(How) Does language support the development of an independent symbolic number system?
Long decades of research have unanimously shown that the linguistic environment plays a vital role in various cognitive development, including the acquisition of numerical skills. Multicultural and multilingual learning environments, where children face the challenge of learning numbers and performing complex math in more than one language or a language different from their mother tongue are particularly concerned with this.
Because languages differ substantially in terms of grammatical and morphological rules, as well as their number systems (e.g., French vs German number words), the acquisition of numerical concepts may become demanding for polyglot children, leading to suboptimal math performance and overall low numeracy skills later in life. Unfortunately, the latter can have serious consequences for both the individuals (e.g., lower socioeconomic status, poor medical decision making) and the state they live in (e.g., low numeracy competencies lead to lower GDP).
To remedy this issue, it is important to identify the crucial factors underlying the early stages of numerical development. In the current project, we focus on the role of language in the very early stages of symbolic number (Arabic numerals, number words) acquisition. The project investigates whether and how the language-specific recursive and combinatorial rules (e.g., statistical regularities, grammatical rules), as well as the amount of linguistic exposure (e.g., mono vs multilingual), are involved in the early stages of number acquisition. An additional goal of the current project is to design and evaluate a novel training protocol to foster the development of the number system in young children. Providing such insights into the foundational role the language plays in the early stages of numerical is essential for the development of school curricula and/or proper intervention programs aiming at improving the numerical competencies of people living in a multilingual society.
Main Researchers Involved: Mila Marinova, Christine Schiltz
Duration: 2021 - 2024
Understanding the role of spatial number representations and spatial skills in
mathematical abilities in young children
Mathematical cognition is tightly linked to spatial processing. This is apparent even when considering the most basic aspect of mathematics, namely the spatial representation of numerical magnitude on a mental number line. This has led to the influential idea that spatial number representations play a foundational role in the development of mathematical abilities. Yet, until now evidence is lacking.
First, the cognitive mechanisms underlying spatial number representations are still debated. Second, it remains unclear whether spatial number representations are a scaffold for more complex forms of mathematics, and whether their potential foundational role spans all mathematical subdomains
across development and education. Third, the general role of spatial skills in spatial number representations and in their potential use in mathematical activities has never been investigated.
The present project proposes to answer these outstanding questions by a large-scale developmental study measuring the triangular relation between spatial number representations, spatial skills and mathematical abilities across development. While we focus on preschool children transitioning to primary school, collaborations at Ghent University will complement the present findings with data acquired in older children and young adults. Children are tested in the last year of preschool as well as in first grade to determine whether spatial number representations consist of temporarily constructed position-space associations in working memory (as expressed by the Ordinal Position Effect, OPE) and/or whether they are encoded in long-term memory (as expressed by the Spatial-Numerical Association of Response Codes Effect, SNARC).
In addition, the link between these indices of spatial number representations
and a variety of early mathematical abilities and spatial skills will be assessed. The latter measures will be complemented with data from standardized curricular tests (https://epstan.lu) in first grade. Overall, better knowledge of the mechanisms underlying mathematical development will not only inform general school practices but also directly benefit expertise and support
centers relating to education.
Researchers Involved: Carrie Georges, Tânia Ramos, Christine Schiltz
Duration: 2021 - 2024
Funding: FNR INTER (SpaNuMaDev)
Towards biomarkers of evaluation, development and fluency of reading
In the UL-International Research Partnership project will study the neuronal correlates of reading using an EEG approach in collaboration with Prof. Bruno Rossion (Neurosciences des Systèmes et de la Cognition, Département BIOSIS, CNRS, Nancy). In Nancy the focus will lie on optimizing the FPVS approach to study reading in neurotypical adults, aiming to develop biomarkers of reading ability. In Luxembourg we will evaluate reading abilities in different language settings, investigating monolinguals from different languages, as well as bilinguals. Furthermore the project will examine brain activations when learning to read in a first or a second language.
Researchers Involved: Aliette Lochy, Christine Schiltz
Duration: September 2022 – August 2026
Funding: FHSE, University of Luxembourg, Université de Lorraine
The brain network of numbers in deaf children & adults
Deaf individuals are slower and less accurate than their hearing counterparts in developing the concept of number, especially when tasks involve the use of symbolic codes (e.g. Arabic numbers). The current project aims at examining the role of deafness in accessing the representation of numbers through different formats (non-symbolic, iconic, symbolic), as well as characterizing its developmental trajectory in deaf children.
