Adrian and Rich, (01)
Last week, John Bottoms sent a note to Ontolog Forum about
that web site. Following is a note I sent in reply. (02)
John (03)
-------- Original Message --------
Subject: Re: [ontolog-forum] Semantic mapping of the brain
Date: Thu, 20 Dec 2012 10:12:06 -0500
From: John F Sowa (04)
John, (05)
Thanks for the pointer to the video and the web site. For anyone who
would like more detail, the authors of the article in _Neuron_ have
posted a preprint on their web site: (06)
http://gallantlab.org/publications/2012.Huth.etal.SMALL.pdf (07)
Unfortunately, the images cited in the article are all gathered
together at the end of the PDF. (08)
JB
> It is impressive work, however, I think of it more of a taxonomy
> viewer than a semantic viewer. It seems to be missing any notion
> of the control system behind the brain mappings. (09)
I'm sure that the authors would love to have much more detail
about all those topics, but it's extremely difficult to get more. (010)
They partitioned the brain into 30,000 voxels, each of which
contains millions of neurons. The information from each voxel
only shows how much blood is flowing to that region for an
association to a word selected from WordNet. (011)
This procedure is a big advance over the methods available in the
20th century. The experimenters would certainly love to know what is
happening at the level of individual neurons and the interconnections
among them. But that would require a million times finer resolution. (012)
Any known methods for digging deeper into the brain are *extremely*
invasive. Neuroscientists plant electrodes that can tap into single
neurons of rats, cats, and monkeys. A lot of data is gathered from
humans as a byproduct of brain surgery. But it's unethical to dig
deeper for experimental purposes. (013)
JB
> To use the brain explorer tool go to: http://gallantlab.org/
> and click on "Brain VIewer" (014)
That's an interesting toy to play with. You can click on any part
of the WordNet hierarchy and see what parts of the brain light up.
But many voxels scattered all over the brain light up for each term. (015)
For a very brief intro and summary of what those areas do, how they're
interconnected, and how they're related to semantics, see Section 2
(slides 13-45) of http://www.jfsowa.com/talks/goal.pdf (016)
Slide 36 (from a book by Peter MacNeilage) shows the information flow
among major areas of the cortex during speech generation. This diagram
was derived from a very large number of studies and sources. When you
see different parts of the brain lighting up in "Brain Viewer", that
is the result of many kinds of operations going on. Disentangling
them to determine what each part contributes is not easy. (017)
JB
> Maybe they could use some ontology assistance. (018)
People working in cognitive science have been collaborating for years.
As an example, following is a review of the book by MacNeilage by
another cognitive scientist (Bart de Boer): (019)
http://uvafon.hum.uva.nl/bart/publications.html (020)
Go up a level in that directory for other papers by de Boer. (021)
John (022)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J (023)
|