research suggests we don't actually perceive separate objects and
actions. We understand them in terms of how we have processed their
relationships and associations with other objects and actions.
Alex Huth, a doctoral student in neuroscience at the University of California at Berkeley and a team of colleagues at the university's Gallant Lab
wanted to learn how the brain makes sense of the thousands of visual
images we see every day. Their findings appear to refute the long-held
view that each category of objects and actions is represented in a
separate part of the brain. Instead, it seems we create complex,
intricately related and overlapping groupings represented in what the
researchers call "a continuous semantic space."
Posted on December 21, 2012 by Zachary Urbina in Neuroscience
Researchers have even mapped how we organize things.
Our brains create semantic neighborhoods, populated with things we see
things as living and nonliving; moving, like cars and motorcycles, and
stationary, like buildings and the sky, and we group things we
understand social, such as people and verbs. Listen to Huth's
explanation of this extraordinary research here, and find an interactive brain map here. Categories that activate the same brain area are shown in similar colors.
A Berkeley news story by Yasmin Anwar
says new insights into brain organization can help with diagnosis and
treatment of brain disorders, and further creation of brain-machine
interfaces. It may also be useful for facial and image recognition
recorded brain function of five volunteers using functional Magnetic
Resonance Imaging fMRI as they watched hours of movie clips. The story
explains they built a model of how 30,000 locations in the cortex
responded to each of 1,700 categories of objects and actions seen in the
movies. Then they used principal components analysis (CPA) a
statistical method that can summarize large data sets, to find the
"semantic space" common to the brains of all subjects.
A thoughtful Scientific American blog by Ben Thomas
observes the research shows we are skilled at relating our visual input
to "other chunks of reality to which our brains have assigned certain
characteristics." And he raises a profound question: If our brains use
association to define what an object or action is, does that suggest
"meaning" itself is just another word for "association?" He thinks more
understanding of semantic coding is needed to answer that question. But
whatever the answer, he writes, the research shows "even our most
abstract concepts depend on our own real-world experiences."
The paper by Huth and colleagues, "A continuous semantic space describes the representation of thousands of object and action categories across the human brain," appears in the December 2012 issue of the magazine Neuron.