While previous studies mainly highlighted behavioural differences between deaf and hearing individuals, we would like to investigate the cerebral basis of the representations of numbers. The non-symbolic representation of numbers corresponds to any collection of non-symbolic quantities. The iconic representation of numbers corresponds to the finger counting and finger-montring habits of hearing individuals. The symbolic representation of numbers includes Arabic digits, written number words and numbers in sign language (symbolic only for deaf signers). Are the brain areas underlying the symbolic representation of numbers the same for Arabic digits and signed numbers in deaf individuals? Do the non-symbolic, iconic and symbolic number networks share the same neural correlates and do they follow the same developmental trajectory in hearing and deaf individuals? To investigate this question, a FPVS EEG paradigm will be used in groups of congenitally deaf children/adults having sign language as their mother tongue, and control groups of hearing individuals. The FPVS paradigm comprises the presentation of: 1) a base stimulus at a high-rate frequency and 2) a deviant stimulus at a lower frequency. If the brain is able to discriminate the deviant stimuli and to generalize across them, a selective response should be observed at the rate of the deviant presentation. In the first experiment numbers will be inserted periodically within streams of letters. In the second experiment canonical number representations will be inserted periodically within streams of noncanonical number representations.
Researchers Involved: Aliette Lochy, Christine Schiltz
Duration: February 2022 – December 2023
Funding: University of Luxembourg, WBI (DEAF_FPVS)
Neural impact of teaching method in learning to read an artificial script
Several methods for teaching to read are commonly used in classrooms, sometimes with fierce debates between proponents of different approaches. The two main families of teaching methods can be divided into phonics and global approaches. The phonics is based on teaching first the correspondences between individual characters and sounds (grapheme-phonemes), hence relying on the so-called assembly route. The global method on the contrary, focuses on teaching the link between the whole word-form to an oral item of the vocabulary. Previous research in ecological settings with children suggested that the global method induce a reliance on the left and right hemispheres when reading, which could reflect TTa strategy related to visual object recognition, by comparison to left hemispheric structures dedicated to reading. In adults, using artificial script allow us to manipulate precisely the characteristics of the relationship between the phonological form (eg., meaningful word, or pseudoword) and the written form, focusing either on individual elements or on the global form. This research program will inform us on the impact, both at the level of performance (reading fluency, transfer), and at the neural level, of the two teaching methods.
Researchers Involved: Aliette Lochy, Christine Schiltz
Duration: September 2021 – December 2023
Funding: University of Luxembourg, UCLouvain
Thinking about money: How does the use and processing of money influence the acquisition of numeracy?
This project is part of the 3E-Doctoral Training Unit, which gathers an interdisciplinary consortium of economists, psychologists, and social scientists conducting experiments involving human subjects. In the present experiment, the math performance of elementary school students and the use of real money as manipulative material is explored. Money is an interesting case with respect to the abstract nature of numerals, which children understand gradually during the development of their cognitive skills, because it reunites in a manner the abstract nature of number symbols with material value. However, the research on money processing is scarce.
More specifically, this study examines how the use of money influences the number processing and mathematics performance of 3rd grade students. First, students are introduced to manipulative materials consisting of real euro coins and banknotes in a teaching session of 45’ in classrooms where they practice using them while solving math tasks. One week later, an assessment of math tasks with and without manipulatives is conducted.
Demographic data, data on students’ out-of-school experience with money, and parents’ subjective financial wellbeing are collected from parents’ questionnaires. In addition, considering the large number of students with immigration background in Luxembourg and the multilingual environment in which they are growing, this research aims to explore the effect of the manipulatives in regard to the students’ general math and language fluency, as well as the students’ home language(s).
Exploring the influence of money as manipulative material while considering the background and pre-established knowledge of the students in Luxembourg could contribute to the research on mathematics education and offer further understanding of the cognitive and social aspects of our ability to deal with money.
Researchers Involved: Styliani (Stella) Politi, Christine Schiltz
Duration: October 2021 - October 2022
Funding: FNR PRIDE (MoneyMathEdu)
Processing of numerosity across formats
In this project we assess the access to number semantics with the use of FPVS-EEG. More specifically, in children we have tested the automatic discrimination of canonical configurations among non-canonical ones, and we have compared neural responses to dots and fingers (Marlair et al., 2021). We also examine the access to an abstract representation of numerosity across different formats (fingers, Arabic numbers, words, dots) using the well-known distance effect. First results show greater discrimination responses for distant numerosities than close ones, that furthermore correlate with the distance effect assessed at the behavioral level with comparison tasks. Further research questions involve the role of explicit instructions during the task, as well as cross-modal presentations (visual/auditory).
