A Structured Mereological Analysis of Intelligence and Individual Differences in Eurasian Jays (Garrulus glandarius) and Rooks (Corvus frugilegus): A Quantitative, Qualitative and Philosophical Facet Theory Research Program

July 15, 2016

The title of this project is detailed and specifies that the research methodology to be employed in this research will mainly be that of facet theory and specifically this will take quantitative, qualitative or philosophical orientations to a facet theoretical enquiry. Below, I present a brief outline of a burgeoning research project and I provide details of the background to the research. 


            Overall, the research asks the question,


            “what is the multi-dimensional[1] structure of intelligence and intelligent behaviour in two species of corvids?”


This question is developed and sub-divided such that the first of the major component of the project is concerned with the perception of corvid behaviour in a sample of scientists whose work involves them with ethological and laboratory based research into corvid intellectual and other behaviour related to individual differences between birds and species of corvids. The second major strand of the research views the actual intellectually related behaviour of two specific Eurasian species of corvids: Eurasian Jays (Garrulus glandarius) (BirdLife International, 2012) and Rooks (Corvus frugilegus) (Rook (bird), (2015). Many types of intellectually related behaviour have been studied in Eurasian Jays (ses for example: Legg and Clayton, 2014). Rooks have also been investigated in terms of their intellectual behaviours (see for example: Cheke, etal, 2011; Reid, 1982; Seed, etal, 2006, 2008; Zucca, etal, 2008) and Federspiel (2010) investigated several forms of intelligence as these are displayed in both of the above two species.


My research addresses corvid intelligence and individual differences in terms of these being ways in which specific corvids or species of corvids differ within species and between the two species groups. To enable a greater understanding of the structural qualities of avian cognitive behaviours proxy measures are needed for intelligence as it is not possible to give birds paper and pencil tests similar to those human beings may be given. That is to say, in order to access corvid intelligence and other differences in their psychological make-up, overt behaviours must be observed and intelligence and psychological difference inferred from what is seen. There are many obvious and other less apparent difficulties associated with measuring a variety of behaviours that are associated with intelligence in animals. One of these concerns is the apparent paucity of research conducted specifically into the conceptual understanding and delimiting of definitions of avian intelligence and avian intelligence related behaviour or the categorial structure of this construct. A further issue is the almost total lack of a clearly established literature on avian psychometrics or put more simply, how intelligence in birds can be measured in a conceptually coherent, rigorous, replicable, reliable and valid manner[2].


In stage one of the research, a non-probability based sample of research scientists from Cambridge University will receive, in a questionnaire survey, a series of questions about pre-determined corvid behaviours that are usually thought to demonstrate intellectual ability. These behaviours have appeared in the avian intelligence literature and have been cited as indicative of distinct strands of intelligence. Along with these behaviours I will also include behavioural types that have been found to account for intelligent behaviour in humans. Together, the avian and human behaviours that will be used to structure my enquiries are: perspective taking; co-operative problem solving; creating novel tools to solve specific problems; using working memory; fluid reasoning; visual comprehension; experience projection; visual spatial ability; processing speed (during the course of this research these categories are reduced in number to avoid overlapping content between human and animal behaviour types). The behaviours I have included are uncontroversial indicators of intelligence. I could have perhaps have chosen other behavioural exemplars but my objective at this initial stage of my research is to employ a series of varied indicators of intelligence rather than perform an exhaustive assessment of all forms of avian intelligence. During the procedure participants will be asked to complete a brief structured assessment of corvid intelligence in which respondents express their beliefs about the frequency with which specific birds successfully use each form of intellectually behaviour. 



[1] When I speak about the multi-dimensional structure or nature of corvid intelligence I am suggesting that no simple or singular conception of intelligence is adequate to describe this cognitive process. Throughout this research I will employ facet theory as a guiding structure within which I will both conceive of and analyse my research observations and other data. Facet theory was developed as a multi-dimensional theory generating procedure analysed through multi-dimensional scaling (MDS) procedures. At later stages in this research program I will use such non-metric statistical MDS analyses. In these procedures events or items such as the different forms of intellectual behaviour of corvids are located in theoretical space in such a way that items in the analysis are located closer to other items if these items are more similar in terms of a non-metric measure of association. A ranking procedure to some extent obviates difficulties in comparing different psychological constructs and the need for these constructs to be of a linear metric character.


[2] An illustration of the lack of literature in the area of the measurement of intelligence in birds is shown by the fact that in books on animal intelligence one page out of over 450 are given over to testing (Reznikova, 2007, p58). An early example of an approach to measuring animal intelligence (in this case this was learning) is shown by studies that employed the Wisconsin General Test Apparatus (WGTA) that was developed in the 1930s and later adapted (Huton, etal, 1959) and used with rats (Rolin, etal, 1963) and primates (Purnelle and Rumbaugh, 1965).

Please reload

  • LinkedIn Clean Grey

© 2015 Paul Hackett