Daniel Everett described in his book 
[Dont sleep there are snakes][1] the language of an Amazonian tribe (Piraha) which has no words for numbers. This tribe also uses "immediacy of experience" for believing other people. The author also argues that this language is a counter-example to the idea of a language instinct as proposed by Whorf and Chomsky.

Caleb Everett in [Numbers and the Making of Us: Counting and the Course of Human Cultures][2] argues, based on the language of the Piraha, that counting is a cultural achievement and not innately laid down in the structure of the brain. 
On page 274, in "Notes" to the section "6. Quantities in the Minds of Young Children", Caleb E. mentions Piaget as follows:

- 10. Jacques Mehler and  Thomas Bever, “Cognitive Capacity of Very Young Children,” Science 3797 (1967): 141–142. See also the enlightening discussion on this topic in Dehaene,  The Number Sense: How the Mind Creates Mathematics, particularly as it relates to the work of Piaget. I should mention, however, that an insightful reviewer notes that there have been issues replicating the results of Mehler and Bever with very young children.
  

A review of Caleb Everett's book is in the [New York Times][3]. 

# Addition:

In the second part of his book, Caleb E. summarizes empirical evidence that people with anumerate languages and human babies have an innate sense to recognize quantities up to 3 precisely, and another innate sense to recognize the larger quantity between two aproximately (like: 16 is more than 8). To connect the two and to start to count needs exposure to language with number words and is a slow process. The approximate number sense is also there in rats, etc. 

On page 168 he writes:

- As Elizabeth Brannon and Joonkoo Park, two animal cognition specialists, recently suggested: “it is challenging to understand how such a primitive system that is not capable of representing exact large numbers could give rise to the formal mathematics that is uniquely human.”20  The quantitative thought with which we and other species are innately equipped is orders of magnitude removed from the kinds of quantitative thought that most humans eventually possess.  This suggests that biological explanations of such thought are inherently limited. Most of our numerical cognition owes relatively little to our neurobiological equipment and owes much more to the ways we manipulate that equipment.  This manipulation can only be possible if there are external tools interacting with our innate mechanisms for quantity differentiation.  The principal external tools in question are numbers, symbolic representations of quantities that are linguistically reified and used in culturally variant ways.  The existence of numbers explains the gulf between actual human numerical thought and the numerical thought to which we are innately predisposed.

  [1]: https://www.amazon.com/Dont-Sleep-There-Are-Snakes/dp/0307386120/ref=sr_1_1
  [2]: https://www.amazon.com/Numbers-Making-Us-Counting-Cultures/dp/0674504437/ref=sr_1_13
  [3]: https://www.nytimes.com/2017/05/04/books/review/Humanity-of-Numbers.html