[Comp-neuro] Grandmother cells a failed concept? Or is Barlow right that grandmother cells exist and “can now be recorded from and studied reliably?"

Asim Roy ASIM.ROY at asu.edu
Sun Feb 16 07:29:42 CET 2020

Ann-Sophie Barwich recently published an article in Frontiers in Neuroscience titled “The Value of Failure in Science: The Story of Grandmother Cells in Neuroscience” (https://www.frontiersin.org/articles/10.3389/fnins.2019.01121/full) where she argues for the need for failure analysis in science and, in the context of neuroscience, analyzed the grandmother cell concept as a failed one. Although I was one of the reviewers who endorsed the article for publication (publication of alternative viewpoints is essential in science), I felt that Barwich’s arguments against grandmother cells were deeply flawed. I point out these flaws in a Commentary in Frontiers in Neuroscience: “Commentary: The Value of Failure in Science: The Story of Grandmother Cells in Neuroscience:” https://www.frontiersin.org/articles/10.3389/fnins.2020.00059/full

I think Reddy and Thorpe’s 2014 article clears the confusion about concept cells. With Reddy and Thorpe’s characterization of concept cells, one would find no difference between concept cells and grandmother cells. And that’s essentially what I have done in the Commentary – show that concept cells and grandmother cells have identical properties. I also point out in the Commentary that it’s a false claim by Barwich (and in other prior articles that Barwich references) that concept cells can be part of a sparse population coding scheme. Given that concept cells have “meaning” on a stand-alone basis, they can’t be part of any population coding scheme. And it’s also a false claim in Barwich that grandmother cells do not use associative learning, because grandmother cells have always been characterized as “multimodal” going all the way back to Gross (2002). In addition, the neurophysiological evidence on multisensory neurons clearly point to a single-cell based abstract cognitive system in the brain, as claimed by Roy (2017).

Overall, concept cell findings confirm the prediction of Barlow (2009)<https://www.frontiersin.org/articles/10.3389/fnins.2020.00059/full#B1> that grandmother cells exist and “can now be recorded from and studied reliably.”


Barlow, H. B. (2009). “Grandmother cells, symmetry, and invariance: how the term arose and what the facts suggest,” in The Cognitive Neurosciences, 4th Edn., ed M. Gazzaniga (Cambridge, MA: MIT Press), 309–320.
Reddy, L., and Thorpe, S. J. (2014). Concept cells through associative learning of high-level representations. Neuron 84, 248–251. doi: 10.1016/j.neuron.2014.10.004
Roy, A. (2017). The theory of localist representation and of a purely abstract cognitive system: the evidence from cortical columns, category cells, and multisensory neurons. Front. Psychol. 8:186. doi: 10.3389/fpsyg.2017.00186

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.tnb.ua.ac.be/pipermail/comp-neuro/attachments/20200216/e5eb859d/attachment.html>

More information about the Comp-neuro mailing list