Archive for the ‘Uncategorized’ Category

Recalling whether or not you can recall

Friday, May 27th, 2005

Can our brain sense that it is going to be able to recall something before it actually does?

Ken Jennings’ brain can. Jennings (picture) won more than $2,000,000 in 72 straight appearances on the US game show Jeopardy. In this game players “ring in” if they think they know the answer, then have five seconds to give that answer. Jennings, in many cases, is clearly ringing in before he knows the answer, then often takes nearly the entire five seconds to come up with the inevitably correct answer.

It thus appears that Jennings has the metacognitive ability to sense whether or not he knows the answer before that answer has actually been retrieved from the recesses of his brain. The explanation that he simply rings in on topics which he knows well (movies in his case) is too simple. It seems that something more sophisticated is at work—he appears to actually know whether or not he knows something before he has fully retrieved that knowledge.

In computer science terms, perhaps Ken’s strategy could be compared to “precompiling” a database query, or calculating retrieval cost in advance.

What is the model for human memory that could explain this? Is the “do-I-know-it” recall act a rougher, abbreviated, accelerated version of full retrieval, or is it a different process altogether, possibly accessing an alternative, compact, pre-indexed “high-speed” version of the knowledge?

Quiet Oboes

Monday, May 16th, 2005

Myra Burg makes these fabulous objects, tubes layered in incredible, rich varieties of colors and textures. She calls them “Quiet Oboes”.

Eide Neuroleaning Blog

Friday, May 13th, 2005

I’d like to introduce Numenware readers to the Eide Neurolearning Blog, one of the best out there. It has everything a blog should: well-selected content, engaging writing, good links, frequent posts, and nice pictures. If you’re interested in the brain and learning, read this.


Saturday, May 7th, 2005

What is the relationship between music and the brain? A leading theoretician on that topic was Dr. Gordon Shaw (left), who died last month, and was most famous for discovering the so-called Mozart effect (addtional link), which evolved into the folk “meme” which claimed that listening to classical music makes your smarter. Before long, people were playing Mozart to their babies in their cribs. In fact, what Shaw had shown was simply that listening to Mozart improved performance on spatio-temporal tasks for ten minutes.

But the Mozart business overshadowed the immense body of ground-breaking research that Shaw carried out. Working at the University of California at Irvine, Shaw focused on cortical organization, developing his unique, columnarly-based trion model. A list of his papers is on the web site of the MIND Institute, the group Shaw founded to continue his brain research and explore applications to elementary education, in the form of the Math+Music program which combines non-language based computer math games with specialized piano training.

Shaw’s model for the architecture of the cortex was set forth in his paper entitled “Model of cortical organization embodying a basis for a theory of information processing and memory recall.” The abstract states:

Motivated by V. B. Mountcastle’s organizational principle for neocortical function, and by M. E. Fisher’s model of physical spin systems, we introduce a cooperative model of th cortical column incorporating an idealized substructure, the trion, which represents a localized group of neurons. Computer studies reveal that typical networks composed of a small number of trions (with symmetric interactions) exhibit striking behavior—e.g., hundreds to thousands of quasi-stable, periodic firing patterns, any of which can be selected out and enhanced with ony small changes in interaction strengths by using a Hebb-type algorithm.

I’m wondering how Jeff Hawkins managed to write an entire book about cortical architecture without mentioning Shaw’s work.

A particular intriguing aspect of Shaw’s theory is that humans love music because it resonates with the innate columnar cortical structure. Xiaodan Leng then derived music directly from these theories, yielding eerily human-sounding, classical-like pieces; get your MP3s here!

Neurotheologically, what conclusions can we draw from Shaw’s insights? In the West, we think of religion as having a heavy musical component; after all, every cathedral has its organs, and Bach’s “religious” compositions tickle those trions of yours every bit as well as Mozart does, but this focus on music, at least of the cerebral kind, may be peculiar to Christianity. You don’t hear people talking too much about “Buddhist music”. Perhaps in a neurotaxonomy of established religions, Christianity occupies a position closer to the cortex.


Monday, March 28th, 2005

In biology, the reduction in size of an organ or part (as in the return of the uterus to normal size after childbirth, or the shrinking of the childhood organ known as the thymus); can also refer to something very elaborate or complicated, such as a grammatical construction.

Inventions of the year

Monday, January 31st, 2005

In addition to my many other hats, I’m an inventor. For a bit of year-end diversion, here are three inventions from this year:

1. Magnetic paper money. Place a weak magnetic strip along the narrow edge of a note of currency. This not only allows the money to stick together—instant money clip!—but also lets you stick it on your refrigerator door. In fact, my two US$1 bill prototypes are stuck on my refrigerator door right now.

Amazingly, this seems to not be the subject of any current US patent.

2. Holographic turn signals . Project cars’ turn signals holographically 10 ft. up above the car, so that following cars can see it even if there is another car in-between.

3. Self-announcing wet paint . Using nanoparticle technology, have wet paint automatically display the message “wet paint” over and over. When it dries, the message automatically disappears.

