Do baboons understand numbers?
|May 3, 2013||Posted by News under News|
If you look at the news headlines over at Science Daily for May 3, 2013, you’ll immediately see that their top story is: “Monkey Math: Baboons Understand Numbers.” Which, to put it quite bluntly, is pure poppycock.
The full title of the story linked to reads: Monkey Math: Baboons Show Brain’s Ability to Understand Numbers, which is equally misleading. The Science Daily news story is reprinted from a press release provided by the University of Rochester, a private research university in Rochester, New York, USA. The story’s opening paragraph makes the sensationalistic claim that olive baboons possess “the ability to understand numbers”:
Opposing thumbs, expressive faces, complex social systems: it’s hard to miss the similarities between apes and humans. Now a new study with a troop of zoo baboons and lots of peanuts shows that a less obvious trait — the ability to understand numbers — also is shared by humans and their primate cousins.
(For those who are interested, the citation for the study is:
Barnard A.M., Hughes K.D., Gerhardt R.R., DiVincenti L. Jr., Bovee J.M. and Cantlon J.F. (2013). Inherently analog quantity representations in olive baboons (Papio anubis). Frontiers in Comparative Psychology 4:253. doi: 10.3389/fpsyg.2013.00253.)
What can the baboons really do?
But if you actually read the story, you’ll find that all it says is that:
(i) Baboons can, without any prior training, discriminate between two cups containing different numbers of items (e.g. peanuts) and pick the cup that contains more items;
(ii) They’re very good at doing this when there’s a large relative difference between the number of items in the two cups (e.g. two peanuts versus seven peanuts), but they’re not so good when there’s a small relative difference (e.g. six versus seven);
(iii) They don’t need to count in order to make this discrimination – instead, they make use of what the researchers call an analog system, which allows them to make “more than” or “less than” discriminations when there’s a large relative difference between the number of items in two groups;
(iv) Young children who haven’t yet learned to count can do this too;
(v) According to the story, animals such as “birds, lemurs, chimpanzees, and even fish” can do this too, but this is the first time that non-human animals have been shown to be able to do it without undergoing any training;
(vi) Even with practice, the baboons don’t get significantly better at discriminating between the relative number of items in the two cups, showing that learning has little or nothing to do with it.
Why understanding numbers isn’t the same thing as making snap “numeric judgments”
What’s wrong with this story is that it confuses the ability to understand numbers with the ability to make numeric judgments, including the perceptual judgment that one set contains more (or fewer) items than another. Judgments of this sort are commonly referred to in the psychological literature as judgments of relative numerosity. The ability to understand numbers, on the other hand, is much more sophisticated. For starters, it presupposes an ability to count: you can’t properly be said to understand the concept of “five” unless you know that it’s the number after four, three, two and one. The baboons in the story above didn’t have to count, in order to judge which cup contained more peanuts: they were able to tell at a glance.
But the ability to understand numbers doesn’t stop with counting. As former Nature editor Stephen Budiansky points out in his best-selling book, If a Lion Could Talk: Animal Intelligence and the Evolution of Consciousness (The Free Press, New York, 1998), it also involves the ability to make judgments of a more general nature. “A seven-year-old child can tell you whether 173 is bigger than 142 without having to count from 1,” he writes, adding: “Older children implicitly grasp the notion that counting involves the possibility of extending the process indefinitely, but even young children do not need to be taught each number separately: a child who can count to 24 doesn’t need to be taught that there are numbers called 25 and 26” (p. 97).
Can any non-human animals count?
It has been widely alleged that Alex, the late African grey parrot trained by Dr. Irene Pepperberg, was able to count to six. However, this is highly doubtful, for reasons pointed out by science journalist Betty Lindsay in an article published in Nature’s Corner magazine (Issue 012). True counting, she observed, is a complex ability, which requires four things: first, the ability to produce a standard series of number tags (e.g. 1, 2, 3, …); second, the ability to give each item to be counted its own unique number tag; third, the ability to remember which items have already been counted; and finally, the ability to know that the last number tag indicates the total number of items. Lindsay then suggested that for quantities of four or less, Alex may be subitizing: in other words, he may simply perceive the number of objects at a glance, as we do when we count the number of spots on dice. Even Alex’s trainer, Dr. Pepperberg, has acknowledged that Alex need not be able to count: he may have simply been able to quickly and accurately estimate the number of items in a small collection, somewhat more rapidly than humans can. That’s an impressive judgmental ability, to be sure, but counting it ain’t.
