Projects

UIDB/04527/2020; UIDP/04527/2020
CICPSI 2022-2023
Emma Delhaye (PI), Ana Raposo (Co-PI)

Unitization is a strategy that allows to encode associations between several pieces of information into episodic memory as one unique, integrated, entity. Despite the great interest triggered by unitization in the literature, especially as a compensation strategy in populations with decreasing or impaired associative memory (e.g., aging, amnesia), little is known about the mechanisms underlying the formation of unitized representations. This project combines EEG measures, eye-tracking data, and a longitudinal design, with three specific aims: (i) to characterize the temporal dynamics of unitization, (ii) to uncover potential eye-movement correlates, (iii) to understand the contribution of semantic memory to the unitization process.

PTDC/PSI-GER/3281/2020
FCT 2021-2024
Susana Araújo (PI), Tânia Fernandes (Co-PI), Mariona Pascual, Miguel Domingues, Alexandra Reis (Universidade do Algarve), Luís Faísca (Universidade do Algarve), Tiago Guerreiro (Faculdade de Ciências, Universidade de Lisboa), Ana Pires (Faculdade de Ciências, Universidade de Lisboa).

Letter recognition is the foundation of reading, a core skill in educational sphere and also for successful participation in every aspect of modern life. Learning to read is a multisensory experience, which involves successfully linking visual letters with speech sounds and with motor gestures through handwriting. It is already clear that handwriting assists the implementation of reading in the brain and mind, but how does it happen?
Two systems conspire for skilled reading: a visual shape recognition system, in the left ventral occipitotemporal cortex (including the visual word form area, VWFA), and a motor gesture decoding system, in the left dorsal premotor cortex (including the Exner’s area). Our original hypothesis is that handwriting assists letter recognition via segmental, fine-grained information that feeds into the visual perceptual system. We will test it in 3 studies with an artificial-script learning paradigm, by adopting state-of-the-art neurocognitive, electroencephalography (EEG) and eye-tracking methods. By using writing logging data in real time during training (HandSpy), we will also assess if training-induced neural changes are related to handwriting expertise (e.g., writing speed). Establishing this relation is crucial for better understanding of brain-behavior relation.
We will thus give clear answers by targeting 3 specific aims on when and how motor action affects letter perception and interacts with the other fundamental systems in letter processing : 

1) whether handwriting enhances letter representations by tuning distinctive letter features (Study1); 

2) whether it affects early- vs late-stage visual processing and the role of attention (Study2); 

3) how multimodal orthographic, phonological, and motor information is combined during letter recognition (Study3). We will track for the first time neural EEG responses to audio-visual-motor integration, to determine if phonological and motor inputs are integrated or processed separately and if one type prioritizes over another for efficient letter recognition. 

LEMON focus on a crucial cognitive domain; understanding how learning experiences shape letter representations is an important endeavor for successful education policies in typical and dyslexic readers.

PTDC/PSI-ESP/30958/2017
FCT & FEDER 2018-2022
Moreno I. Coco (PI), Ana Raposo (Co-PI), Giorgia D’Innocenzo, Emma Delhaye, Anastasiia Mikhailova, José Santos-Vitor, Isabel Pavão Martins

The interaction between attention and memory is a key feature of our cognitive system, enabling us to successfully interact with the environment. These cognitive components are known to decline due to cognitive aging, especially if pathological, which has repercussions on the daily functioning of older adults and indirectly to the societal fabric as a whole. Our project focuses on the interplay between overt attention and visual memory by examining the role played by semantic and visual properties of naturalistic scenes and real objects in directing eye-movements as visual memories are formed, maintained and accessed. We co-register electrophysiological brain activity (EEG) and eye movements to tap into explicit signatures of overt attention and implicit neural correlates of successful memory, and thus to gain empirical insights about the interplay of such processes. This investigation is carried out in younger, healthy older and pathologically aged populations to distinguish mechanisms that are preserved from those that degrade due to cognitive aging.

PTDC/PSI-GER/28184/2017
FCT & FEDER (POR Lisboa 2020) 2018-2022
Tânia Fernandes (PI), Susana Araújo (Co-PI), Miriam Aguilar, Paulo Ventura, Isabel Leite (Universidade de Évora), Isabel Pavão Martins (Faculdade de Medicina, Universidade de Lisboa), Luis Correia (Faculdade de Ciências, Universidade de Lisboa).

