October 16, 2007

Our mental number line?

Now here’s a post that I’ve been meaning to write for a while. I finally found some time (or motivation) to get it done I guess :)

Firstly, I should state that my research interests mostly lie within the field of social psychology (specifically affiliation and social rejection). But I also have some general interests in topics across a variety of other psychology sub-fields (as you can probably tell, I have a rather large interest in evolutionary psychology, which to me is more of a way of thinking about psychology than a sub-field). Well, this is a post about a general interest of mine, and that is the concept of number. Specifically, this post is going to be about human and animal abilities to order and represent numbers.

Jessica Cantlon and Elizabeth Brannon (2006) are researchers from Duke (you can find Brannon’s website here) who worked on two experiments investigating: 1) the ability of two rhesus macaques to apply a learned ordinal numerical rule, and 2) compared the performance of these same monkeys with that of humans on ordering pairs of numbers.
In the first experiment, the authors trained the rhesus macaques to present in ascending order all possible pairings of stimuli, of which each represented a numerical value between 1 and 9.
The stimuli consisted of a background with 1 through 9 square shapes in various colors and sizes, kind of like the blue boxes in the photo below:

Once the ordinal rule was learned, the macaques were then presented with stimuli that represented novel numerical values (10, 15, 20 and 30). Of the experimental sessions, there were three types of pairings. The first type were familiar-familiar pairs where each stimulus represented a value between 1 and 9. The second type were familiar-novel pairs, in which one stimulus represented a value among 10, 15, 20 or 30 and the other stimulus represented a value between 1 and 9. And lastly, the third type of pairing was novel-novel. This is where both stimuli represented a value among 10, 15, 20 and 30. Once presented with the pairings, the macaques were expected to use the ordinal rule and order the values from smallest to largest, which is what the authors found!

The monkeys ordered these values in the same direction that they originally learned to order the values between 1 and 9.
The findings are particularly interesting considering that the authors used a numerical value up to 30 and previous studies have only presented monkeys with values of up to 10. So not only does this first experiment show that rhesus macaques can compare and apply ordinal rules to unfamiliar values, but these monkeys can also represent and compare values of at least 30 with the possibility of no upper limit!

The second experiment was conducted to compare the performance of these monkeys with human participants and specifically to test how much the number comparisons are controlled by Weber’s law in each species (Weber’s law essentially states that the ability to discriminate between two values depends on the ratio between the two and not the actual absolute values.
For instance, it is easier to discriminate between the values of 1 and 9 than it is to discriminate between the values of 28 and 29).

The same macaques and 11 university students were instructed to pick the smallest value presented out of pairings of the same stimuli.
This time each stimulus represented a numerical value between 2 and 30. The human participants were instructed not to count, but to respond as quickly as possible once the stimuli appeared. The investigators observed that the monkeys and the human participants performed similarly in accuracy. Indeed, the difference between monkeys and humans in accuracy was smaller than the difference between the most and least accurate humans! The monkeys did perform significantly faster than the humans though. This could be attributed to the training, but remains unclear. However, the data from both the human and monkey groups did conform to the pattern predicted by Weber’s Law. Accuracy decreased and reaction time increased as the ratio between the small and large values increased. Whenever the values increased and the distance between them decreased, it took both humans and monkeys more time to order them.

Now, I thought that this was an intriguing article when I first read it last year.
It seems to support the supposition that humans and animals use a similar system for representing and ordering numbers. We seem to have a mental number line in common. Perhaps we evolved an underlying mechanism for non-verbal number representation from a common primate ancestor? Well, anyways, I thought the article was pretty cool and just wanted to share it, haha.


Cantlon, J., & Brannon, E. (2006). Shared system for ordering small and large numbers in monkeys and humans.

October 11, 2007

Self-replicating robots . . . and more

So I was just browsing around for some cool videos and what not, and guess what I found? If you guessed a short talk given by Hod Lipson, then you'd be correct!

As you might recall, I wrote about Lispon back in June
concerning his plenary address at HBES 2007. As I wrote before, he was my favorite speaker over the course of the conference. His topic was evolutionary robotics, which is basically about creating robots that:

"decide for themselves how they want to walk; robots that develop a sense of what they look like; even robots that can, through trial and error, construct other robots just like themselves"
It is really a very interesting area that may even help us understand how the human brain constructs self-models. Is it similar to how these robots develop a sense of themselves?

Well, anyways, Lispon gave a brief version of the same address, with the same video clips, as a TED Talk on October 11, 2007. If you don't feel like visiting the website, I have posted the video below. It's definitely really cool and worth the watch: