Brains Hard-Wired for Math

mcgrew writes "New Scientist is reporting that "non-human primates really can understand the meaning of numerals." The small study of two rhesus monkeys reveals that cells in their brains respond selectively to specific number values — regardless of whether the amount is represented by dots on a screen or an Arabic numeral. For example, a given brain cell in the monkey will respond to the number three, but not the number one. The results suggest that individual cells in human brains might also have a fine-tuned preference for specific numerical values." The report itself is online at PLoS Biology, Semantic Associations between Signs and Numerical Categories in the Prefrontal Cortex."

Not just math

[biocute]

Bottom of the friendly article: The results are not the first to suggest there may be specific brain cells tied to individual concepts. In 2005 researchers discovered that individual neurons become activated by images of specific celebrities such as Jennifer Aniston and Halle Berry.

So I guess it is up to individuals to decide how best to utilize limited brain cells. I'm pretty sure that those monkeys can tied a couple of their brain cells to other concepts given enough training.

Obvious

[Stoutlimb]

In other news, reality is hardwired for math.

Seriously, why wouldn't a brain, which exists to process data in one form or another, respond to math positively at some level? Geometry is math, and that is hardwired in our brains to a high level. Any brain that has to process spacial information in any way must be predisposed to math.

music and singing

[xPsi]

A colleague of mine once pointed out that the ability of most humans to sing (speak for yourself!), play music, and even distinguish different tunes implies an intrinsic hard-wired affinity for numbers since music depends on very specific ratios of frequencies to be gauged and produced accurately real time. You are in effect doing a Fourier transform of the music, finding the strongest peaks, and reproducing them and/or scaling them by fairly exact amounts (in spite of a broad spectrum of other frequencies present creating timbre). On top of that, one is usually doing this accurately in the context of much, much lower frequencies (i.e. rhythms/tempos on the scale of Hertz rather than "tones" on the scale of 100s of Hertz) as well. Of course, not all music is western, 12 tone, tuned the same, etc., etc. etc. But I think there may still be a (fairly well understood??) psycho-acoustic music-math connection in there.

Re:Base?

[xPsi]

I wonder if the brains are wired for specific bases, like base 10.

It is possible, but I'm guessing this is mostly a matter of familiarity and convention. For example, Baylonians used a sexagesimal [wikipedia.org] (base 60 -- a.k.a. "thanks for frickin' 360 degrees guys"...) system. As many programmers know (do I even need to say it on ./?), base 2, 8, and 16 can become second nature pretty quickly with some practice and application.

What about the Pirahã?

[settrans]

The notion that primates are genetically predisposed to have mathematical ability is tenuous. Why should we believe there is some neural circuitry designed explicitly for math? First of all, all studies teaching non-human primates to count involve extensive training of the primates; it doesn't just "click" for them. This would suggest that it is a struggle for them to learn the concept of counting and mathematics. (Of course it doesn't help that TFA is extremely light on the gory details of the methodology and results of the study.)

Secondly, the Pirahã people [wikipedia.org] of Amazonia do not have numbers or counting. Professor Everett, despite months of instruction, was unable to make any progress in teaching them how to count. The Pirahã themselves were highly motivated learners, as they didn't want to be ripped off in trade by visiting merchants, but nevertheless, they had no success in learning the most basic concepts of math. Indeed the Pirahã language has no numerals, and is claimed to have no quantifiers, either.

Relevant readings:
Everett, D.L. (2005). Cultural constraints on grammar and cognition in Pirahã. Current Anthropology, 46, 621-646.
Hauser, M.D., Chomsky, N. and Fitch, W.T. (2002) The faculty of language: what is it, who has it, and how did it evolve? Science, 298, 1569-1579.
Pinker, S. & Jackendoff, R. (in press). The components of language: What's specific to language, and What's specific to humans? In M.H. Christiansen, C. Collins & S. Edelman (Eds.), Language universals. New York: Oxford University Press.

Re:Obvious

[Lemmy Caution]

I agree...

I can't stand the over-use of the expression "hard-wired" when the data only indicates something that is universal. It implies that the structures responsible would develop in that function no matter what, without the experience in the world of, for example, things in sets-of-three, etc.

The data really supports dynamical systems models of cognitive development [indiana.edu] more than pure innatist ones. Just look at what the brain of someone blind from birth develops into, absent visual input.

I highly recommend the books of Andy Clark, particularly his "Being There," as an introduction that starts to explain just how flawed the seemingly harmless phrase "hard-wired" is.

Re:Numbers or numerals?

[jandersen]

It isn't surprising that monkeys can understand an abstraction like 'numbers' - a brain is a neural network, and neural nets are 'abstraction engines' by definition. Consider the nature of abstractions: an abstract concept is one that describes a set of properties that are common to a class of objects. A number, for example, is the property that is common to all sets that are isomorphic in the category of sets (to spell it out: what is common to 'five apples', 'five oranges', 'five cows', ...? The number 5, of course). And what is it a neural net does? It learns to recognise patterns that are shared by all the 'objects' it 'sees' (if you will excuse the metaphor) - in other words, it creates an abstraction.

The numbers 1 and 0, although fundamental to our numerical notation, are not really 'interesting' in nature - 0 is simply 'nothing' and 1 is 'anything', they sort of fade into the background. Being able to recognise other, small numbers can be useful, however. Two fruits is one for me and one for you; if you have four children, but can only see three, then you should go looking for the last one, etc etc.

This is the way evolution works - nothing evolves with any purpose; things evolve because there are new traits that turn out to be beneficial in the given environment. And then, down the line, it sometimes also turns out that a trait that evolved at some point in the past allows the organism to do something entirely new in a new environment. So the monkeys didn't evolve to benefit from written language, it turned out that this is one of the things their brains can learn. The real question here is: Why did brains evolve - and that all starts with biofilms ;-)

Re:Base?

[SQLGuru]

I can count to 1023 on my fingers.....I especially like the number 132.

http://www.intuitor.com/counting/ [intuitor.com]

Layne

Re:music and singing

[xPsi]

Specific intervals are pleasing largely because of the way their overtones line up; that's why pretty much every music system has a third, a fifth and an octave. I'd bet that producing music is done based on memory and calibration, the same way many other actions are done; no math involved.

Point well taken. As you probably know, that what is considered a pleasing tone is very culturally dependent. Most of the world's music involves what to "western ears" sounds microtonal -- but perhaps 3rds, 5ths, and octaves are universal, I'm not sure. IF this were true, it would signal to me that there IS a hardware component to at least detecting (and reproducing) certain mathematical ratios. On the other hand, most of western music is mean tempered and only approximates perfect 3rds, and 5ths. Nevertheless, I also point out that the very notion of calibration (if that is what we are doing) is an intrinsically mathematical process involving mutiplicative or additive scaling. If we are able to do that real time, that is again a signal for a wetware math processor. However I agree using memory may or may not be mathematical (at least in the way were are talking about here).