Defensive Tool Use in a Coconut- Carrying Octopus
Total Page：16
File Type：pdf, Size：1020Kb
Magazine R1069 one to four in the mother tongue of our own body differ in line with of the participants. E continued to culture-specific preferences for how Defensive tool demonstrate the movement sequence to conceive of spatial relations. These use in a coconut- until participants could reproduce it results support the view that, at least by themselves. Then, E rotated them in some domains, cultural diversity carrying octopus 180 degrees around their own axis, goes hand in hand with cognitive and positioned himself behind them diversity, and a cross-cultural (Figure 1: Rotation 1). E asked the perspective should play a central part Julian K. Finn1,2, Tom Tregenza3 participants to ‘dance again’. in understanding how variable adult and Mark D. Norman1 After the participants performed, cognition is built from a common E rotated them back into their original cognitive foundation. The use of tools has become a orientation (Figure 1: Rotation 2). If benchmark for cognitive sophistication. participants coded a RLRR dance in Supplemental Data Originally regarded as a defining egocentric coordinates they should Supplemental data are available at http:// feature of our species, tool-use produce a RLRR sequence after both www.cell.com/current-biology/supplemental/ behaviours have subsequently been Rotations 1 and 2. Alternatively, if S0960-9822(09)01898-3. revealed in other primates and a participants coded a RLRR dance in growing spectrum of mammals allocentric coordinates they should Acknowledgments and birds [1]. Among invertebrates, produce a LRLL sequence after We are indebted to students and teachers however, the acquisition of items that Rotation 1 and a RLRR sequence after in Leipzig and at Khomxa Khoeda Primary are deployed later has not previously Rotation 2 (see also Supplemental School. We’d also like to thank the ≠Akhoe been reported. We repeatedly Movie 1). Any response that did not Hai||om and the farmers of Farm 6. Finally we observed soft-sediment dwelling match one of these two patterns owe thanks to Katja Liebal, Heinz Gretsche, octopuses carrying around coconut was coded as ‘other’. These were Melissa Bowerman, Disa Sauter, Emma shell halves, assembling them as a Cohen, Roger Mundry and Stephen Levinson either mixed responses (RLRR, LRLL, shelter only when needed. Whilst for insightful commentary. This research LRLL) and/or failures to memorize the being carried, the shells offer was funded by the Max Planck Society for sequence (RLRR, LRLL, RLRL). the Advancement of Science. The authors no protection and place a We tested 50 German and 35 Hai||om declare no competing financial interest. requirement on the carrier to use children between 4 and 12 years of age Correspondence and requests for materials a novel and cumbersome form of (German: M = 7;3; SD = 2;7; Hai||om: should be addressed to D.B.M.H (haun@eva. locomotion — ‘stilt-walking’. M = 7;8; SD = 2;0). There were 40 boys mpg.de). To date, invertebrates have and 45 girls (German: 25 boys, 25 girls; generally been regarded as lacking Hai||om: 15 boys, 20 girls). German References the cognitive abilities to engage children produced 60% egocentric, 6% 1. Boas, F. (1911). The Mind of Primitive Man in such sophisticated behaviours. allocentric and 34% other responses. (New York: Macmillan). Putative examples of tool use do exist 2. Pinker, S. (2002). The Blank Slate: The Denial Hai||om children produced 20% of Human Nature in Modern Intellectual Life among invertebrates — perhaps most egocentric, 54% allocentric and 26% (New York: Viking). convincingly in the form of the use of 3. Segall, M.H., Campbell, D.T., and Herskovits, other responses. This difference in M.J. (1966). The Influence of Culture on Visual leaves or pellets of sand to collect and response distribution is statistically Perception (Indianapolis: Bobbs-Merrill). transport food in various ant significant (Fisher-exact,p < 0.0001). 4. Levinson, S.C. (2003). Space in Language and species — but these behaviours have Cognition: Explorations in Cognitive Diversity Extracting ‘other’ responses, German (Cambridge: Cambridge University Press). been regarded as distinct from tool children produced 91% egocentric 5. Haun, D.B.M., Rapold, C.J., Call, J., Janzen, use in higher animals on the grounds G., and Levinson, S.C. (2006). Cognitive and 9% allocentric responses. Hai||om cladistics and cultural override in Hominid that they only occur in response to children produced 27% egocentric spatial cognition. Proc. Natl. Acad. Sci. USA very specific stimuli 2[ ]. This highlights and 73% allocentric responses. 103, 17568–17573. a key feature of widely used functional 6. Majid, A., Bowerman, M., Kita, S., Haun, These distributions were significantly D.B.M., and Levinson, S.C. (2004). Can definitions of tool use 3[ ] — simple different from each other (Fisher-exact, language restructure cognition? The case for behaviours, such as the use space. Trends Cogn. Sci. 8, 108–114. p < 0.0001) and different from chance 7. Li, P., and Gleitman, L. (2002). Turning the of an object (or objects) as shelter, (50%) in both populations (German: tables: language and spatial reasoning. are not generally regarded as p < 0.0001, binomial test; Hai||om: Cognition 83, 265–294. tool use, because the shelter is 8. Levinson, S.C., Kita, S., Haun, D.B.M., and p < 0.05, binomial test). The frequency Rasch, B.H. (2002). Returning the tables: effectively in use all the time, whereas of egocentric vs. allocentric responses Language affects spatial reasoning. Cognition a tool provides no benefit until it is 84, 155–188. did not correlate with age (German: 9. Burgess, N. (2006). Spatial memory: how used for a specific purpose. This p > 0.05, point-biserial; Hai||om: egocentric and allocentric combine. Trends rules out examples such as the use of p > 0.05, point-biserial). The absence Cogn. Sci. 10, 551–557. gastropod shells by hermit crabs, but 10. Graziano, M.S.A., Cooke, D.F., and Taylor, of an increase of locally dominant C.S.R. (2000). Coding the location of the arm includes situations where there is an responses with age is surprising given by sight. Science 290, 1782–1786. immediate cost, but a deferred benefit, previous research documenting an such as dolphins carrying sponges increase in cross-cultural differences 1Max Planck Institute for Psycholinguistics, to protect against abrasion during with age [5]. Samples of younger Nijmegen, 6525 XD, The Netherlands. foraging [4] and where an object is 2 children are needed to document the Max Planck Institute for Evolutionary carried around in a non-functional form Anthropology, Leipzig, 04301, Germany. developmental trajectory of this task. 3University of Portsmouth, King Henry to be deployed when required [5]. In summary, we show that the ways 1st Street, Portsmouth PO1 2DY, UK. The dramatic and complex in which we memorize movements E-mail: [email protected] colour and shape change abilities Current Biology Vol 19 No 23 R1070 range of behavioral states — emerged comes from the requirement of the and active on the seafloor (Figure 1A); octopus to correctly assemble the occupying empty gastropod shells, separate parts (when transporting discarded coconut shell halves two shells) in order to create a single (Figure 1B) or other human refuse; functioning tool. or buried within the substrate (with The behaviour reported here is or without accompanying shells; likely to have evolved using large see Supplemental Movie S1 in the empty bivalve shells prior to the Supplemental Data available on-line relatively recent supply of the clean with this issue). When flushed from and light coconut shell halves shells by the observer, individuals discarded by the coastal human quickly reoccupied the shells. On four communities adjacent to the marine occasions (three in Northern Sulawesi, habitat of this species. one in Gilimanuk, Bali), individuals Ultimately, the collection and use of were observed to travel over objects by animals is likely to form a considerable distances (up to 20 m) continuum stretching from insects to while carrying stacked coconut shell primates, with the definition of tools halves below their body (Figure 1C; providing a perpetual opportunity Movie S1). For all instances of this for debate. However the discovery behaviour, observing divers (JF, MN) of this octopus tiptoeing across the remained static for up to 20 minutes at sea floor with its prized coconut 1–2 metres from stationary octopuses, shells suggests that even marine which emerged from the cover of invertebrates engage in behaviours one or two shells halves, arranged that we once thought the preserve the shell(s) under the arm crown, and of humans. departed. Two shell-less octopuses were also observed to extract Supplemental Data previously un-encountered coconut Supplemental data are available at http:// shells buried in the substrate, aided www.cell.com/current-biology/supplemental/ by jets of water to flush mud from S0960-9822(09)01914-9. shells (Movie S1). References To carry one or more shells, this 1. Hansell, M., and Ruxton, G.D. (2008). octopus manipulates and arranges the Setting tool use within the context of animal construction behaviour. Trends Ecol. Evol. 23, shells so that the concave surfaces 73–78. Figure 1. Veined octopus, Amphioctopus are uppermost, then extends its arms 2. Baber, C. (2003). Cognition and Tool Use: marginatus. around the outside and walks using Forms of Engagement in Human and Animal (A) Emerged on sand. (B) Using coconut shell Use of Tools (Boca Raton: CRC Press). the arms as rigid limbs. We describe 3. Beck, B.B. (1980). Animal Tool Behavior: The halves assembled as shelter. (C) ‘Stilt-walk- this lumbering octopedal gait as ‘stilt Use and Manufacture of Tools (New York: ing’ while carrying two stacked coconut shell Garland STPM Press). halves (see Movie S1). Photos: M. Norman walking’ (see Movie S1).