Relations in Relativity: New Perspectives on Language and Thought -

Poster abstracts




The role of language-specificity in the language production model, G. van Beek


For a long time, it has been assumed that at least a certain part of the human linguistic competence is universal (i.e., universalism). However, based on evidence of the influence of language-specific features on cognition, it is to an increasing extent debated whether this view is true. The linguists Sapir (1921, 1949) and Whorf (1949, 1956) stated that thought is determined by language (i.e., linguistic determinism). Slobin (1996) weakened this hypothesis by formulating the principle of ‘thinking for speaking’, indicating that language functions as a guide of thought, to which the principle ‘seeing for speaking’ has recently been added, indicating that in visual perception more attention is paid to what is encoded in the specific language involved than to what is not encoded in the specific language involved (i.e., linguistic relativity; Carroll, Von Stutterheim, & Nüse, 2004).
The evidence of the influence of language on cognition gave rise to the question whether universal elements in language exist at all and if they exist in which elements they precisely can be found and at what stage of the language production process model as formulated by Levelt (1989) language-specificity starts to exert an influence. Studies of second language acquisition have been done in order to formulate an answer to this question by comparing the information structures that people use in their first language with the information structures that people use when speaking a foreign language. In that way, an overview can be created of the type of information structures that speakers of different languages use regardless of the specific language that they speak and of the type of information structures that are typical for a certain language and therefore difficult to acquire for second language speakers. Studies of the use of the progressive aspect in Dutch, English, and German show under certain circumstances a transfer of L1 information structures to the L2, whereas in other situations the influence of the information structures in the L1 on the information structures in the L2 seems to be less strong. Since the findings with regard to a transfer of information structures are ambiguous, more research is needed with regard to the information structures that people use in their mother tongue and in foreign languages.
Therefore, the present study aims at providing clearer insights into language-specific features in information structures by combining a literature review concerning the debate about language-specificity and its role in the language production process model with a meta-analysis concerning data with regard to information structures that L1 Dutch speakers of German and English use compared to the information structures that L2 English and German speakers of Dutch use, specifically with regard to the use of the progressive aspect and the extent to which the endpoint is mentioned when describing motion events (Flecken, 2010, Van Ierland, 2010). It is shown that the language production process is not only a linear process consisting of the consecutive stages of conceptualization, formulation, and articulation, but that the verbal output of the articulator stage influences in turn the cognitive conceptual input of the process and that the language production process in that way can be seen as a circular cognitive-linguistic interaction model. By proposing this perspective on the language production process model, a new perspective with regard to (the role of language-specificity in) the language production process model has been developed.


Conceptual Event Units of Putting and Taking in Two Unrelated Languages, Rebecca Defina and Asifa Majid


People automatically chunk ongoing dynamic events into discrete conceptual event units. This paper investigates one possible factor in this process, i.e., linguistic structure. This paper tests the claim that describing an event with a serial verb construction will influence a speaker's conceptual event structure. The grammar of Avatime (a Kwa language spoken in Ghana) requires its speakers to describe some, but not all, placement events using a serial verb construction which also encodes the preceding taking event. We tested Avatime and English speakers’ recognition memory for putting and taking events. Avatime speakers were more likely to falsely recognize putting and taking events from episodes associated with take-put serial verb constructions than from other episodes associated with other constructions. English speakers showed no difference in false recognitions between episode types. This demonstrates that memory for episodes is related to the type of language used; and, moreover, across languages different conceptual representations are formed for the same physical episode, paralleling habitual linguistic practices.


Language shapes mental representations of musical pitch, Sarah Dolscheid, Shakila Shayan, Asifa Majid and Daniel Casasanto


