March 17, 2013
by Artem Kaznatcheev
This post is a continuation of Part 1 from last week that introduced and motivated the economic Turing test.

Joseph Henrich
When discussing culture, the first person that springs to mind is
Joseph Henrich. He is the Canada Research Chair in Culture, Cognition and Coevolution, and Professor at the Departments of Psychology and Economics at the University of British Columbia. My most salient association with him is the
cultural brain hypothesis (CBH) that suggests that the human brain developed its size and complexity in order to better transmit cultural information. This idea seems like a nice continuation of Dunbar’s (1998) Social Brain hypothesis (SBH; see Dunbar & Shultz (2007) for a recent review or
this EvoAnth blog post for an overview), although I am still unaware of strong
evidence for the importance of gene-culture co-evolution — a requisite for CBH. Both hypotheses are also essential to studying intelligence; in animals intelligence is usually associated with (properly normalized) brain size and complexity, and social and cultural structure is usually associated with higher intellect.
To most evolutionary game theorists, Henrich is know not for how culture shapes brain development but how behavior in games and concepts of fairness vary across cultures. Henrich et al. (2001) studied the behavior of people from 15 small-scale societies in the prototypical test of fairness: the ultimatum game. They showed a great variability in how fairness is conceived and what operationalist results the conceptions produce across the societies they studied.
In general, the ‘universals’ that researchers learnt from studying western university students were not very universal. The groups studied fell into four categories:
- Three foraging societies,
- Six practicing slash-and-burn horticulture,
- Four nomadic herding groups, and
- Three small-scale farming societies.
These add up to sixteen, since the Sangu of Tanzania were split into farmers and herders. In fact, in the full analysis presented in table 1, the authors consider a total of 18 groups; splitting the Hadza of Tanzania into big and small camp, and the villagers of Zimbabwe into unsettled and resettled. Henrich et al. (2001) conclude that neither homoeconomicus nor the western university student (WEIRD; see Henrich, Heine, & Norenzaya (2010) for a definition and discussion) models accurately describe any of these groups. I am not sure why I should trust this result given a complete lack of statistical analysis, small sample size, and what seems like arithmetic mistakes in the table (for instance the resettled villagers rejected 12 out of 86 offers, but the authors list the rate as 7%). However, even without a detailed statistical analysis it is clear that there is a large variance across societies, and at least some of the societies don’t match economically rational behavior or the behavior of WEIRD participants.
The ultimatum game is an interaction between two participants, one is randomly assigned to be Alice and the other is Bob. Alice is given a couple of days wage in money (either the local currency or other common units of exchange like tobacco) and can decide what proportion of it to offer to Bob. She can choose to offer as little or as much as she wants. Bob is then told what proportion Alice offered and can decide to accept or reject. If Bob accepts then the game ends and each party receives their fraction of the goods. If Bob declines then both Alice and Bob receive nothing and the game terminates. The interaction is completely anonymous and happens only once to avoid effects of reputation or direct reciprocity. In this setting, homoeconomicus would give the lowest possible offer if playing as Alice and accept any non-zero offer as Bob (any money is better than no money).
The groups that most closely match the economists’ model are the Machiguenga of Peru, Quichua of Ecuador, and small camp Hadza. They provide the lowest average offers of 26%-27%. They reject offers 5%, 15%, and 28% of the time, respectively. Only the Tsimane of Bolivia (70 interactions), Achuar of Ecuador (16 interactions), and Ache of Paraguay (51 interactions) have zero offer rejection rates. However, members of all three societies offer a sizeable initial offer, averaging 37%, 42%, and 51%, respectively. A particularly surprising group is the Lamelara of Indonesia that offered on average 58% of their goods, and still rejected 3 out of 8 offers (they also generated 4 out of 20 experimenter generated low offers, since no low offers were given by group members). This behavior is drastically different from rational, and not very close to WEIRD participants that tend to offer around 50% and reject offers below 20% about 40% to 60% of the time. If we are to narrow our lens of human behavior to that of weird participants or economic theorizing than it is easy for us to miss the big picture of the drastic variability of behavior across human cultures.

It’s easy to see what we want instead of the truth when we focus too narrowly.
What does this mean for the economic Turing test? We cannot assume that the judge is able to decide how distinguish man from machine without also mistaking people of different cultures for machines. Without very careful selection of games, a judge can only distinguish members of its own culture from members of others. Thus, it is not a test of rationality but of conformation to social norms. I expect this flaw to extend to the traditional Turing test as well. Even if we eliminate the obvious cultural barrier of language by introducing a universal translator, I suspect that there will still be cultural norms that might force the judge to classify members of other cultures as machines. The operationalization of the Turing test has to be carefully studied with how it interacts with different cultures. More importantly, we need to question if a universal definition of intelligence is possible, or if it is inherently dependent on the culture that defines it.
What does this mean for evolutionary game theory? As an evolutionary game theorist, I often take an engineering perspective: pick a departure from objective rationality observed by the psychologists and design a simple model that reproduces this effect. The dependence of game behavior on culture means that I need to introduce a “culture knob” (either as a free or structural parameter) that can be used to tune my model to capture the variance in behavior observed across cultures. This also means that modelers must remain agnostic to the method of inheritance to allow for both genetic and cultural transmission (see Lansing & Cox (2011) for further considerations on how to use EGT when studying culture). Any conclusions or arguments for biological plausibility made from simulations must be examined carefully and compared to existing cross-cultural data. For example, it doesn’t make sense to conclude that fairness is a biologically evolved universal (Nowak, Page, & Sigmund, 2000) if we see such great variance in the concepts of fairness across different cultures of genetically similar humans.
References
Dunbar, R.I.M. (1998) The social brain hypothesis. Evolutionary Anthropology 6(5): 179-190. [pdf]
Dunbar, R.I.M., & Shultz, S. (2007) Evolution in the Social Brain. Science 317. [pdf]
Henrich, J., Boyd, R., Bowles, S., Camerer, C., Fehr, E., Gintis, H., & McElreath, R. (2001). In Search of Homo Economicus: Behavioral Experiments in 15 Small-Scale Societies American Economic Review, 91 (2), 73-78 DOI: 10.1257/aer.91.2.73
Henrich, J., Heine, S. J., & Norenzayan, A. (2010). The weirdest people in the world. Behavioral and Brain Sciences, 33(2-3), 61-83.
Lansing, J. S., & Cox, M. P. (2011). The Domain of the Replicators. Current Anthropology, 52(1), 105-125.
Nowak, M. A., Page, K. M., & Sigmund, K. (2000). Fairness versus reason in the ultimatum game. Science, 289(5485), 1773-1775.
claimtoken-514691b910cca
Interdisciplinitis: Do entropic forces cause adaptive behavior?
April 21, 2013 by Artem Kaznatcheev 27 Comments
Reinventing the square wheel. Art by Mark Fiore of San Francisco Chronicle.
Ten years later, Elias (1958) drained the pus with surgically precise rhetoric:
I highly recommend reading the whole editorial, it is only one page long and a delight of scientific sarcasm. Unfortunately — as any medical professional will tell you — draining the abscess is treating the symptoms, and without a regime of antibiotics, it is difficult to resolve the underlying cause of interdisciplinitis. Occasionally the symptoms flare up, with the most recent being two days ago in the prestigious Physics Review Letters.
Wissner-Gross & Freer (2013) try to push the relationship between intelligence and entropy maximization by suggesting that the human cognitive niche is explained by causal entropic forces. Entropic force is an apparent macroscopic force that depends on how you define the correspondence between microscopic and macroscopic states. Suppose that you have an ergodic system, in other words: every microscopic state is equally likely (or you have a well-behaved distribution over them) and the system transitions between microscopic states at random such that its long term behavior mimics the state distribution (i.e. the ensemble average and time-average distributions are the same). If you define a macroscopic variable, such that some value of the variable corresponds to more microscopic states than other values then when you talk about the system at the macroscopic level, it will seem like a force is pushing the system towards the macroscopic states with larger microscopic support. This force is called entropic because it is proportional to the entropy gradient.
Instead of defining their microstates as configurations of their system, the authors focus on possible paths the system can follow for time
into the future. The macroscopic states are then the initial configurations of those paths. They calculate the force corresponding to this micro-macro split and use it as a real force acting on the macrosystem. The result is a dynamics that tends towards configurations where the system has the most freedom for future paths; the physics way of saying that “intelligence is keeping your options open”.
In most cases to directly invoke the entropic force as a real force would be unreasonably, but the authors use a cognitive justification. Suppose that the agent uses a Monte Carlo simulation of paths out to a time horizon %latex \tau$ and then moves in accordance to the expected results of its’ simulation then the agents motion would be guided by the entropic force. The authors study the behavior of such an agent in four models: particle in a box, inverted pendulum, a tool use puzzle, and a “social cooperation” puzzle. Unfortunately, these tasks are enough to both falsify the authors’ theory and show that they do not understand the sort of questions behavioral scientists are asking.
If you are locked in a small empty (maybe padded, after reading this blog too much) room for an extended amount of time, where would you chose to sit? I would suspect most people would sit in the corner or near one of the walls, where they can rest. That is where I would sit. However, if adaptive behavior is meant to follow Wissner-Gross & Freer (2013) then, as the particle in their first model, you would be expected to remain in the middle of the room. More generally, you could modify any of the authors’ tasks by having the experimenter remove two random objects from the agents’ environment whenever they complete the task of securing a goal object. If these objects are manipulable by the agents, then the authors would predict that the agents would not complete their task, regardless of what the objects are since there are more future paths with the option to manipulate two objects instead of one. Of course, in a real setting, it would depend on what these objects are (food versus neutral) on if the agents would prefer them. None of this is built into the theory, so it is hard to take this as the claimed general theory of adaptive behavior. Of course, it could be that the authors leave “the filling in of the outline to the psychologists”.
Do their experiments address any questions psychologists are actually interested in? This is most clearly interested with their social cooperation task, which is meant to be an idealization of the following task we can see bonobos accomplishing (first minute of the video):
Yay, bonobos! Is the salient feature of this task that the apes figure out how to get the reward? No, it is actually that bonobos will cooperate in getting the reward regardless of it is in the central bin (to be shared between them) or into side bins (for each to grab their own). However, chimpanzees would work together only if the food is in separate bins and not if it is available in the central bin to be split. In the Wissner-Gross & Freer (2013) approach, both conditions would result in the same behavior. The authors are throwing away the relevant details of the model, and keeping the ones that psychologists don’t care about.
The paper seems to be an obtuse way of saying that “agents prefer to maximize their future possibilities”. This is definitely true in some cases, but false in others. However, it is not news to psychologists. Further, the authors abstraction misses the features psychologists care about while stressed irrelevant ones. It is a prime example of interdisciplinitis, and raises the main question: how can we avoid making the same mistake?
Since I am a computer scientists (and to some extent, physicist) working on interdisciplinary questions, this is particularly important for me. How can I be a good connector of disciplines? The first step seems to publish in journal relevant to the domain of the questions being asked, instead of the domain from which the tools being used originate. Although mathematical tools tends to be more developed in physics than biology or psychology, the ones used in Wissner-Gross & Freer (2013) are not beyond what you would see in the Journal of Mathematical Psychology. Mathematical psychologists tend to be well versed in the basics of information theory, since it tends to be important for understanding Bayesian inference and machine learning. As such, entropic forces can be easily presented to them in much the same way as I presented in this post.
By publishing in a journal specific to the field you are trying to make an impact on, you get feedback on if you are addressing the right questions for your target field instead of simply if others’ in your field (i.e. other physicists) think you are addressing the right questions. If your results get accepted then you also have more impact since they appear in a journal that your target audience reads, instead of one your field focuses on. Lastly, it is a show of respect for the existing work done in your target field. Since the goal is to set up a fruitful collaboration between disciplines, it is important to avoid E.O. Wilson’s mistake of treating researchers in other fields as expendable or irrelevant.
References
Elias, P. (1958). Two famous papers. IRE Transactions on Information Theory, 4(3): 99.
Shannon, Claude E. (1948). A Mathematical Theory of Communication. Bell System Technical Journal. 27(3): 379–423.
Wissner-Gross, A.D., & Freer, C.E. (2013). Causal Entropic Forces Phys. Rev. Lett., 110 (16) : 10.1103/PhysRevLett.110.168702
Filed under Commentary, Models, Reviews Tagged with current events, information theory, intelligence, video