Maintaining Deception

In my final year of undergraduate studies I conducted an independent research project under the guidence of Dr.Sarah Hodge, investigating the evolutionary theories of deception. For a long time scientists have puzzled over how deception is maintained in communication systems. This is because the benefits of using deception are expected to lead to its increase; everyone would end up lying, no-one would be listening and the communication system would break downl. However, this is not the case, so how is deception maintained in communication systems?


There are two main hypotheses that suggest signals should generally be honest; (a) organisms are physically constrained to honesty, with a key example from carotenoid derived colouration [1] and (b) the ‘handicap principle’ where honesty is maintained through costly signals [2]. However, Johnstone and Grafen [3] used game theory to argue that deception can be evolutionary stable if on average signals are honest. According to their work there are three key factors that will influence the success of deceptive signals:

i.The frequency of deceptive signals

Higher frequencies of deceptive signals are expected to reduce the success of deceptive signals.

ii.The cost of not reacting to an honest signal

Higher costs are expected to elicit greater response rates to deceptive signals.

iii.The cost of being deceived

Where lower costs should be coupled with higher responses to deceptive signals.


One example of deception is the kleptoparasitism (theft of food) by fork-tailed drongos (Dicrurus adsimilis) on foraging pied babblers (Turdoides bicolor) and even meerkats (Suricata suricatta) - see the video below. The drongos provide a service to the babblers, acting as sentries and alarm calling when they spot a predator. However, the drongos are found to occasionally make deceptive calls, followed by the theft of food items that the babblers drop [4].

This behaviour is also documented in the greater racket-tailed drongo (Dicrurus paradiseus) which, in mixed species flocks uses alarm calls to startle other birds and gain foraging opportunities [5].

The study on the racket-tailed drongo also found that their alarm calls differed to that of their deceptive calls. This provides receivers with the opportunity to differentiate between calls, raising the question;

Do Johnstone and Grafen's predictions still apply when receivers can learn to distinguish between calls?

Learning to differentiate between calls should prevent individuals becoming victim to deceptive attacks, regardless of the frequency and costs to deception. However, the unpredictable behaviour of signallers generates uncertainty. This causes errors in the receiver’s response to signals [6] meaning deception is maintained at low levels.


Inspired by the phenomenon of kleptoparasitism, I designed and implemented an adaptable game to test Johstone and Graften's predictions when learning is possible. To test the three predictions I conducted three different experiments manipulating either the frequency of deceptive signals, costs to failing to respond to honest signals or costs to being deceived.

I asked my volunteers to play two games where they used chopsticks to pick up mini eggs (yes the chocolate sweets) which were buried in sawdust. While foraging for the treats they were required to react to both honest and dishonest alarm signals which were slightly different. A correct response to an honest signal required the forager to lift their hands up and away from the foraging area before the signal ended. The costs to not responding to an honest signal and being deceived involved a deduction of points from the final foraging score (number of items collected).

Before playing the first game I provided the foragers with an opportunity to learn the two calls, letting them hear each call twice and telling them which was the honest signal. Each game lasted 2 minutes and each forager played two games with the treatments dependent on the experiment. The order in which the treatments were played was randomised. I monitored the response of foragers to signals and their foraging score (number of items collected).


What I found was interesting, none of the predictions held true! There was no difference in the number of responses to deceptive signals in any of the experiments. Additionally, successful deception was maintained at a low level, as expected if foragers were prone to error generated from uncertainty.

These results suggest that when individuals have the opportunity to learn and differentiate between honest and dishonest signals, deception can be maintained irrelevant of the costs and frequency of deception. It is likely that the frequency and costs of decpetion influence the amount of error receivers are prone to.

However,no effect between treatments may have been observed because the high frequency of deception or high cost treatments were not high enough. It would be great to pursue this work and adapt the game to more easily generate larger sample sizes.


I think it would be great to convert the experiments into online citizen science games! For example, rather than using chopsticks foragers could use the computer mouse to search for grubs by interacting with the scene. The game could also involve honest and dishonest signals (and these could be either audible or visual). The game could provide benefits in the form of points and costs in the form of a ‘time out’ or point deductions.

Citizen science games are great because they generate large amounts of data in a short time. They also provide a great tool for public engagement, allowing the public to be a part of the science while learning about the questions at hand! Unfortunately my skills in online computer coding are not quite good enough to produce such a game, but the interest is there to one day learn how.


  1. Johnstone, R. A. (1998). Game theory and communication In Game Theory and Animal Behavior (eds. L. A. Dugatkin and H. K. Reeve), pp. 94–117. Oxford, UK: Oxford University Press.


  1. Zahavi, A. (1975). Mate selection: selection for a handicap. Journal of Theoretical Biology 53: 205–214.


  1. Johnstone, R. A. and Grafen, A. (1993). Dishonesty and the handicap principle. Animal Behaviour 46: 759–764.


  1. Ridley, A.R., Child, M. F. and Bell, M. B. V. (2007). Interspecific audience effects on the alarm-calling behaviour of a kleptoparasitic bird. Biology Letters 3: 589-591


  1. Satischandra, S. H. K., Kodituwakku, P., Kotagama, S. W. and Goodale, E. (2010). Assessing “false” alarm calls by a Drongo (Dicrurus paradiseus) in mixed-species bird flocks. Behavioural Ecology 21: 396-403


  1. Wiley, R. H. (1994). Errors, exaggeration, and deception in animal communication. In Behavioral Mechanisms in Evolutionary Biology (ed. L. A. Real), pp. 157–189. Chicago, IL: Chicago University Press

Copyright Jared Wilson-Aggarwal - JWA 2015