Speakers often use spatial metaphors to talk about musical pitch (e.g., a low note, a high soprano). Previous experiments suggest that English speakers also think about pitches as high or low in space, even when they’re not using language or musical notation (Casasanto, 2010). Do metaphors in language merely reflect pre-existing associations between space and pitch, or might language also shape these non-linguistic metaphorical mappings?
To investigate the role of language in pitch representation, we conducted a pair of non-linguistic space-pitch interference experiments in speakers of two languages that use different spatial metaphors. Dutch speakers usually describe pitches as ‘high’ (hoog) and ‘low’ (laag). Farsi speakers, however, often describe high-frequency pitches as ‘thin’ (naazok) and low-frequency pitches as ‘thick’ (koloft) (Shayan, Ozturk, & Sicoli, 2011).
Do Dutch and Farsi speakers mentally represent pitch differently? To find out, we asked participants to reproduce musical pitches that they heard in the presence of irrelevant spatial information (i.e., lines that varied either in height or in thickness). For the Height Interference experiment, horizontal lines bisected a vertical reference line at one of nine different locations. For the Thickness Interference experiment, a vertical line appeared in the middle of the screen in one of nine thicknesses. In each experiment, the nine different lines were crossed with nine different pitches ranging from C4 to G#4 in semitone increments, to produce 81 distinct trials.
If Dutch and Farsi speakers mentally represent pitch the way they talk about it, using different kinds of spatial representations, they should show contrasting patterns of cross-dimensional interference: Dutch speakers’ pitch estimates should be more strongly affected by irrelevant height information, and Farsi speakers’ by irrelevant thickness information.
As predicted, Dutch speakers’ pitch estimates were significantly modulated by spatial height but not by thickness. Conversely, Farsi speakers’ pitch estimates were modulated by spatial thickness but not by height (2x2 ANOVA on normalized slopes of the effect of space on pitch: F(1,79)=10.73, MSE=.29, p=.002 ).
To determine whether language plays a causal role in shaping pitch representations, we conducted a training experiment. Native Dutch speakers learned to use Farsi-like metaphors, describing pitch relationships in terms of thickness (e.g., a cello sounds ‘thicker’ than a flute). After training, Dutch speakers showed a significant effect of Thickness Interference in the non-linguistic pitch reproduction task, similar to native Farsi speakers: on average, pitches accompanied by thicker lines were reproduced as lower in pitch (effect of thickness on pitch: Slope=-1.46, p=.003). By contrast, when participants received the same amount of linguistic training with a height-pitch mapping, they showed no effect of Thickness Interference (Slope=-0.08, ns).
Language thus plays a causal role in shaping nonlinguistic representations of musical pitch.


Language modulates perception differently in the left and right hemifield, Jolien Francken, Peter Hagoort and Floris P. de Lange


Language does not function as an isolated cognitive system, but is actually tightly linked with perceptual and motor functions. In this study we ask whether words that convey motion can interact with the visual perception of motion. Subjects (n=22, right‐handed) were primed with motion words (e.g., rise) that could be congruent, incongruent or neutral with regard to a visual motion stimulus that was presented briefly after the word. Subjects performed a motion detection task on the visual motion stimulus, while ignoring the words. Crucially, the motion stimuli were presented either in the right visual field (RVF) or in the left visual field (LVF), to examine whether words processed by the lefthemispheric language system would exert a stronger influence on the subsequent motion stimuli when presented in the RVF compared to the LVF. Motion words that were congruent with visual motion both speeded up reaction times and increased performance (d‐prime) compared to incongruent words, but only for RVF visual motion. A congruency effect for perceptual bias (negative criterion shift) was only present in the LVF. Next, we tested whether these effects were merely the result of conceptual priming of the motion direction. Therefore, we substituted the motion words by arrow shapes while keeping the design identical. This resulted in equally strong congruency effects in RTs, d‐prime and criterion for the LVF and RVF. These experiments show that motion words influence motion perception in the RVF and LVF in different ways. Moreover, we show this is not just the result of conceptual priming but a language‐dependent effect.



What learners mean: Cross-linguistic Influence in Danish L2 Motion Event Categorization, Moiken Jessen

The present study examines the semantic categories in the motion domain by adult learners of Danish from two typologically different L1s - Turkish and German, and compares their categorization to that of the Turkish and German native speakers. In contrast to previous L2 studies which have tended to use retellings, narratives or picture description tasks, this study uses as stimuli a series of 39 videos depicting human and animal locomotion of various kinds. Following the line of investigation employed by Majid et al. (2007) on the study of L1 semantic categorization in closely related languages, we use the statistical technique of cluster analysis in order to group scenes together based on their similarity (determined by shared verbs per scene).
The analysis of the data examines the nature of semantic categories in the learner language, i.e. their boundaries, and their relationship to one another, and compares the learners’ categorization to that of the native speakers of all the L1s involved. The results of the study show that the meanings assigned by the two learner groups to the Danish motion verbs ‘walk’, løbe ‘run’ and kravle ‘crawl’ are influenced by the semantic categorization of their respective L1s. In addition, there is evidence that typological membership might play a role with respect to the processes involved in the restructuring of meaning/conceptual categories. These processes involve both conceptual transfer from the learners’ L1 to the L2 as well as conceptual convergence (Jarvis and Pavlenko, 2008).