2005 was also the year in which my invention for “microlinking” images on a web page with text was awarded a US patent (see image), within the context of presenting game commentary and educational materials for strategy games such as chess or Go.

News flash: Mental Processing Is Continuous, Not Like a Computer

Sunday, January 30th, 2005

Michael Spivey (pictured) and his collaborators at Cornell did a cute little study, where they tracked the movement of the subject’s mouse as they tried to move it toward a picture on the computer screen corresponding to the word they had just heard. When the two pictures were of items whose lexical representations are “close” (think “candy” vs. “candle”) the mouse darted tentatively toward one before converging on the correct one. As the authors say in the abstract, “competition between partially active lexical representations are revealed in the shape of the movement trajectories”.

This seems like a useful thing to do, while hardly qualifying as a breakthrough.

But does Spivey really believe the stuff they quote him as saying in the brain-damaged press release put out by Cornell? He supposedly said

For decades, the cognitive and neural sciences have treated mental processes as though they involved passing discrete packets of information in a strictly feed-forward fashion from one cognitive module to the next or in a string of individuated binary symbols—like a digital computer.”

No they didn’t. That was one, discredited, easily refuted model. He goes on:

In [our new] model, perception and cognition are mathematically described as a continuous trajectory through a high-dimensional mental space; the neural activation patterns flow back and forth to produce nonlinear, self-organized, emergent properties—like a biological organism.

Well, gee whiz. Maybe that’s because it is.

PS. It does not seem altogether irrelevant that Spivey is apparently the inventor of the Spivey Test, where instead of computers trying to act like humans, humans try to act like computers.


Wednesday, January 26th, 2005

Providing introductory or preparatory instruction.

Neuroscience resources

Wednesday, January 26th, 2005

Today let me introduce you to some neuroscience resources I’ve found useful.

Mind Hacks is a website linked to the popular book of the same name, by Tom Stafford and Matt Webb. It’s readable, fun, and informative. These guys must be good, since they were smart enough to introduce Numenware to their readers.

Although not specifically about neuroscience, The Loom is a general-interest science blog by Carl Zimmer, a popular science writer whose most recent book is Soul Made Flesh, an account of the origins of neurology in 17th century. This site is far from mere daily musings: each article is well-researched and well-written but not too long. My favorite recent article is one on the fantastic diversity of beetle horns and how they evolved. Carl is really leveraging the blog format in an interesting way to extend the ways in which he can get his ideas out and connect with readers.

Carl’s critical review of Dean Hamer’s The God Gene is worth a look.

Scientific American Mind is an irregularly published new magazine whose subtitle is “Thought, Ideas, Brain Science”. The most recent issue has interesting articles on topics such as sign language, and how it is parsed and generarated using Wernicke’s and Broca’s areas, just like spoken language; neurolinguistic programming; Koch’s consciousness research; neuromarketing; and magnetic brain stimulation, with today’s most disturbing factoid: a DoD project to develop a TMS helmet that will animate exhausted soldiers back into battle.

Scientific American desperately needs a new web strategy. They think the web is nothing more than a way for people to buy articles for $3 each, after being allowed to read one or two paragraphs. I don’t think so. At least put some content up to make the site more bookmarkable and visitable. Or start a blog.

Increasing fMRI resolution

Tuesday, January 25th, 2005

Neuroimaging techniques are limited by their spatial and temporal resolution. fMRI has a spatial resolution of about 3-4mm. Now researchers have developed statistical techniques to narrow fMRI resolution down to less than a millimeter, fine enough to see the “orientation columns” in the primary visual cortex (V1), so named because each reacts to lines and edges at a particular orientation.

The research is described in an article in Nature Neuroscience, Decoding the visual and subjective contents of the human brain. The authors conclude their abstract with the observation

Our approach provides a framework for the readout of fine-tuned representations in the human brain and their subjective contents.

Much more of science than people realize is about computers processing massive amounts of data. Theories of the cosmos are basically just computers crunching numbers; the Hubble photos we see are massively computer enhanced. From an information theoretic point of view, it’s highly likely that orders of magnitude of additional useful information can be extracted from even the current generation of fMRI hardware, given the right algorithms. That’s the approach these researchers have taken, highlighting the importance of the field called computational neuroscience.

V1 is the part of the visual cortex at the back of your head which takes the first crack at signals coming in from your eyeballs. It’s just one synapse away from the retina. The interesting thing about this research was not just the ability to look at individual orientation columns, but how this ability was used. Subjects were told to focus on one of two overlapping grids of lines of different orientations. The researchers then used the enhanced scanning technique to verify that the relevant orientation columns were activated. This demonstrates how higher-level, conscious functioning (“attention”) can drive lower-level brain functions—the sort of top-down mechanism that Jeff Hawkins focuses on in On Intelligence.

I do have to object to the headline the NYT came up with for its article about this news: “Improved Scanning Technique Uses Brain as Portal to Thought”.

Now, we just have to hone in on the “God” columns in the temporal lobe, or wherever they are.