This study examines the neural underpinnings of Critical Thinking (CT) using functional magnetic resonance imaging (fMRI) in high school students. It is integrated in a more comprehensive study whose main goal is to uncover the links between writing and comprehension skills in one's mother tongue and the cognitive skills of critical thinking among high school students. While previous research has explored behavioral aspects of CT, few studies have investigated its neural correlates. Understanding how the brain processes CT tasks can provide valuable insights for improving educational strategies and informing pedagogical interventions.

The central research questions include: (i) What are the neural markers of CT as measured by fMRI? (iii) How does neural activity in language-related brain regions correspond to CT performance?

The study is framed within three theoretical perspectives: (i) Vygotsky’s socio-interactionist theory (1962, 1991), which positions language as fundamental for cognitive development; (ii) Hagoort’s MUC model (2013), which describes neural mechanisms underlying language processing; and (iii) Facione’s CT model (1990), categorizing key CT competencies.

Since Vygotsky's foundational theory, language is accepted as the main symbolic system of

human groups, representing a qualitative leap in the evolution of the species (Vigotsky, 1962, 1991; van Berkum & Nieuwland, 2019). This theory also considers the human brain as the biological basis for learning (Vygotsky, 1962, 1991). But how does language act to develop such activities? And how does it impact learning in an educational context, such as CT learning? Bronckart (2006, 2007) defends that situated language-practices (or tex-discourses) have a major contribution in developing human knowledge and skills. As code shared by the set of its speakers, language allows the construction of socially shared meanings and the construction of knowledge considered necessary for a society at a given socio-historical context.

In neuroscience, the social interactionist view of language has played a core role in dialogues on human language (see Hagoort, 2019 for the state of the art). We adopt the “MUC” neurobiological model of language (Hagoort, 2013), according to which language processing involves three functional components: Memory, Unification and Control. The memory component houses the linguistic knowledge accumulated during the course of language acquisition and encoded in neocortical memory structures. It contains knowledge on phonology, morphology, and syntactic language organization. The Unification component refers to the activities of concatenating the linguistic pieces retrieved from Memory (syntactic, semantic, phonological, lexical level information) in constructing larger structures supported by contextual information. The Control component is responsible for linking language to joint actions and social interaction such as selecting the appropriate target language, taking turns in a dialogue, and adjusting the register to the different communicative situations.

These two complementary language approaches will be applied to assess Facione’s (1990)

cognitive dimension of CT that includes the skills of (1) interpretation, (2) analysis, (3)

evaluation, (4) inference, (5) explanation and (6) self-regulation that require language abilities of comprehension and writing.

A cross-sectional mixed-methods approach is employed, integrating behavioral assessments with fMRI. To assess the cognitive dimension of critical thinking skills, we use a validated Portuguese- language Critical Thinking Test (Lopes et al., 2018). fMRI will analyze neural activation patterns in regions associated with language processing according to the MUC model of language comprehension (Hagoort, 2013). We adopt a naturalistic neuroimaging protocol from the longitudinal study of Meshulam et al. (2021). This protocol involves measuring neural activity using fMRI while students passively listen to naturalistic auditory samples from the Critical Thinking Test text, followed by a behavioral assessment of their learning outcomes. The assessment consists of 6 CT test questions evaluating the following skills: (1) interpretation, (2) analysis, (3) evaluation, (4) inference, (5) explanation and (6) self-regulation. Given the high costs involved in fMRI a subset of 6 high school students (two per grade level) will undergo fMRI scans while performing CT-related tasks. We will use intersubject correlation (ISC) analysis to identify brain regions exhibiting stimulus-driven responses across levels of processing (Nastase et al., 2019). This approach will allow us to simultaneously map brain activity in a variety of language comprehension areas spanning the MUC model: temporal cortex, the angular gyrus, Broca’s area (Brodmann areas 44 and 45) and adjacent cortex (Brodmann areas 47 and 6), and the anterior cingulate cortex. We hypothesize that students with more advanced CT skills and better learning outcomes will have strong neural engagement with the TPC stimuli across language areas, particularly in the Unification and Control areas. Learning outcomes assessment will consist of the 6 questions of the TPC test also described above. fMRI data will be analyzed using Python 3 (www.python.org) and R (www.r-project.org), using the Brain Imaging Analysis Kit (http://brainiak.org). Behavioral outcomes, based on written responses to the TPC test, will be evaluated by two independent judges using the predefined analysis criteria. Neural activity patterns and behavioral learning outcomes will be examined both individually and collectively, as well as compared across the three high school grade levels. Ethical approval will be obtained, ensuring compliance with neuroscience research standards.

