Friday 18 October 2013

PRESENTATION: How the brain processes surprise and prediction, by Dr. Marta Garrido (Queensland Brain Institute)

The School of Psychology is proudly hosting a talk by:

Dr. Marta I. Garrido
DECRA Research Fellow
Queensland Brain Institute
The University of Queensland


Title: Sensitivity to statistical structure in the human brain


Date: Thursday 24th October 2013, 12-1pm in Keats Reading Room (AVLG17) (video streaming to Science Offices Meeting Room, E1.32 at Ourimbah). If you would like to meet with Dr. Garrido, please contact Dr. Juanita Todd (juanita.todd@newcastle.edu.au).


Abstract: Salient or oddball events in the environment may signal potential rewards or threats. The ability to detect these events is fundamental to survival and adaptive behaviour. In this talk, I will discuss behavioural and magnetoencephalographical (MEG) data that demonstrate the brain’s ability to learn about, and detect changes to, statistical structure in sounds sequences. In addition, I will discuss the usefulness of the oddball paradigm and model-based connectivity in demonstrating functional asymmetries in audio-spatial perception.



Thursday 3 October 2013

PRESENTATION: How objects are represented in the human brain, by Dr. Tom Carlson (Macquarie University)

The School of Psychology is proudly hosting a talk by Dr. Tom
Carlson (ARC Future Fellow, Department of Cognitive Science
Macquarie University)

WHEN: Thursday 10th October 12-1pm

WHERE: AVLG17, Aviation Building, Callaghan Campus
Also videoconferenced to Ourimbah E1.32.

TITLE: Object representations in the human brain: How they
emerge, and their connection to semantic knowledge and
perceptual decision-making

ABSTRACT: Our capacity to recognize visual objects is mediated
by multiple visual areas, in particular inferior temporal cortex
(ITC). Recent technological and methodological advancements have
afforded researchers the capacity to measure the complex
geometrical structure of brain representations. In this talk, I
will describe several studies that examine how the structure of
object representations emerges in the brain, and how this
structure relates to behavior. I will first describe experiments
that use brain decoding methods and magnetoencephalography (MEG)
to study the emergence of object category structure. Our findings
suggest that hierarchical category structure (e.g. human face is
a member of the superordinate category human) is a process of
accumulating evidence in which categories solidified at early
stages of processing inform more abstract category representations
that emerge later. I will then describe two studies that examine correspondences between the structure of object representations in
human ITC and human word usage patterns and decision-making
behavior. These studies show that the structure of object
representations in ITC correlate with human word usage patterns in
language; and ITC’s representational structure predicts reaction
times for object categorization. These studies collectively provide
an important link between the geometry of brain representations and
functional behavior.

Tuesday 1 October 2013

PRESENTATION: Dr. Titia Benders (Nijmegen, the Netherlands). How do babies learn language?

The School of Psychology is proudly hosting a talk by Dr. Titia
Bender
(Center for Language Studies, Radboud University Nijmegen
The Netherlands)

Thursday, 3rd October, 2013, 12-1pm in AVLG17 (Aviation Building, Callaghan Campus).

How do babies learn language from the way we talk to them?


Babies learn about their native language by listening
to the speech that surrounds them. To understand how language-
learning happens in this situation, we must know: 1) what kind of
speech babies are confronted with and 2) what babies learn from
this.

Babies hear a special type of speech, because most adults speak
differently to babies. It has been proposed that parents speak
to babies in a special register to express their positive affect
and to teach them the language. I will show more fine-grained
analyses of infant-directed speech, which reveal that all parents
do is express positive affect when they speak to their baby. I
will argue that earlier results that were interpreted as language
teaching, were actually happy speech in disguise.

Even though parents do not actively teach their baby language,
babies manage to learn language from what they hear. Listening
experiments provide evidence what babies learn about the language,
but cannot reveal how they learn it. I will illustrate that the
combination of analyses of language input, listening experiments
with babies, and computational modeling can bring us a long way
in answering the question of how babies use the speech that
surrounds them to learn their native language.