To create an e-mail account we must pass a word-recognition test on distorted letters: swift for human readers but a hard challenge for advanced spambots. CAPTCHA (Completely Automated Public Turing test to tell Computers and Humans Apart) shows how little we know on orthographic processing, i.e, access to abstract letter identities: from pixels into words. VOrtEx tests a new proposal aimed to unravel these representations. We and others showed that reading not only puts heavy demands on visual processing but also rests along the ventral occipitotemporal (vOT) stream due to partial recycling of object recognition, the original function of vOT. Thus, the orthographic code is either distinct of other visual representations or is just object recognition modulated by language. How can we test it, if most vOT properties are suited for reading which, in turn, depends on mapping of orthography into phonology in an interactive system? Our breakthrough idea is that it can, by testing two visual adaptations required for reading but that depart from the original properties of object recognition: narrow spatial integration of features in letters, i.e., reduced crowding, and discrimination of mirror images (d – b). VOrtEx answers 3 core questions: When in processing do these adaptations occur in the mature reading system? What is their developmental trajectory? Do they crash in the reading disorder? To track the full spectrum, studies will test adult readers (Study1), illiterate, ex-illiterate and literate adults, 1st-6th-graders (Study2), and dyslexic readers (Study3). We predict these adaptations occur early on in processing by adults, reach full action after few years of reading experience, but are hampered in dyslexia. The strong theory-driven nature from Cognitive Psychology, Neuroscience and Computational Science and the multi-method groundbreaking approach are keys to success. Fundamental and translational research beyond national borders. VOrtEx replies to challenges on Health, Inclusive, and Security Societies with dissemination and outreach plans along Portugal2020: “if smart growth is about knowledge and innovation, investment in literacy skills is a prerequisite for achieving such growth” (European Commission; p.11).

PTDC/PSI-ESP/28414/2017
FCT 2018-2022
Ana Raposo (PI), María J. Maraver (Co-PI), Ana Lapa, Leonel Garcia-Marques, Sara Cardoso, Angel Fernandez, Pedro Albuquerque, Bridgid Finn, Paula Carneiro

The debate over whether unsuccessful retrieval enhances or harms subsequent learning is yet unresolved. The errorful learning literature proposes that error generation can enhance future learning if followed by corrective feedback, because retrieval of a wrong answer provides a good context for incorporating the corrective information. The project tackles this unresolved question using distinct populations (adults and adolescents) and experimental paradigms that involve an episodic retrieval that occurs into a deeper semantic network, leading to a richer context for encoding the corrective feedback. We employ pragmatic inference sentences, often used to study false memories, that represent more closely what happens in the real world and offer a way of studying errors in which participants are confident that their incorrect answers are right. The results allow to establish the advantages and disadvantages of using unsuccessful retrieval in classroom settings, offering important insights to educational practice.

IF/00533/2015
FCT 2017-2021
Susana Araújo

This project aims to shed light on how a highly specialized system for letter processing emerges in the human brain. We will track i) when during reading acquisition the letter-specialized system emerges, i.e., if through rudimentary phonological recoding operations vs. from recognizing letter in words; ii) which mechanisms bolster a functional tuning to fine-grained orthographic information (fine tuning), focusing on the role of reading proficiency, strategies for reading, and the opacity of the letter-sound mappings reinforced during learning to read; and iii) whether orthographic processing may be constrained by general principles of visual object processing, such as statistical learning. We will aim at clarifying these questions using high-temporal resolution methods such as event-related potentials and eye-movement recordings and testing individuals varying in reading competence (proficient vs. dyslexic readers).

PTDC/MHC-PCN/1267/2014
FCT 2016-2018
Leonel Garcia-Marques (PI), Ana Sofia Santos, Sara Cardoso, Ana Raposo, Tomás Palma, Ludmila Nunes, Yana Weinstein, Diana Orghian, Paula Carneiro

We follow the literature that stresses the role of retrieval in memory and learning, and explore three new aspects: i) the powers of retrieval as learning episodes; ii) the problem-solving nature of memory tasks and the interplay between memory and knowledge; iii) the hidden costs of retrieval mode. In a series of experiments: i) we expect participants across study-test cycles to improve the ways they meet test requirements and disregard test-irrelevant dimensions; ii) we manipulate the relevance of different types of concept-based knowledge and let participants learn about it across study-test cycles; iii) we test how new information presented during a retrieval mode tend not to be learned, and how misleading information that seems relevant to reconstruct the episode ends up incorporated giving rise to false memories.

180/14
Bial Foundation 2016-2018
Tânia Fernandes (PI), Ana Raposo (PI), Isabel Pavão Martins (Faculdade de Medicina, Universidade de Lisboa), José Pimentel (Grupo de Cirurgia de Epilepsia, Hospital de Santa Maria).