Researchers Involved: Aliette Lochy, Christine Schiltz
Duration: September 2021 – December 2023
Funding: University of Luxembourg, UCLouvain
Color and number associations
Color is often used in teaching numerical properties. For example, the Cuisenaire rods, that systematically vary in color and length, are used by children learning about quantity. Despite this, there is a lack of empirical evidence supporting the use of color, or the way in which color is applied, in learning materials. In this project, the empirical basis of associating numbers with colors is investigated. We are particularly interested in whether conceptual groups of colors support the formation of numerical properties, as opposed to the coloration of individual items. We are also interested in the timecourse of learning color-number associations, and their neural bases.
Researchers Involved: Talia Retter, Christine Schiltz
Duration: October 2019 – September 2024
Funding: University of Luxembourg
Reading words, reading numbers
Words and numbers are both constituted of strings of visual symbols, sharing some similarities but also striking differences in the cognitive processes deployed to read them aloud. In this research program, I am interested in a systematic comparison of the two codes, from the visual recognition of individual elements, to the nature of cognitive and neural representations involved when reading numbers. Several approaches are developed in order to understand and characterize better the commonalties and specificities of the two notations. With FPVS-EEG, we aim at understanding the specialization of the VOTc for these symbols, both in children and adults (e.g., Lochy & Schiltz, 2019: How early does our visual system distinguishes arbitrary symbols like letters vs digits according to their cognitive domain category?). With behavioral experiments, we assess the different reading routes for words and numbers (e.g., Lochy & Schiltz, 2022: are some familiar numbers recognized in an input lexicon?). Finally with clinical approach we test the selectivity of impairments in neurodevelopmental disorders like dyslexia.
Researchers Involved: Aliette Lochy, Christine Schiltz
Duration: January 2019 –
Funding: University of Luxembourg
Visual recognition of faces and words
Both faces and written word recognition involve specialized brain regions of the VOTc, and both involve a high degree of visual expertise. However, word reading is culturally very recent in history of humanity, and is acquired only via instruction; while face recognition on the contrary develops spontaneously and rapidly in infants. This research program, in collaboration with Bruno Rossion at University of Lorraine, involves understanding changes that occur in development on neural responses for faces, whether they are related to reading acquisition, whether there exists any neural competition between these two categories, and what are the factors that drive the right hemispheric specialization for faces.
Researchers Involved: Aliette Lochy, Christine Schiltz
Duration: January 2013 –
Funding: University of Luxembourg, Université de Lorraine (CNRS-CRAN)
The reading brain: Understanding how neural representations for words emerge and how they are shaped by teaching
The current project focuses on reading acquisition, and more specifically what characterizes the emergence of lexical representations (where words are considered to be stored and retrieved from an orthographic lexicon) at the neural level in typical and dyslexic children or adults, in natural settings or when they learn new words.
Using a highly sensitive paradigm that combines frequency-tagging and EEG recordings (Fast Periodic Visual Stimulation, FPVS), as well as behavioural tests, we aim at understanding how words are represented in the brain. FPVS allows to assess automatic and implicit discrimination between categories of stimuli (for instance words among non-words). This powerful approach will be applied to answer several research questions without requiring any linguistic task:
-
What drives and characterize the emergence of lexical responses for known words, both in children and in adults, typical and dyslexics?
We expect automatic neural responses to words among pseudowords to emerge at 10 years-old and to be related to behavioural reading fluency and vocabulary. Further, comparing dyslexics to age-matched and younger children will help characterize this disorder at the brain level as reflecting hypoactivation, maturational delay or compensatory mechanisms.
-
Do irregular and regular words entail similar brain responses?
In opaque languages like French, some words are difficult to read because they contain irregularities in the letters-to-sounds mappings. Studying if irregular words entail specific neural responses, both in children and in adults, typical and dyslexics, will help understand the nature of representations underlying reading.
-
Are some teaching methods more efficient than others for learning irregular words?
Three studies will focus on learning new words to examine the impact of pre-existing vocabulary knowledge (meaning and spoken form), as well as teaching method, on the novel neural representations created both in children and in adults.
Researchers Involved: Amaury Barillon, Aliette Lochy, Christine Schiltz
Duration: September 2022 – August 2025
Funding: FNR CORE (READINGBRAIN)
Electrophysiological biomarkers of language hemispheric lateralization and language ability, using fast periodic stimulation
Researchers Involved: Marion Marchive, Aliette Lochy, Christine Schiltz
Duration: September 2022 – August 2025
Funding: ANR, BioSerenity
Learning mathematics in social context
This research is about how to promote collaborative learning in mathematics for preschoolers.
