Forum: Featured Discussions: What can animal models tell us, if anything, about human wisdom?

Animal models are powerful and profitable tools for understanding both basic biological processes (such as transcription and translation) as well as much more complex ones, such as cancer origination and progression, organ development, and immune processes. This truth underlies the use of rodents and primates for trials of promising medical advances, including pharmaceuticals and various dietary and exercise regimens.  Similarly, species that are much more distantly related to humans have proven to be the most powerful genetic models for illuminating a wide range of biological processes and, equally importantly, for demonstrating how processes at the more simple end of the spectrum can be combined and tuned to produce the complex.  

A nice example of this is the wealth of research on the molecular genetics of learning and memory. Genetic and biochemical manipulations of the fruit fly (or more correctly, vinegar fly), Drosophila melanogaster, have shown that there are distinct, qualitatively different short-term, medium-term, and at least two types of long-term memories (1). These types of memory are not simply different versions of the same thing, but rather, different genetic pathways and enzymatic interactions are mobilized when data are stored as different types of memory. Blocking the formation of one type of memory with drugs does not prevent the data from being stored as a different type of memory.

But is it possible to use animal models such as these to study something as elusive and ill-defined as human wisdom? This depends, in large part, on how we define wisdom. At one level, it is clear that wisdom requires biological processes, such as the ability to learn and remember. As discussed above, it is clear that animal models have something to offer at this level.

But are there elements of wisdom that are not a biological processes, and therefore not amenable to study using model organisms? One interesting possibility is that wisdom may arise as an emergent property of society – interactions among individuals may fuel the growth of wisdom in a way that is impossible for an individual in solitude. This point was made quite nicely by John Cacioppo and Valerie Tiberius, in an October 2008 post on this site. Thus, in social groups generally, whether they be human or insect, wisdom may a property of individuals that is manifested most strongly when social connections permit extension to a group larger than ourselves.

This type of wisdom brings to mind the concept of “transactive memory” (developed, I believe, by Daniel Wegner (2), and discussed by Malcolm Gladwell in his 2002 book, The Tipping Point (3)). According to this view, individuals possess their own personal knowledge, but also take advantage of the data stored in the minds of the people in their close social network. Within the social group, individuals can then specialize on particular types of knowledge and develop different proficiencies, and need only remember which members of the group possess the knowledge outside their particular areas of expertise. To me, this is highly reminiscent of the division of labor among members of a social insect colony – foragers, nurses, soldiers, queens, etc.

For human individuals, this social dimension must certainly permit the development of individual wisdom. But I would argue that organisms with much less sophisticated neural circuitry can also leverage these social connections to make wise decisions – in some cases more effectively than do humans.

A nice example of this is illustrated in the recent, high-profile study by Susan Edwards and Stephen Pratt, on the ant, Temnothorax curvispinosus, published in Proceedings of the Royal Society of London-Series B (4). To understand their study, we must first know something about human decision-making. Although humans possess a magnificent capacity for reason and thought, we also harbor some cognitive blind spots (5). One of these is evident when consumers attempt to select between two equally attractive alternatives that vary in different characteristics – for example, car A is comfortable, but ugly whereas car B is uncomfortable, but attractive. Normally, each car would have a 50% chance of being selected by the consumer (if truly equally attractive). This is considered a “rational” decision, because the frequency that each car is chosen matches its value relative to the alternative. For our purposes, we can consider this the wise choice.

However, imagine that a third option is presented: a choice (that we’ll call car C) that is horribly uncomfortable, but is the most attractive, appealing-looking car you’ve ever seen. In fact, car C is so uncomfortable that you would probably never choose it over car A or B. Nevertheless, simply having car C presented as an option can skew the perception of the original two items. In this example, including car C as an option would lead to an overvaluation of attractiveness, and you’d be more likely to choose car B over car A. This is now an “irrational”, or unwise, decision. The relative value of A versus B has not changed, but the selection probabilities have. Interestingly, this type of irrational decision-making is not limited to humans, but has also been observed in animals such as honeybees and jays, as well (6).

Now back to the study of Edwards and Pratt. Temnothorax ants typically nest in small nooks and crevices, like inside acorns or hollow twigs, or in the narrow cracks between flakes of stone. Many of these refuges are ephemeral, and this has driven the evolution of a highly effective behavioral system for colony relocation. When forced to relocate, scouts investigate nearby sites and gauge their suitability for nesting. Temnothorax particularly like nooks with small entrances and dark interiors. By presenting nests with entrances of various sizes and interiors of various darkness, researchers have been able to identify combinations that are equally attractive, but also suboptimal in one of the characters (entrance too wide or interior too bright). Now what happens when presented with the ant equivalent of car C? Edwards and Pratt show that the ants don’t fall into the unwise (or irrational) choice - the first two options are still favored at equal frequency.

So how do ants avoid making the irrational choice? Are ants smarter than people? Actually, the ants make the appropriate decision because they know less. No individual ant surveys all the choices, but instead, each scout examines one prospective nesting site, then returns to the colony to recruit other ants to investigate. If these new ants are enthusiastic about the new site, they return to the old nest to recruit more investigators. Good nesting sites continue to attract more and more ants until a threshold number of ants are convinced that this should be the colony’s new home, and all the ants move to the new location en masse. Although poorer sites may initially attract some new recruits, the number of ants going to investigate gradually dwindles as the competing superior site gains momentum.

Thus, the wise choice is made by the ants because their decision-making is distributed across many small minds, rather than in a single large brain where all the information is gathered together. In these ant societies, extraneous, distracting choices do not taint the perception of other options.

I would be interested to hear from the computer scientists about this topic. There are clear parallels to distributed networking, and I know that Daniel Wegner has done at least a little work comparing computer networks to human transactional memory (7).

-Neil Tsutsui, Assistant Professor, University of California, Berkeley

1.    Skoulakis, E. M. C. & Grammenoudi, S. Dunces and da Vincis: The genetics of learning and memory in Drosophila. Cellular and Molecular Life Sciences 63, 975-988 (2006).
2.    Wegner, D. M., Erber, R. & Raymond, P. Transactive memory in close relationships. Journal of Personality and Social Psychology 61, 923-929 (1991).
3.    Gladwell, M. The Tipping Point: How Little Things Can Make a Big Difference (Back Bay Books, Boston, MA, 2002).
4.    Edwards, S. C. & Pratt, S. C. Rationality in collective decision-making by ant colonies. Proceedings of the Royal Society of London B - Biological Sciences (2009).
5.    Huber, J., Payne, J. W. & ***, C. Adding Asymmetrically Dominated Alternatives - Violations of Regularity and the Similarity Hypothesis. Journal of Consumer Research 9, 90-98 (1982).
6.    Shafir, S., Waite, T. A. & Smith, B. H. Context-dependent violations of rational choice in honeybees (Apis mellifera) and gray jays (Perisoreus canadensis). Behavioral Ecology and Sociobiology 51, 180-187 (2002).
7.    Wegner, D. M. A computer network model of human transactive memory. Social Cognition 13, 319-339 (1995).

Photo from Flickr Creative Commons.

Neil, thanks for the interesting post.  How would the analogy between the ants and humans work if we instead considered individual ants the analogue of, say, neurons, and the ant colony as an individual human (e.g., the way Douglas Hofstandter does in Ant Fugue)?  In that case, it looks like the colony behaves rationally because of the "group dynamics" of the individual ants, just as humans can behave rationally because of the group dynamics of their brains' neurons.

Of course, your example uses a case of humans apparently behaving irrationally.  But I'm not convinced that these studies demonstrate irrational behavior.  While the objective features of cars A and B have not changed with the introduction of car C, the subject's subjective preferences have been influenced by car C so that she now favors car B.  Why not just see this as her choice being based on her having gained relevant information about her preferences?  The only reason A and B were considered "objectively" equal to begin with is because they were considered "subjectively" equal in terms of the subject's preferences. 

We may, however, worry about the rationality of the decision-making if the subject has no idea why she is being influenced in the way she is (and even makes up reasons that have nothing to do with the causal influence of car C on her preferences)--indeed, I worry about that.  Of course, neither the ants nor the ant colony know anything about why they are making the choices they make, I presume.