Limits...
The benefit of symbols: monkeys show linear, human-like, accuracy when using symbols to represent scalar value.

Livingstone MS, Srihasam K, Morocz IA - Anim Cogn (2010)

Bottom Line: When humans and animals estimate numbers of items, their error rate is proportional to the number.To date, however, only humans show the capacity to represent large numbers symbolically, which endows them with increased precision, especially for large numbers, and with tools for manipulating numbers.This ability depends critically on our capacity to acquire and represent explicit symbols.

View Article: PubMed Central - PubMed

Affiliation: Department of Neurobiology, Harvard Medical School, Boston, MA 02115, USA. mlivingstone@hms.harvard.edu

ABSTRACT
When humans and animals estimate numbers of items, their error rate is proportional to the number. To date, however, only humans show the capacity to represent large numbers symbolically, which endows them with increased precision, especially for large numbers, and with tools for manipulating numbers. This ability depends critically on our capacity to acquire and represent explicit symbols. Here we show that when rhesus monkeys are trained to use an explicit symbol system, they too show more precise, and linear, scaling than they do using a one-to-one corresponding numerosity representation. We also found that when taught two different types of representations for reward amount, the monkeys systematically undervalued the less precise representation. The results indicate that monkeys, like humans, can learn alternative mechanisms for representing a single value scale and that performance variability and relative value depend on the distinguishability of each representation.

Show MeSH
Generalization and abstraction behavior of monkeys trained to associate dot arrays and numerals independently with the corresponding number of drops of juice. a Monkey choosing 12 shapes over 10. The stimuli are shapes of different colors (the colors were more distinctive than they appear on this video image). b Average performance of 4 monkeys for the first 2 days on this task. c Equal-area control, total dot area was always constant. The monkey chooses 21 small dots instead of 3 large dots. d Average performance of 4 monkeys for the first 2 days of the equal-area control task. e Transitivity of dot array learning and numeral learning. The monkey chooses the numeral N (= 20 drops) over 10 dots (= 10 drops). f Average performance of 4 monkeys over the first 4 days on the numeral vs. dots choice task. Their average performance did not change over the 4 days
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC2921054&req=5

Fig3: Generalization and abstraction behavior of monkeys trained to associate dot arrays and numerals independently with the corresponding number of drops of juice. a Monkey choosing 12 shapes over 10. The stimuli are shapes of different colors (the colors were more distinctive than they appear on this video image). b Average performance of 4 monkeys for the first 2 days on this task. c Equal-area control, total dot area was always constant. The monkey chooses 21 small dots instead of 3 large dots. d Average performance of 4 monkeys for the first 2 days of the equal-area control task. e Transitivity of dot array learning and numeral learning. The monkey chooses the numeral N (= 20 drops) over 10 dots (= 10 drops). f Average performance of 4 monkeys over the first 4 days on the numeral vs. dots choice task. Their average performance did not change over the 4 days

Mentions: The monkeys were trained alternately on choosing between pairs of numerals and pairs of dot patterns. They started with 0 and 1 and progressed to the next higher number when their performance on the highest number reached >75% averaged over all possible combinations with other numbers, for both dots and numerals. The reward pulses were long (yielding several drops per pulse) when the maximum numbers were small, and decreased in length as the monkeys attained larger maximum numbers. They reached a stable level of performance for both dots and numerals from 0 to 21 after 4 months of training. Then they were tested on 0–21 for both dots and numerals, on alternate days, for 1 month to ensure that their performance was stable, and then tested for 2 months to obtain the data presented here (Fig. 1). Control experiments (Fig. 3) were run subsequently. Only data from the first 200 trials each day, excluding the first 10 trials, were used for analysis because after 200 trials the monkeys usually started playing around and working sporadically, and during the first few trials the experimenter was still in the room, which distracted them.Fig. 1


The benefit of symbols: monkeys show linear, human-like, accuracy when using symbols to represent scalar value.

Livingstone MS, Srihasam K, Morocz IA - Anim Cogn (2010)

Generalization and abstraction behavior of monkeys trained to associate dot arrays and numerals independently with the corresponding number of drops of juice. a Monkey choosing 12 shapes over 10. The stimuli are shapes of different colors (the colors were more distinctive than they appear on this video image). b Average performance of 4 monkeys for the first 2 days on this task. c Equal-area control, total dot area was always constant. The monkey chooses 21 small dots instead of 3 large dots. d Average performance of 4 monkeys for the first 2 days of the equal-area control task. e Transitivity of dot array learning and numeral learning. The monkey chooses the numeral N (= 20 drops) over 10 dots (= 10 drops). f Average performance of 4 monkeys over the first 4 days on the numeral vs. dots choice task. Their average performance did not change over the 4 days
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC2921054&req=5

Fig3: Generalization and abstraction behavior of monkeys trained to associate dot arrays and numerals independently with the corresponding number of drops of juice. a Monkey choosing 12 shapes over 10. The stimuli are shapes of different colors (the colors were more distinctive than they appear on this video image). b Average performance of 4 monkeys for the first 2 days on this task. c Equal-area control, total dot area was always constant. The monkey chooses 21 small dots instead of 3 large dots. d Average performance of 4 monkeys for the first 2 days of the equal-area control task. e Transitivity of dot array learning and numeral learning. The monkey chooses the numeral N (= 20 drops) over 10 dots (= 10 drops). f Average performance of 4 monkeys over the first 4 days on the numeral vs. dots choice task. Their average performance did not change over the 4 days
Mentions: The monkeys were trained alternately on choosing between pairs of numerals and pairs of dot patterns. They started with 0 and 1 and progressed to the next higher number when their performance on the highest number reached >75% averaged over all possible combinations with other numbers, for both dots and numerals. The reward pulses were long (yielding several drops per pulse) when the maximum numbers were small, and decreased in length as the monkeys attained larger maximum numbers. They reached a stable level of performance for both dots and numerals from 0 to 21 after 4 months of training. Then they were tested on 0–21 for both dots and numerals, on alternate days, for 1 month to ensure that their performance was stable, and then tested for 2 months to obtain the data presented here (Fig. 1). Control experiments (Fig. 3) were run subsequently. Only data from the first 200 trials each day, excluding the first 10 trials, were used for analysis because after 200 trials the monkeys usually started playing around and working sporadically, and during the first few trials the experimenter was still in the room, which distracted them.Fig. 1

Bottom Line: When humans and animals estimate numbers of items, their error rate is proportional to the number.To date, however, only humans show the capacity to represent large numbers symbolically, which endows them with increased precision, especially for large numbers, and with tools for manipulating numbers.This ability depends critically on our capacity to acquire and represent explicit symbols.

View Article: PubMed Central - PubMed

Affiliation: Department of Neurobiology, Harvard Medical School, Boston, MA 02115, USA. mlivingstone@hms.harvard.edu

ABSTRACT
When humans and animals estimate numbers of items, their error rate is proportional to the number. To date, however, only humans show the capacity to represent large numbers symbolically, which endows them with increased precision, especially for large numbers, and with tools for manipulating numbers. This ability depends critically on our capacity to acquire and represent explicit symbols. Here we show that when rhesus monkeys are trained to use an explicit symbol system, they too show more precise, and linear, scaling than they do using a one-to-one corresponding numerosity representation. We also found that when taught two different types of representations for reward amount, the monkeys systematically undervalued the less precise representation. The results indicate that monkeys, like humans, can learn alternative mechanisms for representing a single value scale and that performance variability and relative value depend on the distinguishability of each representation.

Show MeSH