Specifically, Gelman and Gallistel reviewed experimental
studies with innumerate Pirah and Munduruku Indian subjects who either did not have true number words
or had consistent, unambiguous words only for one and two but more loosely
used words for three and four