United States

Normal aging is associated with an increase in word finding problems. Competing explanations posit that age impairs access to phonological representations (transmission deficit) or leads to a deterioration of semantic representations (under-activation). Because these accounts are difficult to disentangle in a highly interactive language system, we employed a neuro-biologically grounded spiking network model which was lesioned to reflect transmission deficits or under-activation. Results of three simulated picture naming paradigms were in line with the transmission deficit account that normal aging impairs access to representations during word production. These initial findings suggest that this is a promising approach for understanding age-related changes to language ability in an interactive system.

CogSci 2021

AgeNet: A Neurobiological Model of Age-related Word Retrieval Deficits

### Looking for Cogsci 2023? Please [click here](https://underline.io/events/405/reception) to switch to 2023 edition.


**Dear Colleagues,**

Welcome to the 43rd Annual Conference of the Cognitive Science Society! 

This year’s conference brings together the latest research from all Cognitive Science and also highlights this year’s theme **Comparative Cognition: Animal Minds.** 

The program features four plenary speakers: Friederike Range (University of Veterinary Medicine, Vienna), Onur Güntürkün (Ruhr University Bochum), Julia Fischer (University of Göttingen), and Erich Jarvis (The Rockefeller University). The keynotes and invited symposia showcase the broad spectrum of ideas that encompass the conference theme of comparative cognition. 

Further, the program includes the Rumelhart Prize keynote presentation by Susan Goldin-Meadow (University of Chicago), and symposia to celebrate the winner of the Elman Prize: Wei Ji Ma (New York University), the Heineken Award winner Robert Zatorre (McGill University), and the Robert J. Glushko Dissertation Prizes session.

The program committee for CogSci 2021 received 1,025 submissions, including 710 full papers, 275 member abstracts, 8 publication-based talks, as well as 15 proposals for symposia, and 17 for workshops and tutorials. After a rigorous review process, the committee selected 152 papers for oral presentation and inclusion in the conference proceedings (14.8%), 332 papers for poster presentation and inclusion in the proceedings (32.4%), and 175 papers for poster presentation with inclusion of abstracts in the proceedings (17.1%) We also selected 275 submitted member abstracts. In addition, we accepted 6 publication-based talks, 12 symposia, and 9 workshops and tutorials to make for an exciting and inclusive program.

We hope that you enjoy the conference this year!<br><br>

**Your Program Chairs,**
![](https://assets.underline.io/uploads/markdown_image/1/image/99984908caba5e6453e5fc5ba34f3507.jpg)

**Learn more about the Vienna Cognitive Science Hub at the University of Vienna, hosts of CogSci 2021**

<iframe width="560" height="315" src="https://www.youtube.com/embed/464h1m3iU68" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe>

To review the proceedings for the conference or review the searchable program, please visit the [website](https://cognitivesciencesociety.org/program/).


The Cognitive Science Society brings together researchers from around the world who hold a common goal: understanding the nature of the human mind. The mission of the Society is to promote Cognitive Science as a discipline, and to foster scientific interchange among researchers in various areas of study, including Artificial Intelligence, Linguistics, Anthropology, Psychology, Neuroscience, Philosophy, and Education. The Society is a non-profit professional organization and its activities include sponsoring an annual conference and publishing the journals Cognitive Science and TopiCS.

poster

Biological motion perception is supported by a network of regions in the occipito-temporal areas, primarily in superior temporal sulcus (STS), and premotor cortex (PMC). How biological motion is processed outside the focus of attention and whether it is modulated by attentional load remain unknown. We investigated the bottom-up processing of biological motion under different levels of attentional load (high vs. low) with functional magnetic resonance imaging (N=13). In line with previous work, we found that fronto-parietal attention regions were significantly more activated when the attentional load was high than when it was low. Importantly, biological motion under low attentional load yielded activity in STS and PMC, whereas biological motion under high load was restricted only to the low-level motion sensitive areas. These results show that biological motion is processed outside the focus of attention and it is modulated by attentional load.

Biological Motion Perception under Attentional Load

The concept of accessibility has often been evoked to explain reference resolution. According to the Givenness Hierarchy theory, a referent’s accessibility in the mental state of a comprehender is encoded in the form of the referent as part of its lexical semantic representation. However, the current literature has not reached a consensus on what accessibility exactly means and how to best quantify it. The factors that modulate accessibility show a great extent of overlap with another independently motivated concept of predictability, raising the possibility of a “Predictability Hierarchy” that mirrors Givenness Hierarchy. In a self-paced reading study, the current study examines whether there is such “Predictability Hierarchy” by manipulating the predictability and the form of a referent presented to the participants. Our results indicate that although there is no strong evidence for approximating the Givenness Hierarchy with a “Predictability Hierarchy,” there is some preliminary evidence for a partial correlation between the form and the predictability of a referent.

Is there a predictability hierarchy in reference resolution?

This study focused on associative learning for new words in the first language and haptic stimuli. In the first, healthy Japanese participants made nine subjective evaluations of haptic stimuli using five-point semantic differential scales (e.g., regarding stickiness, scored from 1 [not sticky] to 5 [sticky]). In the second and third, the participants carried out two learning tasks for associative pairs of a new word in Japanese and haptic stimulus (H), or new words only (W). In the fourth one, after each learning task, participants performed recognition and free recall tasks. The results of the recognition tasks showed that the accuracy rates of W were better than those of H, whereas the response times of W were faster than those of H. Further, preference of haptic features negatively correlated with free recall scores of H; however, there was no significant difference between the free recall scores of H and W.

