Cultural evolution of systematically structured behaviour in a non-human primate

Nicolas Claidière, Kenny Smith, Simon Kirby, Joël FagotProceedings of the Royal Society B Biological Sciences 2014 · DOI

TL;DR

This study demonstrates that baboons, when placed in a controlled experimental setting, can exhibit key aspects of human cultural evolution, including progressive performance increases and the emergence of systematic, lineage-specific structures. This suggests that the foundations for cumulative culture may be more widespread in the primate lineage than previously thought.

What Problem Does This Paper Address?

Human culture is incredibly complex and has been a driving force behind our species' success. While it's clear that other animals also exhibit forms of culture (learned behaviors passed down socially), these animal cultures often lack the complexity and cumulative nature seen in humans. A key question is whether the unique complexity of human culture stems from uniquely human cognitive abilities or if it's a more general outcome of social transmission processes. This paper investigates whether non-human primates can exhibit the systematic structure and lineage specificity that are hallmarks of human cumulative culture, using a novel experimental paradigm.

Methods

The researchers adapted a "transmission chain" experimental paradigm, previously used with humans, for baboons. This involves a sequence of individuals where the output of one becomes the input for the next.

  1. Pattern Reproduction Task: Baboons were trained to interact with a touch screen to reproduce patterns of red squares on a grid.
  2. Iterated Learning Procedure:
    • Between-Individuals Transmission: A series of baboons (forming a "chain") were tasked with reproducing patterns. The actual patterns of squares touched by one baboon became the target patterns for the next baboon in the chain. This was repeated for multiple "generations" of transmission.
  3. Within-Individuals Transmission: As a control, a separate procedure exposed individual baboons to their own previous outputs to see if similar structures emerged without inter-individual transmission.
  4. Statistical Analysis: Generalized linear mixed models (GLMMs) were used to analyze performance (success rates) and the emergence of specific patterns (tetrominos) across generations and trial types (transmission vs. random).

Datasets

The study did not use pre-existing datasets. Instead, it generated its own data through the experimental protocol with 15 baboons. The "dataset" for each transmission trial consisted of the specific grid patterns reproduced by an individual baboon. The initial target patterns for the first baboon in each chain were randomly generated.

Key Insights and Ideas

  • Baboons can learn and transmit complex patterns: The experiment successfully trained baboons and demonstrated their ability to learn and reproduce specific visual patterns.
  • Cultural transmission leads to performance improvement: Over successive generations in the transmission chains, baboons showed a significant increase in their success rate, indicating a cumulative improvement in performance.
  • Systematic structures emerge spontaneously: The baboons' reproduced patterns evolved to favor "tetrominos" (configurations where all four red squares are connected), a rare but stable pattern. This suggests that systematic structure can arise even when the task's function (getting a reward) is arbitrary with respect to the pattern's form.
  • Structure is driven by system-wide regularities, not just individual bias: The emergence of tetrominos wasn't simply due to individual baboons preferring them. Performance on tetromino grids improved significantly within transmission trials over generations, suggesting a feedback loop where the presence of tetrominos in a set makes them easier to learn and reproduce.
  • Lineage specificity exists: Different transmission chains (lineages) developed distinct distributions of grid types, mirroring the diversity of human languages or bird song dialects. This highlights that cultural evolution isn't a single, convergent path.
  • Social transmission is key for structure: The within-individual transmission control showed that inter-individual transmission was crucial for the development of systematic structure, suggesting that social learning is not just about copying but about shaping and refining cultural elements over time.

Key Results

  • Performance Increase: Success rates in transmission trials increased significantly over generations (estimated 16.0% per generation), while performance on matched random trials slightly decreased.
  • Tetromino Emergence: The proportion of tetromino grids increased dramatically over generations in transmission trials (estimated 20.8% per generation) compared to random trials (estimated 2.3% per generation).
  • Tetromino Advantage: Performance on tetromino grids was significantly higher (estimated 112% higher odds of success) in transmission trials compared to non-tetromino grids, but lower in random trials.
  • Lineage Specificity: Statistical analysis showed significant differences in the distribution of grid types across the six independent transmission chains at the final generation, indicating lineage-specific cultural evolution.
  • Social Transmission Benefit: Between-individual transmission chains produced a higher proportion of tetrominos (30% higher odds) compared to within-individual chains, underscoring the importance of social learning.

Limitations and Future Work

  • Laboratory Setting: The experiment was conducted in a controlled laboratory environment. The authors acknowledge that this doesn't fully replicate the complexity of naturalistic cultural transmission and its ecological drivers.
  • Arbitrary Task: The task's "function" (getting a reward) was arbitrary with respect to the visual patterns. Future work could explore how functional constraints might influence the emergence of systematicity in other domains.
  • Origin of Profound Difference: While the study shows baboons can exhibit key aspects of human cultural evolution, it doesn't fully explain the profound difference in complexity between human and non-human cultures in the real world. The authors suggest that the arbitrary link between form and function in their task might be a key factor in generating systematicity, and that this might be less prevalent in many animal cultures where behaviors are tightly linked to specific functions.
  • Future Directions: The authors suggest that further research could explore the role of different cognitive biases, the impact of varying transmission fidelity, and how these experimental findings relate to the evolution of more complex human cultural phenomena like language.

Summary updated

Abstract

Culture pervades human life and is at the origin of the success of our species. A wide range of other animals have culture too, but often in a limited form that does not complexify through the gradual accumulation of innovations. We developed a new paradigm to study cultural evolution in primates in order to better evaluate our closest relatives' cultural capacities. Previous studies using transmission chain experimental paradigms, in which the behavioural output of one individual becomes the target behaviour for the next individual in the chain, show that cultural transmission can lead to the progressive emergence of systematically structured behaviours in humans. Inspired by this work, we combined a pattern reproduction task on touch screens with an iterated learning procedure to develop transmission chains of baboons (Papio papio). Using this procedure, we show that baboons can exhibit three fundamental aspects of human cultural evolution: a progressive increase in performance, the emergence of systematic structure and the presence of lineage specificity. Our results shed new light on human uniqueness: we share with our closest relatives essential capacities to produce human-like cultural evolution.
BibTeX
@inproceedings{claidiere2014cultural,
  author = {Nicolas Claidière and Kenny Smith and Simon Kirby and Joël Fagot},
  title = {{Cultural evolution of systematically structured behaviour in a non-human primate}},
  booktitle = {Proceedings of the Royal Society B Biological Sciences},
  year = {2014},
  doi = {10.1098/rspb.2014.1541},
}
Figure