This is a summary of Where the Sidewalk Ends: The Limits of Social Constructionism
Author: David Peterson, Sociology Department, Northwestern University
Click here to check the original version.
This paper pushes back on both strong social constructionism and strong nativism, using the concept of number as its test case. Peterson argues that neither side has it right on its own and proposes a more precise framework for sorting out where culture actually shapes cognition and where it doesn’t.
Social processes
The sociology of knowledge generally holds that meaning comes from social processes like tradition, ritual, and institution, not from objects themselves or from individual minds working in isolation. This idea traces back to Durkheim, who argued that basic categories of thought, things like time, space, number, and causality, come from society itself rather than from pure experience or from an innate rational structure.
Durkheim’s claim was radical: society is the soil from which even our most basic cognitive categories grow. Critics have long pointed out a circularity problem here. If society produces the categories of thought, how did society itself get organized enough to do that in the first place, since social life presumably needs those categories to function?
Sociologist Andrew Bergesen mounted a direct challenge to Durkheim using developmental psychology. He cited studies showing that infants, who haven’t been socialized yet, already display awareness of space, number, cause, and object categories. His conclusion was that these categories must be innate rather than socially produced.
Peterson agrees that Bergesen’s critique is valuable and empirically grounded, but argues his nativist conclusion goes too far. Peterson’s central move is distinguishing between numerical concepts, which are innate and tacit, and concepts of number, which are explicit and depend on cultural transmission.
Numerical concepts
The evidence for innate numerical concepts is substantial. Infants show two separate innate systems for handling quantity. The first, called parallel individuation, lets infants accurately track and compare small sets up to about three items; beyond three, the system breaks down completely, even in cases where the ratio should make discrimination easy.
The second, analog magnitude estimation, lets infants and even animals like rats make rough comparisons of larger quantities, but these estimates follow Weber’s law, meaning accuracy depends on the ratio between two amounts rather than their exact difference. Neither system amounts to a true concept of natural number, since natural number requires understanding that you can always add one to get a larger number, a rule neither innate system captures.
Peterson brings in child development research to show the transition from these innate systems to explicit number isn’t smooth or automatic. Children take six to nine months to move from correctly using “one” to correctly using “two,” despite already distinguishing these quantities as infants. Around age three and a half, children undergo a sudden shift from being “subset knowers,” who only handle sets of one to three, to “cardinal principle knowers,” who grasp that counting and set size are unified. This is a qualitative leap, not a gradual accumulation of memory or capacity.
Discontinuity
Cross-cultural evidence reinforces this discontinuity. The Pirahã, an Amazonian tribe, use a number system limited to roughly “one,” “two,” and “many.” When asked to match a set of visible items with an equivalent set, they perform well, but once the original set is hidden from view, their errors match the imprecise pattern of the analog magnitude system rather than exact counting.
The Mundurukú, another Amazonian group with no number words past five, show the same pattern once tested beyond five items. Neither group lacks intelligence or contact with quantities; they simply never developed the cultural technology of natural number, and their cognition stays limited to the innate systems as a result.
From this, Peterson concludes that natural number requires a pre-existing cultural domain of knowledge to develop at all. Without it, a child or a whole culture may never move past the two innate but limited systems, no matter how much they interact with the physical world. This directly contradicts embodied-cognition theories that try to derive number purely from physical interaction with objects, since the Pirahã and Mundurukú interact with objects constantly without ever arriving at natural number.
A broader framework
Peterson then generalizes this into a broader framework describing four distinct ways culture can shape cognition, replacing the flat “natural versus socially constructed” dichotomy:
Social facilitation applies to domains like number and formal logic, where a culture either has the domain available to transmit or it doesn’t, and the underlying structure can’t be reshaped once it exists.
Social division applies to continuous domains with no natural boundaries, like pitch or color. Cultures impose arbitrary but internally consistent divisions, which explains why the number of basic color terms varies across languages while showing consistent patterns once a given number of terms exists.
Social specification applies to underspecified innate tendencies, like infants’ preference for faces or their inclination to attribute agency to certain kinds of objects. These starting points get shaped into culturally specific forms through both direct teaching and ordinary exposure.
Social construction is reserved for concepts with real historical variability and no innate grounding at all, like systems of government or musical styles. These differ from the other three categories because they aren’t built on top of any universal cognitive substrate.
Peterson’s conclusion
Peterson’s conclusion argues that sociology of knowledge needs to engage seriously with cognitive and developmental psychology rather than treating all mental categories as either fully innate or fully constructed. Both Durkheim and Bergesen capture part of the truth. Durkheim was right that culture provides indispensable scaffolding for developing full number concepts. Bergesen was right that a real innate cognitive substrate exists. The mistake, in Peterson’s view, is collapsing these into a single all-or-nothing claim instead of examining how the two interact differently across different domains of thought.