences in Spanish, but they also suggest that phonetic details in the signal can shape this process, details that may be related to reduction processes in speech production (e.g. Davidson, 2006; Munson, 2001; Van Son & Pols, 2003).We therefore pursue two hypotheses in a speech production task. First, we ask whether word-initial [e]precedingsC is in fact so predictable that speakers leave it out (such that Spanish speakers may produce and hear examples of "illicit" #sC clusters in natural Spanish speech), and that the reduction of articulatory gestures may nonetheless preserve sufficient acoustic detail to support identification of certain vowels in this position. 

Nov 7, 2016 - John McWhorter (Columbia University)

Title: The missing Spanish creoles are still missing: revisiting afrogenesis and its implications for a coherent theory of creole genesis

Abstract: Theories that plantation creoles were all born as pidgins at West African coast slave castles, including that proposed in McWhorter (2000), have not fared well among creolists, amidst a preference for supposing that creoles are born, or not, according to factors local to a given context.In this paper I review some of the responses to McWhorter (2000) and spell out why, especially in light of research since, the “Afrogenesis” paradigm is still worth serious consideration. A key fact is the following. Many creolists argue that a creole did not appear when there was extensive black-white contact and many slaves were locally-born, a scenario most often associated with the Spanish Caribbean and Reunion and now proposed for South American colonies by Sessarego (2014) and Díaz-Campos & Clements (2008). However, conditions were of just this kind in early St. Kitts and Barbados, where most scholars now locate the birth of English-based and French-based plantation creoles. The disparity in outcomes between these locations means that after fifty years, there is no coherent theory of how or why creoles come to be. I argue that only Afrogenesis shows the way out of this conundrum.

Dec 2, 2016 - Laurel Brehm (Penn State University)

Title: Distinguishing Discrete and Gradient Category Structure in Language

Abstract: Work in cognitive psychology underscores the probabilistic (gradient) nature of mental classes, but traditional linguistic analysis rests upon the discrete separation of classes. I present work that uses memory errors to examine the mental representation of verb-particle constructions (VPCs, e.g., *make up* the story, *cut up* the meat). VPCs are diverse in terms of their semantic and syntactic properties; an outstanding question is how this variability connects with the class structure in the mental representation of VPCs. To experimentally examine this question, I present a novel paradigm that elicits illusory conjunctions of sentence elements-- memory errors that are sensitive to linguistic structure. Applying piecewise regression on these error data demonstrates that illusory conjunctions of verbs and particles follow a graded cline rather than discrete classes, supporting the presence of gradience in the mind's representation of linguistic elements.

Dec 9, 2016 - Chaleece Sandberg (Penn State University)

Title: Development of a Theoretically-Based Culturally-Relevant Therapy for Anomia in Bilingual Aphasia

Abstract: Training abstract word retrieval improves generative naming for trained items, and promotes generalization to generative naming of concrete words in the same context category. However, this therapy has not yet been adapted for bilingual persons with aphasia (PWA). This study aimed first to develop culturally and linguistically relevant stimuli for the extension of this therapy to bilingual PWA, and second to conduct this therapy with a test case. Pertinent differences regarding cultural relevance and linguistic content of several context-categories across three languages will be discussed as will the results of the test case.