Associative learning of new word forms in a first language and haptic features in a single-day experiment

There is a growing interest in alternative explanations to the dual-system account of how people learn category structures varying in their optimal decision bounds (unidimensional and information-integration structures). Recognition memory performance and hippocampal activation patterns in these tasks are two interesting findings, which have not been formally explained. Here, we carry out a formal simulation with SUSTAIN (Love, Medin, & Gureckis, 2004), an adaptive model of category learning, which had great success in accounting for recognition memory performance and fMRI activity patterns. We show, for the first time, that a formal single-system model of category learning can accommodate recognition performance after learning and is consistent with fMRI data obtained while participants learned these structures.

SUSTAIN captures category learning, recognition, and hippocampal activation in a unidimensional vs information-integration task

Domesticated animals and modern humans show a set of behavioral and molecular changes that converge on shared genetic targets. There is evidence that these changes disproportionately target neurotransmission, in particular the glutamatergic signaling system. This has led to proposals that attenuation of glutamatergic signaling may be crucial for the downregulation of the stress response in both modern humans and domesticates and the potentiation of exploratory motor output in our species. Here, we use a deep learning method (ExPecto) to predict gene expression in silico of H. sapiens-specific (relative to the closest extinct human species, Neanderthals and Denisovans) variants of genes involved in glutamatergic signaling. This approach allows us to hone hypotheses about the functional implications of genetic changes in our species' recent evolution, including proposals as to the neurobiological substrates of H. sapiens so called 'self-domestication hypothesis'.

Gene expression under human self-domestication: an in silico exploration of modern human high-frequency variants

Information in speech appears in bursts. To optimize speech perception, the brain aligns these bursts of information with slow rhythms in neural excitability (<10 Hz). How does the brain track the timing of external events? Here we tested whether neural stimulus-tracking depends on participants’ expertise, or on a language-specific mechanism. We recorded electroencephalography (EEG) in participants who were experts in either ballet or in sign language, while they watched videos of ballet or sign language. We show that stimulus-tracking depends on expertise: Dancers' brain activity more closely tracked videos of dance, whereas signers' brain activity more closely tracked videos of sign language. This effect of expertise emerged at frontal channels, but not at occipital channels. These results suggest that frontal cortex forms temporal predictions based on expert knowledge. The brain may use the same predictive mechanisms to optimize perception of temporally-predictable stimuli such as speech, music, sign language, and dance.

Expertise modulates neural tracking of dance and sign language

While there are interesting correspondences between form and meaning in many languages, psycholinguists conventionally consider them to be marginal, as they affect only a small subset of words. As such, a common simplification in computational models is to use empirically- or theoretically-motivated representations for form and random vectors for semantics. We recently introduced a novel model of human speech recognition, EARSHOT, which maps spectral slices (form) to pseudo-semantic patterns (sparse random vectors [SRVs]). Here, we replace SRVs with SkipGram vectors. Empirically-based semantics allow the model to learn more quickly and, surprisingly, exhibit more realistic form competition effects. These improved form competition effects do not depend on the particular form-to-meaning mapping in the training lexicon; rather, they arise as a result of the nontrivial output structure. These results suggest that while form-meaning mappings may be mainly arbitrary, realistic semantics afford important computational qualities that promote better fits to human behavior.

Distributed semantics in a neural network model of human speech recognition

Is numerosity an abstraction that arises downstream of basic perceptual processing, or does the brain process numerosity like a primary perceptual feature such as color or motion? Here, we tested whether visual cortex computes number automatically, even when number is task-irrelevant and is being processed unconsciously. We recorded electroencephalography while subjects watched dotcloud stimuli that alternated in numerical content at 15 Hz, under one of three conditions: judging numerosity, judging convex hull, or making no judgments. Under all three conditions, we observed oscillatory activity at 15 Hz in early visual cortex. The strength of this signal depended on the numerical content of the stimuli in all three conditions and did not differ significantly across the conditions. Results show that number can be computed spontaneously by visual cortex, even when participants are attending to non-numerical information, consistent with the proposal that number is a primary perceptual attribute of visual stimuli.

The human visual system spontaneously computes approximate number

How does scientific knowledge grow? Some argue that science advances by developing a set of core concepts, while others argue complementary or competing core/periphery structures allow for scientific progress. We conducted a large-scale analysis of abstracts from the American Physical Society, spanning 58 years. For each year, we created a network of concepts, where single- and multi-word noun phrases were linked if they appeared in the same abstract. Our results suggest that a single core––rather than multiple cores––best explains the structure of our concept networks. We also see that only a small subset of core concepts (around 10%) remain within the core. Finally, we see a decline in the relative size of the core, driven by the growth of scientific production. The decline in core size is negatively associated with discoveries that push the scientific frontier, which suggests that strong core/periphery structures may be important for innovation.

Network Dynamics of Scientific Knowledge Reveal a Single Conceptual Core that Declines Over Time

Recent research has proposed that adjective form (i.e., whether adjectives typically occur before or after the nouns they modify) interacts with considerations from efficient communication to determine the rate at which we use adjectives to resolve reference to objects. According to this efficiency hypothesis, languages with pre-nominal adjectives use modifying adjectives at a higher rate in an effort to aid incremental reference resolution. We test this broad typological prediction in a large-scale corpus analysis of 74 languages, finding that languages that favor pre-nominal adjectives indeed do exhibit higher rates of adjectival modification than languages that favor post-nominal adjectives.

AgeNet: A Neurobiological Model of Age-related Word Retrieval Deficits

Downloads

Next from CogSci 2021

Biological Motion Perception under Attentional Load

Similar lecture

A common framework for quantifying the learnability of nouns and verbs

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES