Original ArticleLanguage and cooperation in hominin scavenging☆
Section snippets
Language and the origins of cooperation
Bickerton (2009) makes the bold claim that “without understanding how language evolved, we can never hope to explain or understand ourselves” (p. 12). Language, he argues, is fundamental to the story of human evolution. Not only is language the product of an evolved, human-like mind, it is a key ingredient in the recipe for one. Without the cooperative outcomes made possible through language, we would lack access to high calorie meat necessary for larger, more complex brains. Language is
Experimental design and procedures
Our project is unrepentantly descriptive. Rather than testing formal hypotheses for a set of treatment conditions specified in advance, we develop some facts with a series of successive treatment conditions that illuminate the process by which language solves a cooperative problem. Whereas many experiments are nomothetic, i.e., designed to test theoretical or empirical “propositions” about the world, ours is heuristic or exploratory in nature (Smith 1982). We are probing a new a line of inquiry
Free Chat results and the selection of words for the Bounded treatment
We evaluate the performance of a session by averaging, per period, the net health added from scavenging guarded meat. We calculate that metric by summing the total health added from (guarded) meat in a period and subtracting from it the total amount of health lost from engaging the tiger. We then average these results for the first 14 and last 13 periods in a session. As the figures make clear, it takes some time for the participants to get their bearings in this virtual world. Presenting the
The Bounded treatment
Fig. 5 (b) reports the average net health gains from scavenging guarded meat in the Bounded treatment. Notice how the six sessions, as a set, replicate the results from the Free Chat baseline: One session (B2) excels in cooperating against the tigers, three others yield positive results in the second half of the session (B3–5), and two sessions perform quite poorly (B1 and B6). In other words, this is a nontrivial problem in an uncertain world. Using a Kolmogorov–Smirnov test for a difference
The Bounded – {Recruitment} treatment
Given the prominence of recruitment in Bickerton's hypothesis, and of recruitment words in B2, for our next treatment condition we exclude the 6 recruitment words in the first column of Table 2. The complete list is presented in Table 3. This is the sole difference between B and this Bounded – {Recruitment} (B–R) treatment.
Fig. 5 (c) reports the results by session for this treatment. Notice again that one successful session (B–R4) stands apart from the rest, and that two other sessions find a
The Bounded – {Recruitment} – {Scavenging} treatment
When participants in B–R cannot use recruitment words in column 1 of Table 2, they fall back on the scavenging words in column 2. In our next treatment condition, we remove the scavenging words harvest, defend, and switch from those available. See Table 4 for the complete list of words and symbols available to the participants. This is the sole difference between the B–R treatment and what we will call Bounded – {Recruitment} – {Scavenging} (B–RS).
Fig. 5 (d) reports the results from the
The Bounded – {Recruitment} – {Scavenging} – {Meat} treatment
When participants in B–RS could not use recruitment words, they relied heavily on the word meat as both a verb and an object. In our next treatment condition, we remove that one word—meat—from the list of available words/symbols (see Table 5). This is the sole difference between the B–RS treatment and what we will call the Bounded – {Recruitment} – {Scavenging} – {Meat} (B–RSM) treatment.
Fig. 6 reports the results from this treatment by session. Notice for a fifth time that one session (B–RSM2)
The {Pointers} + {Dispositions} treatment
With the elimination of a single word, the B–RSM treatment very nearly eliminated cooperation. The words we, us, and people functioned to create a common identity for the purpose of pursuing a common, concrete end. For our final treatment condition, we test whether the lack of any words stymies cooperation altogether. In what we will call the {Pointers} + {Dispositions} (P + D) treatment, only the pointing and disposition symbols in the last two columns of Table 2 are available to the participants
Discussion and conclusion
The first take away from comparing our treatment conditions is that our scavenging problem is a tough problem to solve (see Fig. 5, Fig. 6, Fig. 7). It is important to recognize that the results did not have to turn out this way. The problem could have been so easy that every session solved it, or so difficult that it was never solved. Instead, we have a severe test of our participants' cooperative powers: Only one session in each treatment regularly solves the problem. In these successful
References (21)
- et al.
Economics and identity
Quarterly Journal of Economics
(2000) - et al.
The hominid entry into the cognitive niche
- et al.
Rapid cultural adaptation can facilitate the evolution of large-scale cooperation
Behavioral Ecology and Sociobiology
(2010) Adam's tongue: how humans made language, how language made humans
(2009)More than nature needs: language, mind, and evolution
(2014)- et al.
Confrontational scavenging as a possible source for language and cooperation
BMC Evolutionary Biology
(2011) Proto-discourse and the emergence of compositionality
Interaction Studies
(2008)The symbolic species: the Co-evolution of language and the brain
(1997)- et al.
Early hominid hunting and scavenging: A Zooarcheological review
Evolutionary Anthropology: Issues, News, and Reviews
(2003) Public goods: A survey of experimental research
Cited by (7)
The influence of language on the evolution of cooperation
2023, Evolution and Human BehaviorMutual choice of partner and communication in a repeated prisoner's dilemma
2018, Journal of Behavioral and Experimental EconomicsCitation Excerpt :In order to throw light on the mechanisms behind these results, we turn to how the chat is used to shape cooperation. In analyzing the chat transcripts, we follow a procedure similar to that of Wilson and Harris (2017).14 We divide the chat into three phases: (1) establishing a common end, (2) forming pairs to achieve a common end (this phase is only relevant for the conditions with partner choice), and (3) maintaining the cooperative spirit to achieve the common end established in phase 1.
Computer simulation of scavenging by hominins and giant hyenas in the late Early Pleistocene
2023, Scientific ReportsComparative economics: how studying other primates helps us better understand the evolution of our own economic decision making
2023, Philosophical Transactions of the Royal Society B: Biological SciencesEvolutionary ecology of language origins through confrontational scavenging
2023, Philosophical Transactions of the Royal Society B: Biological SciencesThe Effect of a Guided Imagery Suggestion-based Intervention on Human Cooperation in the Stag Hunt Game
2022, North American Journal of Psychology
- ☆
We gratefully acknowledge the financial support of Chapman University and the National Science Foundation (SES 1123803). Wilson specifically thanks the Property and Environment Research Center (PERC) for their generous hospitality and financial support as a Lone Mountain Fellow during the summer of 2015, and Harris thanks the International Foundation for Research in Experimental Economics and Thomas W. Smith Foundation for their generous support to work on this project as a Summer Scholar from 2012-15. We also thank Jeffrey Kirchner for his creative software programming; Bailey Ennis, Jeremiah Ludwinski, Keane Tarrosa, and Sean Weller for their research assistance on this project in its nascence; Joshua Ball, Benjamin Chalmers, Gillian Courtney, Alex Feldman, J. Trent Gerdeman, and Mary Howard for diligently fact checking our narratives; and Jan Osborn and Stephen Semos for their meticulous comments on an early draft of the paper. Finally, we thank Frank Wolak, Erez Yoeli, seminar participants at PERC, and conference participants at the Economic Science Association meetings in Dallas for valuable conversations and comments that have improved the paper. The data and source code are available upon request.