The role of contextual factors in the use of demonstratives: Differences between Turkish and Dutch, David Peeters and Asli Ozyurek


An important feature of language is that it enables human beings to refer to entities, actions and events in the external world. In everyday interaction, one can refer to concrete entities in the extra-linguistic physical environment of a conversation by using demonstratives such as this and that. The choice of which demonstrative to use has been explained both in terms of the distance of the referent [1], and in terms of contextual factors such as joint attention [2][3]. These claims have never been tested in a controlled setting and across different languages. Therefore, we tested demonstrative choice in a controlled elicitation task in two languages that previously have only been studied observationally: Turkish and Dutch.
In our study, twenty-nine Turkish and twenty-four Dutch participants were presented with pictures including a speaker, an addressee and an object (the referent). They were asked which demonstrative they would use in the depicted situations. We found that, in addition to the distance of the referent, the focus of attention of the addressee on the referent and the type of sentence in which a demonstrative was used, influenced demonstrative choice in Turkish. In Dutch, only the distance of the referent and the sentence type influenced demonstrative choice.
Our cross-linguistic findings show that in different languages, people take into account both similar and different aspects of triadic situations to select a demonstrative. The controlled study of referring acts in triadic situations is a valuable extension to observational research, in that it gives us the possibility to look more specifically into the interplay between language, attention, and other contextual factors influencing how people refer to entities in the world.


Exposing L2 Acquisition Costs of Typological Differences, Job Schepens, Frans van der Slik and Roeland van Hout


The L1 acquisition device provides a homogeneous mechanism for establishing communication with speakers of the same language. In comparison, the mechanism of the L2 acquisition device for establishing communication with speakers of a different language is more heterogeneous, which might have to do with the availability of an L1. Measuring the effect of the mother tongue on L2 acquisition may be useful to understand the relative dependence between language and its function in the mind, either as an L1 or as an L2. We study the dependency between language use and its linguistic features by exposing L2 acquisition costs of typological differences between the L1 and L2. We assume that  acquisition cost is the result of a distance effect based on the degree of difference between the typological characteristics of Dutch and the specific mother tongue of the learner. Computing acquisition costs of specific linguistic properties might give us new insights into the relation of linguistic differences to cognitive processes.
We quantify linguistic effects on L2 acquisition cost using a large state data base of L2 test scores of immigrants who are learning to communicate within the Dutch linguistic niche. The Dutch niche has always been relatively small although Dutch has been used as a second language considerably.

Acquiring the sounds of your language: taking statistics or doing semantics?, Sophie ter Schure


Infants learn the sounds of their language between 6 and 10 months, before they have a productive lexicon (Kuhl et al., 1992; Werker and Tees, 1984). But how do English infants learn that /æ/ and /E/ are different sounds, while Dutch infants recognize them as belonging to one category, when they have no lexicon to inform them that indeed /pEn/ and /pæn/ mean something different in English, but not in Dutch?
Work by Maye and colleagues (2002) suggests that infants keep track of distributions of distinguishing sound characteristics. For example, /E/ and /æ/ can be distinguished by their first formant. If the input contains two frequency peaks in that specific continuum of formant values, infants will create two vowel categories. If there is one frequency peak, infants will only learn one category.
Although similar ‘statistical bookkeeping’ has been found for infants in other domains (e.g. Saffran et al., 1996; Plunkett et al., 2008), failure to replicate these results for sound learning means that researchers have returned to the idea of lexical learning of sound categories: infants learn to distinguish two sounds if they occur with different referents. This poster contains a proposal to disentangle the two options, by testing distributional vs. lexical learning of two sounds with the same experimental paradigm.


Japanese adults say /pata/, but hear /tapa/, Sho Tsuji, Nayeli Gonzalez-Gomez, Victoria Medina, Thierry Nazzi and Reiko Mazuka


The labial-coronal effect, originally described as a bias to initiate a word with a labial consonant-vowel-coronal consonant (LC) sequence (e.g., /pat/), has been found in the inventories of a majority of languages (MacNeilage & Davis, 2000), infant (MacNeilage & Davis, 2000) and adult (Rochet-Capellan & Schwartz, 2007) speech production, as well as  in adult (Sato, Vallee, Schwartz, & Rousset, 2007) and infant (Nazzi, Bertoncini, & Bijeljac-Babic, 2009) speech perception. The prevalence of this bias has been explained with constraints on the human speech production system, and its perceptual correlates have motivated the suggestion of a perception-production link (Sato et al., 2007), both assuming a universal cause for this bias rather than an influence of ambient language. However, all of the above-mentioned studies were conducted in populations whose native languages have an LC bias. The current study therefore examined speakers of Japanese, a language that has been claimed to possess more CL than LC sequences. After conducting an analysis of Japanese corpora, we identified a subgroup of consonants (plosives) exhibiting a CL bias and thus qualifying this claim. Focusing on this subgroup of consonants, we found diverging results for production and perception such that Japanese speakers exhibited an articulatory LC bias, but a perceptual CL bias. We thus find support for a universal, articulatory motivated LC bias in production, supporting a motor explanation of the LC effect, while perceptual biases are influenced by distributional frequencies of the native language.


