What's in a Name?
“What’s in a name?” Juliet asks in William Shakespeare’s famous play. When it comes to how humans store information on the names of objects and events, recent studies in the field of cognitive neuroscience reveal that she was right to focus on the rose’s sweet smell. Leonardo Fernandino, PhD, assistant professor of neurology, published results in PNAS showing that sensory-motor experiences were more predictive of brain activation patterns than two other types of information structures when study subjects considered names of common things. Dr. Fernandino’s team measured which regions of the brain activated when 44 participants read names of more than 600 common objects and concepts on a screen while undergoing functional magnetic resonance imaging (fMRI) scans.
Dr. Fernandino’s study compared brain activation models based on experiences of objects and events with two other leading hypotheses of how concepts are stored in the brain. Proponents of the taxonomic model believe that learning a concept consists of associating an arbitrary neural pattern to a particular kind of thing in the world, such that the neural representation of a natural kind is essentially a symbol that stands for that category of things. Thus, one would recognize something as a tree by activating the symbolic representation for “tree” in their brain. Other scientists focus more on the brain’s ability to associate specific words based on how often they appear together during natural language use. This is how machine learning is used to train computer programs about words, and by processing enough text the software begins to determine that “wolf” is more similar to “dog” than to “cat,” or that “peeling” goes with “orange” but not with “watermelon.” This type of information is called distributional because it is based on how a word is distributed among others in speech or text.
By directly comparing experiential, taxonomic and distributional models of concept representation, Dr. Fernandino’s team found strong evidence that experiential information is fundamental to further understanding how conceptual knowledge is stored and retrieved. Our sensory, motor, spatial, temporal and emotional memories form a critical part of “what” is in a name and how our brains remember it.
The study was co-authored by Drs. Tony Tong, Lisa Conant, Colin Humphries and Jeffrey Binder from MCW's neurology department. Funding was provided by the National Institute on Deafness and Other Communication Disorders and the Advancing a Healthier Wisconsin Endowment at MCW.
Research and narration by Leonardo Fernandino, PhD
Artwork, animation, and soundtrack by Michael Shank
Dr. Fernandino’s study compared brain activation models based on experiences of objects and events with two other leading hypotheses of how concepts are stored in the brain. Proponents of the taxonomic model believe that learning a concept consists of associating an arbitrary neural pattern to a particular kind of thing in the world, such that the neural representation of a natural kind is essentially a symbol that stands for that category of things. Thus, one would recognize something as a tree by activating the symbolic representation for “tree” in their brain. Other scientists focus more on the brain’s ability to associate specific words based on how often they appear together during natural language use. This is how machine learning is used to train computer programs about words, and by processing enough text the software begins to determine that “wolf” is more similar to “dog” than to “cat,” or that “peeling” goes with “orange” but not with “watermelon.” This type of information is called distributional because it is based on how a word is distributed among others in speech or text.
By directly comparing experiential, taxonomic and distributional models of concept representation, Dr. Fernandino’s team found strong evidence that experiential information is fundamental to further understanding how conceptual knowledge is stored and retrieved. Our sensory, motor, spatial, temporal and emotional memories form a critical part of “what” is in a name and how our brains remember it.
The study was co-authored by Drs. Tony Tong, Lisa Conant, Colin Humphries and Jeffrey Binder from MCW's neurology department. Funding was provided by the National Institute on Deafness and Other Communication Disorders and the Advancing a Healthier Wisconsin Endowment at MCW.
Research and narration by Leonardo Fernandino, PhD
Artwork, animation, and soundtrack by Michael Shank