Researchers Involved: Hoyeon Lee, Christine Schiltz
Duration: 2022 – 2025
Funding: C21 Consortium
Diagnosing specific learning disorders in mathematics in a multilingual education context
When suspecting a learning disorder in mathematics, individuals’ strengths and weaknesses are identified in an extensive diagnostic process to provide adequate support. As most diagnostic instruments rely on language to give instructions and present tasks, proficiency in the test language is likely to influence the student’s test performance. This becomes even more important in a culturally and linguistically diverse setting such as in Luxembourg. To date, there are no tests that are specifically adapted to Luxembourg, implicating that the tools currently used have been developed abroad. However, these tests mostly do not consider linguistic heterogeneity, and given that many children have highly varying test language proficiency, this may result in over- or under-identification of learning disorders.
In this project, we are developing a diagnostic test battery in the area of mathematics for children in grade 3, which takes into account the multilingual education context in Luxembourg. Based on diagnostic guidelines and neurocognitive models of number processing, a series of tasks have been devised. The instructions aim to have a reduced language load and the contents are adapted to the national curriculum of traditional public schools. The project encompasses the pretesting of the developed materials, followed by the validation of the new instrument, including the standardization of the administration process and the collection of normative data.
This project is part of the larger LESEMATH project of the Luxembourg Centre for Educational Testing (LUCET) of the University of Luxembourg.
Researchers Involved: Vera Hilger, Christine Schiltz
Duration: 2021 – 2024
Funding: Ministère de l’Éducation nationale, de l’Enfance et de la Jeunesse
Arithmetic and visuo-spatial attention
The ability to process numbers and perform mental computations is part of any educated adult cognitive toolkit but it requires long and effortful learning. Research shows that space processing is a fundamental determinant of human numerical skills, and that it significantly contributes to mental calculation processes. However, how number learning and space processing interact to give rise to adult numerical abilities is still unknown. The objective of the present project is to determine when and how spatial and numerical representations interact to give rise to spatial-numerical associations (SNAs) observed in mental arithmetic, in order to understand the learning mechanisms influencing mental arithmetic.
The project combines behavioural, oculometric and neuroimaging methods in healthy French-, German- and Arabic-speaking adults, children, and individuals born without the ability to move their eyes horizontally (Moebius syndrome). It was carried out thanks to the join expertise of two research groups in Belgium (UCLouvain) and Luxembourg (UL).
-
The first part of the project was led by the Belgian team of the UCLouvain composed of Nicolas Masson (post-doctoral researcher) and Mauro Pesenti (PI) and investigated SNAs in populations with different reading/writing habits, as well as in atypical individuals born with the inability to perform lateral eye movements (i.e., Moebius individuals).
-
The second part of the project was led by the Luxembourg team of the UL composed of Nicolas Masson (post-doctoral researcher) and Christine Schiltz (PI) and examined the developmental trajectory of SNAs in mental arithmetic.
The project helped to identify the putative neuro-cognitive roots and the developmental trajectory of the spatial-numerical associations in mental arithmetic. The main findings are:
1. From part 1 led by the UCLouvain team
-
SNAs in mental arithmetic are not related to the orientation of writing/reading, as evident from performance comparison of French (left-to-right readers/writers) vs. Arabic (right-to-left readers/writers) adult speakers (publication)
-
Oculomotor programming is functional for mental arithmetic, since interfering with eye movements with an eye-abduction paradigm in healthy adult participants led to impairments in arithmetic problems with carrying and borrowing processes (publication).
-
Horizontal SNAs in mental arithmetic reflect the cognitive strategies recruited by the participants while vertical SNAs are related to the absolute magnitude of the numbers that are manipulated (in preparation).
2. From part 2 led by the UL team
-
The neural correlates of spatial biases in mental arithmetic (i.e., addition and subtraction) can be investigated using a steady-state EEG approach (manuscript in preparation).
-
Single-digit additions and multiplications do not evoke the same neuronal responses in healthy adult participants. The neural response to sums that are slightly larger than the correct response suggests that solving processes rely on visuo-spatial attentional rather than verbal retrieval procedures. Since multiplications are known to rely on the latter, this indicates that simple additions are solved by functionally distinct mechanisms (in preparation).
-
Spatial attention is recruited to solve arithmetic problems, even in early stages of learning mathematics. With practice in arithmetic, children use space to anticipatively search for the answer (under review).
Researchers Involved: Nicolas Masson, Talia Retter, Christine Schiltz
Duration: 2018 – 2022
Funding: FNR INTER (SNAMATH: Spatial-Numerical associations in mental arithmetic)