Speak softly and carry a big stick.
—Theodore Roosevelt
We were thrilled to receive four extremely thoughtful and constructively critical commentaries on our focus article. Apart from anything else, we are pleased to see that all respondents recognized the importance of strategic signaling in both nature and international politics, and the value of exploring their differences and similarities.
We especially appreciated Rebecca Sear’s goal to ask big picture questions about how one would overcome the obstacles of getting evolutionary insights out of the labs of biologists and into the laps of policymakers. As with the growing field of evolutionary medicine, we agree that an evidence-based experimental approach will be the best way to identify and hone effective strategies—the proof of the utility of these ideas is in the pudding. Can they improve our predictions and success in international politics?
Armando Geller is concerned to point out that humans are a very complex species, but we disagree that this means there are no insights from successful strategies in other species. The key point is that 3.5 billion years of life on Earth reveals recurrent behavioral, organizational and mathematical patterns that appear across contexts and species, and transcend cognitive sophistication. For example, although human brains are remarkably complex, the movement of crowds can be modeled with surprisingly simple rules.
Humans are clever, but this hardly makes us immune to fundamental principles of nature. Almost every line drawn in the sand between humans and non-humans—tool use, language, deliberate planning for the future, sense of self—has dissolved with further examination of the animal world, or the realization that basic mechanisms of interaction trump cognition. We are different, but not as different as we’d like to believe. The advantage of analogies in natural history is in part through the sheer diversity of biology—ideas generalized from biology have been tested in all sorts of environments by all sorts of organisms. As with all generalized analogies, exceptions can be found and contingencies must be considered, but this same argument could be applied to generalized theories of international relationships that contain no evolutionary or biological guidance.
Geller is correct that human cognition adds a complicating dimension in applying signaling theory to human behavior, but cognitive capacity does not nullify the predictions of signaling models. Evolutionary models (e.g., ESS models, optimal foraging theory, life history theory) offer predictions about phenotypic strategies that often ignore the underlying mechanisms, including cognition, physiology, and even genetics, that produce these strategies, as long as these proximate mechanisms do not significantly constrain optimal outcomes. These models allow us to determine the selective pressures that favor successful strategies. While human cognition may enable humans to devise uniquely sophisticated signaling strategies, there is no reason to believe that human signals are somehow outside the reach of selection pressures.
From the opposite perspective, the critique that past theories of international relationships and politics have essentially come to similar conclusions as some of our evolutionarily inspired insights (as noted by Fearon) is not troubling to us—rather, it goes to support our claim that there are fundamental patterns that know no species boundaries. The tragedy is that different disciplines are reinventing the wheel. Biologists observe numerous examples of “convergent evolution” where very similar structures or behaviors arose through very different evolutionary pathways. Where our suppositions align with those resulting from separate intellectual thought processes would seem to strengthen their conclusions, not weaken them. Nonetheless, we would argue that a biologically-based theory has the distinct advantage of a large empirical database populated by the diversity of life, and thus knowledge of how such systems vary across a wide variety of contexts, and its use of signaling and communication to survive for billions of years in hostile and unpredictable environments.
Note also that, crucially, signaling mechanisms may arise from a variety of selection mechanisms. Humans are different from deer, and one such difference is that we can observe and copy successful strategies (of, say, signaling). However, this process of selection and replication is just another evolutionary process—cultural evolution instead of genetic evolution. The consequence? Equilibria are the same (all else being equal), but evolution is faster. Geller writes that “signaling systems … have emerged as outcomes of purposive interactions among human entities”. This is certainly the case, but (1) those outcomes are influenced by evolved psychology (which may include the products of genetic evolution of signaling mechanisms).
Focusing on the differences between human and non-human signaling systems, as Geller advises, is indeed an important part of our analysis. However, the beauty of evolutionary biology is it gives us a framework against which to compare when and what our one species does differently from the millions of others.
Olivier Morin offers a fascinating example of lying working in the interests of the state. However, there are two very important problems with using this example as a weapon against the importance of costly signaling.
First, we know from evolutionary game theory that cheaters can (and often will) exist at a low frequency—an example of frequency dependent selection. Indeed, models show that costly signaling, specifically, can remain perfectly stable in the presence of cheats who do not emit costly signals (Johnstone 1997; Searcy & Nowicki 2005). Thus, a Bismarck or two does nothing to upset the fact that costly signaling prevails among the majority.
Second, this one example (about Bismark), compelling as it is, worked precisely because other actors expected honesty. Indeed, diplomatic history is marked more by stringent etiquette than the violations that undermine them. More striking than cases like this are the widespread rules to which they are exceptions. Those widespread rules of the game are, among other things, honest signals: they persist even if they are sometimes breached. We agree with Morin that signalers may benefit by manipulating information. But this is only possible where honest signaling is the norm. There has to be an expectation for cheats to exploit it. As Teddy Roosevelt would advise us, if you carry a big stick, most of the time you do not have to shout.
