The origins of number sense: a commentary on “Is there an innate sense of number in the brain?”

Abstract

The question of whether a “sense of number” is innate has been posed in a new article by Lorenzi et al. (2025). The article explores the behavioral and neurobiological evidence from newborn animals to delve into the evolutionary origins of a sense of number. Lorenzi et al.(2025) raises new questions, interpretations, and ideas for future work to understand how number sense has evolved in humans and nonhuman animals. In this commentary, I discuss the arguments for an innate number sense, evaluate the implications for numerical cognition, and suggest how future work could fill the current gaps in our knowledge.

A sense of number

Is a “sense of number” innate? That is the question posed in a new article by Lorenzi et al. (2025). To solve this question, the authors reason that both the behavior and neurobiological processes need to be examined. The authors present combined behavioral and neurobiological evidence of an innate “sense of number” with a focus on newborn animals.

Simultaneously collecting behavioral and neurobiological data to answer the question of whether number sense is innate is not an easy feat. As Lorenzi et al. (2025) emphasize, invasive studies of newborns are impossible for many species and the visual systems of many altricial species are not developed enough at birth to process visual stimuli. Therefore, Lorenzi et al. (2025) have relied on case studies from newborn animals with a specific focus on zebrafish and chicks. These two species allow behavioral and neurobiological data to be collected during spontaneous numerical tasks—tests of animals lacking prior experience with numerical stimuli.

Lorenzi et al. (2025) raises some intriguing questions, interpretations, implications, and provides experimental frameworks for future work to build upon. In this short commentary, I discuss the implications of an innate sense of number, the missing links in our understanding of how number sense has evolved, and how future work could fill the gaps in our knowledge.

Implications of an innate sense of number

There is debate over whether number sense, and more advanced numerical cognition, have evolved in nonhuman animals (hereafter “animals”), and indeed, if animals are even capable of performing complex numerical tasks (Shaki and Fischer 2015; Rugani et al. 2015a, 2015b, 2016; Núñez 2017a; Nieder 2018, 2020, 2025; Howard et al. 2019b; Howard et al. 2020; Shaki and Fischer 2020; MaBouDi et al. 2021; Giurfa et al. 2022; Giurfa et al. 2023; Pitt et al. 2023). A sense of number, or approximate number system (ANS), is a nonsymbolic representational system of processing quantities above the subitizing range of approximately four to seven items (Núñez 2017b). Animals processing quantities above four items are considered to be using the ANS mechanism of numerosity judgment. The ANS is thought to have an ancient evolutionary foundation, shared between humans and nonhuman animals (Feigenson et al. 2004; Brannon and Merritt 2011; DeWind et al. 2015). Numerical cognition, on the other hand, requires the exact and sometimes symbolic use of number including arithmetic operations and symbolic representation of numerosities (e.g. using Arabic numerals to represent numerosities) (Núñez 2017a).

Different perspectives claim a “sense of number” has evolved in animals and is an innate mechanism; however, other views posit it is the result of empiricism (sensory stimulation) or learning. Current evidence presented by Lorenzi et al. (2025), among others (Dehaene 2001; Nieder 2005, 2016a, 2017, 2020, 2021, 2025; Lorenzi et al. 2024 but see: Leibovich et al. 2017; Núñez 2017a, 2017b), suggests that animals do possess a sense of number at birth, but this innate sense is further developed by sensory experience and experience with numerosities Lorenzi et al. (2025). The authors write

“…learning is not essential to estimate numerosity.”

It may not be essential, but learning does help shape our numerical and quantity processing abilities. Learning, and even the method of conditioning, clearly plays an important role in advancing how animals perform on numerical tasks (for examples see: (Agrillo et al. 2017; Howard et al. 2019a)). The impact of learning on numerical discrimination influences both behavioral performance and, importantly, neural mechanisms. For example, the proportion of number neurons in the prefrontal cortex of rhesus monkeys (Macaca mulatta) increased from 14% when the monkeys were untrained to 20% following training on a numerical discrimination task. Performance on numerical discrimination also improved following training (Viswanathan and Nieder 2015). Thus, while learning and experience may not be necessary for a sense of number, learning certainly plays a role in performance and the neural mechanisms involved.

Could numerical cognition be innate?

If a sense of number is indeed innate and has evolved in humans and animals, we should next investigate whether numerical cognition also has an innate origin. There are debates as to whether numerical cognition—the exact, and sometimes symbolic, use of number—is only accessible to humans through the development of language and culture (Nieder 2017; Núñez 2017a, 2017b). Exploring whether animals can spontaneously perform tasks such as symbolic representation of numerosity (e.g. Arabic numerals), manipulation of symbolic numerical representations, and arithmetic (e.g. addition/subtraction) will enable us to determine if advanced numerical concepts, beyond a basic sense of number, have evolved in animals. Past examples of spontaneous arithmetic in some animals (Sulkowski and Hauser 2001; Flombaum et al. 2005) have paved the way for possible methods and controls for consideration.

Outstanding questions and future research

While Lorenzi et al. (2025) has provided insight into the evolutionary origins of a sense of number, many outstanding questions remain. Future work should expand upon the limited research into the neurobiology of number sense in invertebrates (Nieder 2016b; Bengochea et al. 2023). Such research would inform us of the evolution of number sense in species with highly different brain sizes and structures. This knowledge may allow us to further elucidate whether number sense evolved separately in vertebrates and invertebrates, or whether a common ancestor possessed a sense of number.

Lorenzi et al. (2025) argues that experiments testing visually naïve animals reveals whether animals have a true innate sense of number or if empiricism plays a significant role in number sense. Future work could test this concept in other species, as currently this seems to have only been shown in newborn chicks (Lorenzi et al. 2024). Such experiments would again inform the origins and evolution of number sense across species.

A sense of number is not restricted to modality. Animals that cannot be reared to be visually naïve (e.g. zebrafish) could be tested in other sensory domains that could be controlled such as auditory, tactile, or olfaction senses. Cross-examination experiments where animals are stimulated with visual numerosity stimuli, then tested in auditory, tactile, or olfactory domains for their familiarity with number sense in a different domain would be informative for understanding the roles of innateness, empiricism, and learning in number sense. Do animals have a stronger number sense when exposed to quantity independent of sensory stimulation type, or does the experience need to match the sense?

It seems that an innate sense of number has evolved in many species, possibly pointing to a common ancestor (Nieder 2017; Giurfa 2019; Giurfa et al. 2022). However, number sense is also shaped by experience with sensory input and numerosities. Lorenzi et al. (2025) have introduced intriguing concepts and ideas for future research into the origins of number sense and numerical cognition in animals.

Author contributions

Scarlett Howard: Conceptualization, Project administration, Writing—original draft, Writing—review & editing.

Funding

No funding to declare.

Conflict of interest statement: The author declares no conflict of interest.

References