This study aims to uncover neurophysiological responses related to language comprehension areas involved in critical thinking cognitive skills. We expect to provide evidence for a large study that establishes a neurocognitive model of CT, guiding educational practices by linking CT competencies to specific brain networks and identifying differences in neural activation between high and low CT performers. Expected findings include: 1) providing fMRI-based evidence linking CT to language-processing; 2) insights into the neurocognitive basis of CT skill development; 3)contributing to scientific knowledge on the role of language to learning in school learning contexts; 4) getting evidence on the links between language ability and CT that support a future large study. The comprehension of these processes could improve education, regarding CT, by using the potential of our more powerful learning tool - language - to intentionally develop CT in young students. We expect that this research contributes to an evidence-based understanding of CT, supporting innovative pedagogical approaches that integrate cognitive neuroscience with educational strategies. The findings will offer new directions for curriculum designers and policymakers to enhance CT instruction through targeted interventions based on neural evidence.

Bronckart, J-P. (2006). Les conditions de construction des connaissances humaines. In: M. Carton & J-B. Mayer (Ed.). La société des savoirs: trompe-l'oeil ou perspectives ? (pp. 27-48). L'Harmattan. https://archive-ouverte.unige.ch/unige:37556 Bronckart, J-P. (2007). Desarollo del Lenguaje y Didactica de las Lenguas. Miño Y Dávila. Facione, P. (1990). Critical thinking: A statement of expert consensus for purposes of educational assessment and instruction. The California Academic Press Hagoort, P. (2013). MUC (Memory, Unification, Control) and beyond. Frontiers in Psychology, 4. Article 416 Hagoort, P. (Ed.). (2019). Human Language: from genes and brains to behavior. MIT Press. Lopes, J., Silva, H., & Morais, E. (2018). Test de pensamiento crítico para estudiantes de enseñanza básica y secundaria || Critical thinking test for elementary and secondary students. Revista de Estudios e Investigación en Psicología y Educación, 5(2), 82-91. https://doi.org/10.17979/reipe Meshulam, M., Hasenfratz, L., Hillman, H., Liu, Y-F., Nguyen, M., Norman, K. A., & Uri Hasson (2021). Neural alignment predicts learning outcomes in students taking an introduction to computer science course. Nat Commun 12, 1922. https://doi.org/10.1038/s41467-021-22202-3 Nastase, S.A., Gazzola, V., Hasson, U., Keysers, C., 2019. Measuring shared responses across subjects using intersubject correlation. Soc. Cogn. Affect. Neurosci. 14, 667– 685. doi:10.1093/scan/nsz037. van Berkum, J. J. A., & Nieuwland, M. S. (2019). A Cognitive Neuroscience Perspective on Language Comprehension in Context. In: P. Hagoort (Ed.). Human Language: from genes and brains to behavior (pp. 429-442). MIT Press Vygotsky, L. S. (1962). Thought and Language. MIT Press. https://doi.org/10.1037/11193-000 Vygotsky, L. S. (1991). Genesis of the higher mental functions. In P. Light, S. Sheldon, & M. Woodhead (Eds.), Learning to think (pp. 32–41). Taylor & Frances/Routledge. (Original work published 1966)