It may also be worth noting that such cheats do not always succeed as Bismarck did. During the Vietnam war, Nixon came up with what he called the “mad man” strategy (Sagan & Suri 2003), in which he tried to convince Ho Chi Minh that he was crazy enough to consider the use of nuclear weapons against North Vietnam if negotiations failed. If Ho believed it, the war could be ended early. However, Nixon failed to convince Ho that this was the case. Deceptive signals do not always work.
Morin raised an excellent point when he noted that cost implies reliability and it works by indicating past, rather than future investment. While, as the saying goes, “past returns do not predict future performance”, this is valuable information in a Bayesian context and permits one to modify actions accordingly. We would argue, especially in our information-dense age that “merely” stating a position does indeed carry enormous immediate costs. Just consider how quickly simple, often symbolic gestures—President Obama apologizing to Afghanistan for U.S. troops burning Korans or Rush Limbaugh calling a birth-control advocate a “slut”—result in political and economic repercussions. In fact, in political observation, it is often seen as a sign of a strong leader that they are able to “speak their mind” and by contrast, one who is seen as continually parsing their words to “feel the political winds” is usually considered to be in a position of weakness. Building up a body of strong statements and yet maintaining a leadership position is certainly perceived as both a signal of past and current strength.
Finally, we contest the idea that “most diplomatic threats and promises are not costly signals in the biological sense [i.e., signals that are costly to produce]”. Indeed, it is also important to recognize that not all signals require significant costs to ensure reliability. For example, conventional signals, such as symbolic gestures, can be reliable, if dishonest signalers experience differential benefits. More promising for political signaling are models that demonstrate the importance of future interactions for maintaining low-cost reliable signals. In a study on rhesus macaques, for example, Silk et al. (2000) show that low-cost vocal signals, what the authors refer to as “cheap talk”, can be honest and stabilize even when interests conflict, as long as there is a high expectation of future interaction. These conditions would seem to parallel the reality of many political interactions; the likelihood of future interactions are particularly high, since governments and nations usually survive beyond the lives of their leaders.
Whether states follow through on their threats or commitments is laid bare for all to see in the pages of history, so there are reputational (and domestic political) costs to making claims that are unrealistic or unbelievable. But furthermore, threats and commitments very often require massive costs, in shoring up political capital to make them, binding alliance treaties, and military deployments. The United States only signaled that it was serious about invading Iraq in 1991 when there were half a million troops on the ground.
The importance of information in strategic signaling and negotiation was also recognized by James Fearon; a political scientist who has thought deeply about these ideas for a number years and has incorporated perspectives from other disciplines. His main point is that our predictions are not surprising because they have already been arrived at in political science and economics only means that there are fundamental patterns that transcend species and context. In the context of international relations, these same successful strategies may in fact have arisen precisely because they have been shaped by a process of (cultural) evolution. Our essay attempts to explain why and how those effective strategies were selected.
Fearon also notes that signals vary widely in international relations as well as in biology. Biological insights may be useful for this reason alone: the plethora of examples of signaling in nature allow us to generate predictions for what kind of signals are likely to emerge (and be successful) in a variety of different contexts. The key parameters are probably much more varied in nature than they are in politics.
Fearon might be more persuaded by the utility of evolution in its novel predictions for understanding of when and where signaling may succeed or fail. Although above we have stressed the system-level, that is, some selection mechanism generating costly signaling among interacting entities, costly signaling is thought to have been an important part of human cognitive evolution as well. If so, humans may be predisposed to making costly signals of commitment when interacting with other humans (including other state leaders). Thus, states may not need to wait for a process of socialization to generate signals: state leaders will do it anyway. This can be a useful insight, because the proximate mechanisms that cause costly signaling have not changed from the Pleistocene to today, but the social and technological context has changed beyond recognition. This means that an understanding of the evolution of signaling mechanisms can generate predictions for “evolutionary mismatch”—the contexts in which our proximate mechanisms for signaling will be triggered but lead to detrimental outcomes in the modern world.
Fearon may be right to note that “rules of thumb” are unlikely to lead to the optimal exploitation of signals in international politics. This is because rules of thumb work on average, over time, but make many individual errors along the way. Because humans are clever, they should in principle be able to work out what to do on a case-by-case basis rather than having a blanket strategy that works well on the whole but often makes mistakes (perhaps causing wars or disasters in the process). But this does not make the stable strategy less important to understand. In fact, it makes it more important to understand, because only by understanding the likely strategies of other states can we determine when to play by the rules or to seize the opportunity to bluff and win cheaply.