Even more absurd is the risible claim, published in Science Daily back in 2005, that Alex has “a zero-like concept” – despite his trainer’s admission that he had never been taught the concept of zero as a quantity. Apparently, what Alex was able to do was call out the word “None,” when he was unable to see any group of identically colored objects on a tray, having the same number of items as a number called out by the trainer. Did he have a “null concept”, as his trainer suggested? Or did he simply mean: “Can’t answer that”?
Getting back to the baboons: what the experiment in the study cited above showed was that they are able to make judgments of relative numerosity: they can tell at a glance which set contains more or fewer items. It turns out that this ability is found in a wide variety of animals, including some that are not even conscious! Fish have this ability too: a recent study found that female mosquitofish were able to discriminate between two shoals with a 1:2 numerosity ratio (2 vs. 4, 4 vs. 8 and 8 vs. 16 fish) but were unable to discriminate a 2:3 ratio (8 vs. 12 fish). However, Professor James D. Rose (Department of Zoology and Physiology, University of Wyoming) has concluded on scientific grounds that fish are totally lacking in consciousness: they can respond to events occurring in their surroundings without being consciously aware of anything. In case readers are wondering how Rose arrived at his conclusion that fish are not conscious, here are the references:
(1) The Neurobehavioral Nature of Fishes and the Question of Awareness and Pain (Reviews in Fisheries Science, 10(1): 1-38, 2002).
(2) “Can fish really feel pain?” by J. D. Rose, R. Arlinghaus, S. J. Cooke, B. K. Diggles, W. Sawynok, E. D. Stevens, and C. D. L. Wynne (Fish and Fisheries, published online 20 December 2012, doi: 10.1111/faf.12010).
The neurological basis of animals’ ability to make judgments of relative numerosity is now fairly well understood, and has been described in an recent article by Manuela Piazza and Veronique Izard, entitled, How Humans Count: Numerosity and the Parietal Cortex (The Neuroscientist, Volume 15 Number 3, June 2009 261-273). However, the neurological basis of the human ability to grasp the concept of a number is far more mysterious, as the authors acknowledge:
In humans, during development, with the introduction of symbols for numbers and the implementation of the counting routines, the parietal system undergoes profound (yet still largely mysterious) modifications, such that the neural machinery previously evolved to represent approximate numerosity gets partially “recycled” to support the representation of exact number.
Are baboons really as smart as three-year-old children?
The story about the baboons in Science Daily also claimed, incorrectly, that the baboons’ ability to make “more than” or “less than” discriminations is only found in human children aged three and over, implying that the baboons are as smart as three-year-olds. Study co-author Jenna M. Bovee, described in the story as “the elephant handler at the Seneca Park Zoo who is also the primary keeper for the baboons,” was quoted as saying:
“A lot of people don’t realize how smart these animals are. Baboons can show you that five is more than two. That’s as accurate as a typical three year old, so you have to give them that credit.”
This is simply wrong. Recent research has shown that even 6-month-old infants can successfully discriminate between two sets and identify which one has more items, so long as the ratio of the number of items in the two sets exceeds 2:1 (see here, here). What’s more, this ability holds regardless of how big or small the two sets are, according to recent research (see here). In other words, even a 6-month-old human baby can “show you that five is more than two,” as study co-author Jenna Bovee put it.
Now, Ms. Bovee has a Bachelor’s degree in zoology, but no formal qualifications in psychology (see here). It would be extremely uncharitable to blame her for making factual errors about infants’ cognitive abilities: after all, that’s not her specialty. The real responsibility for the error lies with the University of Rochester, which authorized the press release without bothering to check the accuracy of her assertions about infants’ cognitive abilities. That is extraordinarily careless.
So in conclusion, what are we left with, after all this brouhaha about monkeys that can understand numbers? All the story shows is that untrained baboons are reliably able to judge which of two sets of objects contains more items, so long as the ratio of the number of items in the two sets remains sufficiently large (e.g five items versus two). But it turns out that this ability is found even in fish, which are not even conscious! It’s also found in six-month-old infants – a fact which refutes the claim made by one of the authors of the study, that the baboons accomplished a feat that human children are not capable of until the age of three.
In short: the Science Daily story of baboons that can grasp the concept of a number might best be described as “much ado about nothing.” Methinks it is time for science reporters to show a little more skepticism.
NOTE: News would like to gratefully acknowledge that the picture of an olive baboon shown above was taken by Muhammad Mahdi Karim (www.micro2macro.net). The picture is a Wikipedia image.