Perceptual symbols activated by metaphoric expressions, Lisanne van Weelden, G. Salami, and J. Schilperoord


Barsalou (1999) argued that perceptual representations are the building blocks of cognition. Stanfield and Zwaan (2001) and Zwaan, Stanfield, and Yaxley (2002) showed that perceptual symbols are activated through language comprehension indeed, and that they contain information about the implied orientation and shape of the concepts.
The purpose of the present study was to investigate perceptual symbols activated by metaphoric expressions. The linguistic construction for a metaphoric expression ('X is (like) Y') implies a comparison between the target (X) and the source (Y) concept and stimulates us look for correspondences between them. In doing so, perceptual symbols of both the target and the source concepts are activated. We expect that 'is (like)' does not only encourage looking for conceptual similarities between the concepts, but also for perceptual similarities. Hence, we expect the linguistic construction of metaphor to influence ‘thought’ in the sense that the compared concepts activate perceptual symbols which show perceptual similarities. As perceptual symbols are typically schematic, we expect similarities in the outline of both symbols.
In our experiment, subjects were presented with sentences describing a conceptual correspondence between two concepts from different domains of knowledge, such as 'Deze motor is heel snel, net als een luipaard' (‘This motorcycle is very fast, just like a leopard'). The sentences were followed by two line drawings of the mentioned objects. The two drawings had either a similar or dissimilar outline. Subjects made recognition responses as to whether both objects were mentioned in the sentence. The analysis showed that responses were faster when the two line drawings had a similar outline as compared to when they had a dissimilar outline. This result adds to previous evidence about how language shapes visual representations and shows that in metaphor comprehension perceptual symbols with a similar outline are activated.


Forgetting the birth language: Birth language attrition in Chinese adoptees in the Netherlands, Wen Cui Zhou and Mirjam Broersma


As commonly acknowledged, the sound structure of the native language has strong effects on speech perception, a part of the cognitive system. But it is still unclear how long-lasting the native language effects may be on phonological perception and production at a later stage. The current study attempts to explore this issue by investigating perception and production of Chinese speech by  Chinese adoptees in the Netherlands. Previous studies with international adoptees report no evidence of any retention of the birth language by the time adoptees have reached adulthood (Pallier et
al., 2003, Cerebral Cortex, 13, 155-161; Ventureya et al, 2004, Journal of Neurolinguistics, 17, 79-91). Adoptees do, however, perform better when relearning their birth language than novice learners (Hyltenstam et al., 2009, Bilingualism: Language and Cognition, 12, 121–140). This project explores the time course of forgetting the birth language during childhood, examining the effect of age at the time of adoption (AoA) and of length of residence in the Netherlands (LoR) (i.e., time elapsed since adoption). AoA is varied from one to five years, and LoR from one to four years. Chinese adoptees in the
Netherlands between four and nine years old and age-matched Dutch control participants are tested on perception and production of Chinese phonemes and tones before, during, and after extensive training. One group of Mandarin adoptees and one group of Dutch control participants are tested and trained on one Mandarin tone
contrast (tone 2, low-high, vs. tone 3, high-low-high) and one Mandarin retroflex affricate contrast ([tßH] vs. [tß]). One group of Cantonese adoptees and another group of Dutch control participants are tested and trained on one Cantonese tone contrast (tone 2, mid-high, vs. tone 5, low-mid) and one Cantonese dental affricate contrast ([tsH] vs. [ts]). Training and tests consist of XAB discrimination tasks, presented in the form of animated video games.
This study will provide further insights in the attrition of the birth language by investigating changes in birth language processing during childhood soon after adoption, and by investigating how AoA and LoR affect retention, attrition, and relearning of the birth language in international adoptees.

Last checked 2012-05-06 by Ewelina Wnuk

Street address
Wundtlaan 1
6525 XD Nijmegen
The Netherlands

Mailing address
P.O. Box 310
6500 AH Nijmegen
The Netherlands

Phone:   +31-24-3521911
Fax:        +31-24-3521213

Public Outreach Officer
Charlotte Horn