Learning new words is an impressive capacity of the human brain; it is a frequent, swift, and effortless lifelong learning ability. This project aims to investigate the neurocognitive mechanisms by which new words become part of the lexical network. This is a fundamental endeavour that will allow characterizing the dynamic relation between memory and language. It will also inform how word learning may be resilient to neural changes in healthy ageing and in amnesia (i.e., selective episodic memory deficit). We propose two complementary studies, adopting an integrative and interdisciplinary approach. We focus on the interface between memory and language , proposing that learning new words involves two facets that differ in time course and underlying neural mechanisms. The episodic facet is rapidly available and is supported by the hippocampus and the surrounding medial temporal lobe . The lexical facet emerges slowly and relies on a distributed neocortical network, including lateral temporal and frontal regions. After episodic encounters with a new word, a stable lexical representation is formed, which abstracts away specific prior experiences with that word. The new word will then become an active lexical representation, competing for recognition with phonologically similar words. Yet, little is known about the transition from episodic to lexical representations at the neural level and the specific role of the hippocampus on it. We will investigate three hypotheses by combining neuroscience tools with skillful cognitive designs, leveraging the research centre facilities and our team expertise.

IF/00886/2013/CP1194/CT0002
FCT, 2015-2017
Tânia Fernandes

This project tested a new proposal on the impact of learning to read on visual object recognition that considers the role of the ventral vision-for-perception and the dorsal vision-for-action streams on literacy-dependent changes. Prior studies were the first to show that literacy deeply affects the object recognition system. Learning a script with mirror symbols (e.g., b – d) pushes the learner to break mirror invariance, one original property of object recognition that collides with mirror discrimination required by literacy. Impressively, mirror discrimination becomes automatic during nonlinguistic object recognition but three core questions remained open: How deep in terms of processing levels are these literacy-dependent changes? Since visual processing depends on the interplay between the two visual streams what is the role of the dorsal stream in mirror discrimination? How does it interact with literacy? We will aim at answering them, starting from disperse multidisplinary evidence integrated into a coherent proposal. Based on our proposal we will develop software applications directed to children to facilitate learning to read by promoting links between motor actions and mirror discrimination, disseminating new technologies with impact in economy and society- Also, understanding the links between literacy and object processing will help improve literacy instruction and remediation programs for dyslexia, a societal challenge priority of the Europe 2020 strategy for smart, sustainable and inclusive growth.

89/12
Bial Foundation 2013-2015
Ana Raposo (PI), J. Frederico Marques, José Pimentel

Temporal lobe structures are implicated in long-term memory, with the medial portion being associated with episodic memory and the anterolateral region linked to semantic memory. Yet, temporal lobe epilepsy (TLE) studies challenge this view: medial TLE patients have severe amnesia but lateral TLE patients do not present prominent semantic deficits. By testing medial and lateral TLE patients, and using sensitive measures of semantic ability, we aim to elucidate the nature of semantic deficits in epilepsy. Additionally, we examine how intact vs. compromised semantic systems in the two groups differentially affect episodic recollection. Results inform about the scope of temporal lobe to long-term memory and are an important step towards an accurate prognosis of anterior temporal lobe resection in epilepsy.

PTDC/PSI-PCO/118148/2010
FCT 2012-2015
Ana Raposo (PI), Andrea Santi, Joana Carmo, Mara Alves, J. Frederico Marques

Our conceptual knowledge is organized into categories, which help us make sense of the world, quickly, knowledgeably and pragmatically. This project investigates the neurocognitive structure of conceptual knowledge and how it changes along ageing. The main innovative approach is the consideration of category structure as represented by the continuum of typicality ratings. The first endeavour is to widen the scope of typicality to different conceptual levels to get a hierarchical representation of category structure (e.g., sparrow, bird, living thing). The results of this study are then used to parametrically relate hierarchical category structure to brain activation and atrophy in healthy young and older adults.

PTDC/PSI-PCO/113774/2009
FCT 2011-2014
Ana Raposo (PI), Sofia Frade, Mafalda Mendes, J. Frederico Marques

Understanding how semantic and episodic memory systems interplay in the healthy brain is a crucial endeavour as it will ultimately inform about how memory difficulties may be partially overcome if new learning is anchored to prior knowledge. We propose a set of fMRI studies to gain a more precise understanding of: i) the nature of the semantic operations that promote vs. hinder recollection of past events; ii) the role of specific prefrontal cortex (PFC) regions in semantic processing during episodic retrieval (e.g., early semantic generation processes vs. late integration of semantic cues); iii) the co-modulation of PFC and other brain regions, in integrated networks, subserving memory recollection and expression of knowledge.

PTDC/PSI-PCO/114822/2009
FCT 2011-2014
J. Frederico Marques (PI), Jorge Almeida, Ana Raposo

How have different visual pathways contributed to object recognition and to the emergence of different object processing circuits? The first goal of this project is to lay out a map of low-level influences on high-level visual object processing. We adapt psychophysical manipulations that explore these pathways’ physiological responses and measure how object recognition processes are differentially affected. The second goal is to understand how the visual object processing circuits, defined in the first goal, interact with mid-level object-related knowledge. In fMRI and behavioural studies, we use different types of object knowledge (e.g., function, manipulation) and see how they are differentially processed within the processing circuits.