Deceptive-like behaviour in dogs (Canis familiaris)

Abstract

Deception, the use of false signals to modify the behaviour of the receiver, occurs in low frequencies even in stable signalling systems. For example, it can be advantageous for subordinate individuals to deceive in competitive situations. We investigated in a three-way choice task whether dogs are able to mislead a human competitor, i.e. if they are capable of tactical deception. During training, dogs experienced the role of their owner, as always being cooperative, and two unfamiliar humans, one acting ‘cooperatively’ by giving food and the other being ‘competitive’ and keeping the food for themselves. During the test, the dog had the options to lead one of these partners to one of the three potential food locations: one contained a favoured food item, the other a non-preferred food item and the third remained empty. After having led one of the partners, the dog always had the possibility of leading its cooperative owner to one of the food locations. Therefore, a dog would have a direct benefit from misleading the competitive partner since it would then get another chance to receive the preferred food from the owner. On the first test day, the dogs led the cooperative partner to the preferred food box more often than expected by chance and more often than the competitive partner. On the second day, they even led the competitive partner less often to the preferred food than expected by chance and more often to the empty box than the cooperative partner. These results show that dogs distinguished between the cooperative and the competitive partner, and indicate the flexibility of dogs to adjust their behaviour and that they are able to use tactical deception.

Full text

ORIGINAL PAPER
Deceptive-like behaviour in dogs (Canis familiaris)
Marianne T. E. Heberlein
1
•Marta B. Manser
1
•Dennis C. Turner
1,2
Received: 2 January 2016 / Revised: 6 February 2017 / Accepted: 21 February 2017 / Published online: 1 March 2017
ÓSpringer-Verlag Berlin Heidelberg 2017
Abstract Deception, the use of false signals to modify the
behaviour of the receiver, occurs in low frequencies even
in stable signalling systems. For example, it can be
advantageous for subordinate individuals to deceive in
competitive situations. We investigated in a three-way
choice task whether dogs are able to mislead a human
competitor, i.e. if they are capable of tactical deception.
During training, dogs experienced the role of their owner,
as always being cooperative, and two unfamiliar humans,
one acting ‘cooperatively’ by giving food and the other
being ‘competitive’ and keeping the food for themselves.
During the test, the dog had the options to lead one of these
partners to one of the three potential food locations: one
contained a favoured food item, the other a non-preferred
food item and the third remained empty. After having led
one of the partners, the dog always had the possibility of
leading its cooperative owner to one of the food locations.
Therefore, a dog would have a direct benefit from mis-
leading the competitive partner since it would then get
another chance to receive the preferred food from the
owner. On the first test day, the dogs led the cooperative
partner to the preferred food box more often than expected
by chance and more often than the competitive partner. On
the second day, they even led the competitive partner less
often to the preferred food than expected by chance and
more often to the empty box than the cooperative partner.
These results show that dogs distinguished between the
cooperative and the competitive partner, and indicate the
flexibility of dogs to adjust their behaviour and that they
are able to use tactical deception.
Keywords Canis familiaris Cognition Deception 
Dog Misleading Dog–human interaction
Introduction
Behavioural deception involves the use of false signals to
modify the behaviour of a receiver resulting in a benefit for
the signaller and a cost for the receiver (Semple and
McComb 1996). The costs can vary from being very high,
involving the loss of life, or relatively minor, such as
energy expenditure to find a different foraging location.
Deception has been predicted to occur even in stable sig-
nalling systems, where a signal reliably causes a certain
reaction in the recipients (Johnstone and Grafen 1993). The
frequency of misleading incidences depends on the costs to
the receiver, but in general, honest signalling must prevail
(Johnstone and Grafen 1993). For example, male domestic
chickens (Gallus gallus domesticus) produce food calls that
convey information about food quality to a hen, thereby
guiding her to better food sites (Marler et al. 1986), but
males regularly produce the calls even when no food is
available in order to attract the female (Gyger and Marler
1988). Responding to the signal can cause some cost to the
hen, since she might leave a food patch in expectation of a
better one. Therefore, the benefit of her reacting to this call
must be on average higher than the costs or she will stop
responding. The rooster, hence, should emit this signal
‘honestly’ in most cases, for his deceptive call to continue
having the beneficial outcome.
&Marianne T. E. Heberlein
marianne.heberlein@ieu.uzh.ch
1
Animal Behaviour, Department of Evolutionary Biology and
Environmental Studies, University of Zurich,
Winterthurerstrasse 190, 8057 Zurich, Switzerland
2
I.E.T./I.E.A.P., Horgen, Switzerland
123
Anim Cogn (2017) 20:511–520
DOI 10.1007/s10071-017-1078-6
Content courtesy of Springer Nature, terms of use apply. Rights reserved.