Is optimism optimal? Functional causes of apparent behavioural biases

Alasdair I. Houston*, Pete C. Trimmer, Tim W. Fawcett, Andrew D. Higginson, James A R Marshall, John M. McNamara

*Corresponding author for this work

Research output: Contribution to journalArticle (Academic Journal)peer-review

15 Citations (Scopus)


We review the use of the terms 'optimism' and 'pessimism' to characterize particular types of behaviour in non-human animals. Animals can certainly behave as though they are optimistic or pessimistic with respect to specific motivations, as documented by an extensive range of examples in the literature. However, in surveying such examples we find that these terms are often poorly defined and are liable to lead to confusion. Furthermore, when considering behaviour within the framework of optimal decision theory using appropriate currencies, it is often misleading to describe animals as optimistic or pessimistic. There are two common misunderstandings. First, some apparent cases of biased behaviour result from misidentifying the currencies and pay-offs the animals should be maximising. Second, actions that do not maximise short-term pay-offs have sometimes been described as optimistic or pessimistic when in fact they are optimal in the long term; we show how such situations can be understood from the perspective of bandit models. Rather than describing suboptimal, unrealistic behaviour, the terms optimism and pessimism are better restricted to informal usage. Our review highlights the importance of choosing the relevant currency when attempting to predict the action of natural selection.

Original languageEnglish
Pages (from-to)172-178
Number of pages7
JournalBehavioural Processes
Issue number2
Publication statusPublished - 1 Feb 2012


  • Bandit models
  • Long-term currency
  • Natural selection
  • Optimal decision making
  • Optimism
  • Pessimism
  • Short-term currency


Dive into the research topics of 'Is optimism optimal? Functional causes of apparent behavioural biases'. Together they form a unique fingerprint.

Cite this