Learning through exploration: a bird's eye view more

Learning through exploration: a bird's eye view zxd878@bham.ac.uk School of Biosciences, University of Birmingham, Edgbaston, Birmingham, B15 2TT, UK Zoe Demery & Jackie Chappell Introduction - Research has focussed on what different species' cognitive capacities are, rather than how animals structure information, especially by exploration . - We believe exploration is not random; but structured, selective and sensitive to particular categorical features of the environment[1,2]. - We have exploited the natural exploratory tendencies and unique sensorimotor apparatus of parrots to investigate the mechanisms of learning about object properties through vision and manipulation . Key do parrots' visual fields and bill tip organ support their exploratory learning? Questions 1 . How 2. What cues are parrots particularly sensitive to? 3. How does their exploration change with environmental changes? 4. What does this suggest about their internal folk physics? FIG.1 MethodsSenegal parrot (Poicephalus senegalus) visual fields by the ophthalmoscopic - We measured reflex technique and studied touch receptors in their bill to characterise the limits of parrots' cross-modal exploration [3]. - This led to two behavioural experiments to look at the exploration process in detail, specifically kakarikis' (Cyanoramphus novaezelandiae) sensitivity to functional invisible cues (weight, centre-of-mass) versus visible cues (colour, shape)[FIG.2]. - Using a counterbalanced, repeated measures design, we consecutively presented 1 0 parrots individually with rigid, novel objects of different shapes, colours, weight and centres-ofmass . all p < .01 (**) FIG. 2 EXPT. 1 ** FIG. 3 EXPT. 2 ** Conclusionsunlike those of any other birds, with both quite a - Parrot visual fields are broad frontal binocular field and nearly comprehensive view around the head [FIG.1]. - Perhaps the bill tip organ allows the visual field to shift up & back for better predator detection, so once grasped, object exploration seems to be largely haptic. - Corners/indents & highly curved edges across both behavioural experiments [FIG.3] were explored more than flat surfaces. - Experiment 1 : an invisible weight change was explored more than a visible shape change, and even moreso than a less functional colour change [FIG.2,4]. - Experiment 2: an asymetrical shape was explored more than a symmetrical shape, unless the centre-of-mass was unusually on one side [FIG.4]. - Novel, functional or unexpected environmental changes seem to cause more exploration [FIG.4]. FIG. 4 REFERENCES [1 ] Demery, Z., Rios, V. E. A., Sloman, A., Wyatt, J., & Chappell, J. (201 0). Construct to Understand: Learning through Exploration. In Proceedings of the International Symposium on AI-Inspired Biology (pp. 59–61 ). Presented at the Proceedings of the International Symposium on AI-Inspired Biology. [2] Chappell, J., Demery, Z. P., Arriola-Rios, V., & Sloman, A. (in press). How to build an information gathering and processing system: lessons from naturally and artificially intelligent systems. Behavioural Processes. [3] Demery, Z. P., Chappell, J., & Martin, G. R. (2011 ). Vision, touch and object manipulation in Senegal parrots Poicephalus senegalus. Proceedings of the Royal Society B, 1 –8. doi:1 0.1 098/rspb.2011 .0374