Lastly (he made so many important points!), we can defend his penultimate criticism of our suggestion to “deliberately make your gestures of reconciliation spontaneous in order to make them effective”. As he says, this may seem a contradiction—how can you plan something spontaneous? Again, evolution may be way ahead of us here. Robert Trivers’s new book suggests that humans have evolved not just deception, but self-deception, and this is an adaptive strategy to better deceive others (Trivers 2011). Deliberate deception can be ineffective because “behavioral leakage” can give the game away. Genuinely believing you are confident of victory, or willing to reconcile, even if you are not, is more likely to bluff an adversary. Long ago, economist Robert Frank made related arguments about emotions being a strategic device to deter enemies. Rational choice cannot match it because it is too predictable. Nature discovered this many millions of years ago.
We end with an analogy that illustrates the conservative nature of integrating disparate disciplines. The field of evolutionary medicine could be said to have begun with the 1990 Quarterly Review of Biology article showing how and why the field of medicine would benefit from the incorporation of evolutionary thinking and knowledge. Twenty-two years on, there are textbooks, edited volumes, and studies showing how evolutionary ideas can enhance medicine and public health, yet in the US there are no evolutionary medicine degrees, physicians are not trained in evolution, and most medical schools have no evolutionary biologists on their faculty (R. Nesse pers. comm.). Indeed, many of the recommendations are high-level suggestions for future study. While there have been a number of ‘actionable’ discoveries, evolutionary medicine sets the stage for proper empirical studies; all of which must be tested before application.
We believe that nature has similar insights for political science and that when questions can be focused to be more actionable, they can be studied and their successes and failures evaluated. Just like a doctor wouldn’t want to use evolutionary logic alone without a double-blind study to determine treatment strategies, diplomats and negotiators might not want to immediately jump in and adopt our suggestions. However, those that take lessons from life and study their effectiveness might find that 3.5 billion years of life has created effective time-tested strategies (Sagarin & Taylor 2008; Sagarin et al. 2010). It is a short step to investigate whether these work today as well.
References
Johnstone, R. A. 1997. The Evolution of Animal Signals. In Behavioural Ecology: An Evolutionary Approach (ed. J. R. Krebs & Davies). Oxford: Blackwell Science.
Sagan, S. D. & Suri, J. 2003. The madman nuclear alert: Secrecy, signaling, and safety in October 1969. International Security27, 150-183.
Sagarin, R. D., and T. Taylor (eds.) 2008. Natural Security: A Darwinian Approach to a Dangerous World. Berkeley: University of California Press.
Sagarin, R.D., Alcorta, C.S., Atran, S., Blumstein, D.T, Dietl, G.P, Hochberg, M.E., Johnson, D.D.P., Levin, S., Madin, E.P., Madin, J.S., Prescott, E.M., Sosis, R., Taylor, T., Toby, J. and G. Vermeij. 2010. OPINION: Decentralise, adapt and cooperate. Nature465, 292-293.
Searcy, W. A. & Nowicki, S. 2005.The Evolution of Animal Communication: Reliability and Deception in Signaling Systems. Princeton: Princeton University Press.
Silk, J.B., E. Kaldor, and R. Boyd 2000. Cheap talk when interests conflict. Animal Behaviour59, 423-432.
Trivers, R. L. 2011.Deceit and Self-Deception: Fooling Yourself the Better to Fool Others. London: Allen Lane.
Daniel T. Blumstein (1), Scott Atran (2), Scott Field (3), Michael E. Hochberg (4), Dominic D. P. Johnson (5), Raphael Sagarin (6), Richard Sosis (7), and Bradley Thayer (8)
1)Department of Ecology and Evolutionary Biology, University of California, 621 Young Drive South, Los Angeles, CA 90095-1606, USA. 2)CNRS, Institut Jean Nicod-Ecole Normale Supérieure, 29 rue d’Ulm, 75005 Paris, France. 3)National Security Affairs Department, Naval Postgraduate School, 1411 Cunningham Rd., Monterey, CA 93943, USA. 4)Institut des Sciences de l’Evolution, Université Montpellier II, CNRS, Place Eugéne Bataillon, 34095 Montpellier, France. 5)Politics and International Relations, University of Edinburgh, 15a George Square, Edinburgh, EH6 6LG, UK. 6)Institute of the Environment, University of Arizona, Tucson, Arizona 85716, USA. 7)Department of Anthropology, University of Connecticut, Storrs, Connecticut 06269-2176, USA. 8)Department of Political Science, Baylor University, One Bear Place #97276, Waco, Texas 76798, USA. In addition to their primary affiliations, the authors are members of the Natural Security Working Group (www.naturalsecurity.arizona.edu) that has been supported by the National Center for Ecological Analysis and Synthesis, and the Office of Naval Research-Global.
Address correspondence to: D.T. Blumstein, email: marmots@ucla.edu
Why We Experiment
July 7, 2022