Connectionists: Grandmother cells are widely used in the brain - the brain is highly readable and transparent at higher levels

[Asim Roy]

The following article appeared in Frontiers of Cognitive Science on May 24:



Roy A. (2013). “An extension of the localist representation theory: grandmother cells are also widely used in the brain.”



It’s a very short article, just 2000 words. Here’s the link to the article: http://www.frontiersin.org/Cognitive_Science/10.3389/fpsyg.2013.00300/full

Single cell recordings of the past decade have produced a substantial body of evidence for grandmother cells.


Phys.org (MedicalXpress) reported the grandmother cell findings in a story titled: “If you can't beat them, join them: Grandmother cells revisited.” Here’s the link to the story: http://medicalxpress.com/news/2013-05-grandmother-cells-revisited.html



Jerzy Konorski (http://en.wikipedia.org/wiki/Jerzy_Konorski ) and Jerry Lettvin of MIT (http://en.wikipedia.org/wiki/Jerome_Lettvin<http://en.wikipedia.org/wiki/Jerome_Lettvin>) were the original proponents of the grandmother cell idea. Horace Barlow (http://en.wikipedia.org/wiki/Horace_Barlow) of Cambridge University, UK, remains to this day a firm believer in grandmother cells. His 2009 paper "Grandmother Cells, Symmetry, and Invariance: How the Term Arose and What the Facts Suggest" claims grandmother cells "exist" and “can now be recorded from and studied reliably.”


There is some misconception about grandmother cells. The concept of grandmother cells is not just about a single cell encoding (representing) a person or an object. It's much more than that. It's really about the brain's ability to abstract complex concepts (e.g. concepts of categories such as cats, dogs, animals, humans, trees, buildings, anger, laughter and so on) and encode (represent) each using a single cell. And that ability exists in other animals too, not just humans. The evidence for the brain's ability to abstract complex concepts is presented in this paper.



With grandmother cells, mind reading becomes very simple. In fact, the brain is amazingly transparent at some level.



Grandmother cells are a special type of localist cells. In localist representation, units (cells) simply have "meaning and interpretation." Grandmother cells represent complex concepts that are multimodal invariant.



The localist representation theory article appeared in Frontiers of Cognitive Science on Dec. 4, 2012:

Roy A. (2012). “A theory of the brain: localist representation is used widely in the brain.”



It’s also a very short article, just 2000 words. Here’s the link to the article:

http://www.frontiersin.org/Journal/FullText.aspx?s=196&name=cognitive_science&ART_DOI=10.3389/fpsyg.2012.00551


Phys.org (MedicalXpress) wrote a feature story on the localist theory titled: “Do brain cells need to be connected to have meaning?” Here’s the link to the story: http://phys.org/news273783154.html

James McClelland of Stanford University and David Plaut of Carnegie Mellon University presented arguments against the localist theory in that Phys.org story. I posted a separate response to their arguments on the Lifeboat site. Here’s the link to the response:

http://lifeboat.com/blog/2012/12/response-to-plaut-and-mcclelland-on-the-phys-org-story


With best regards,
Asim Roy
Arizona State University
Tempe, Arizona
www.lifeboat.com/ex/bios.asim.roy<http://www.lifeboat.com/ex/bios.asim.roy>





-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/connectionists/attachments/20130527/247d5ba1/attachment.html>