Even in our modern world, where transportation and communication technologies can lead to cultural homogenization by making it easy to share ideas and objects, it's still possible to see distinct regional differences. In the US, for instance, adobe houses are primarily found in the Southwest, while brightly painted clapboard houses have a distinctly New England feel. Differences like these are driven partly by culture: Inhabitants of these areas build, or built, houses like those of their ancestors, who built houses like those of their ancestors, and so on. But local ecology also plays a role: Builders are more likely to use materials that are abundant nearby, and will produce homes specifically tailored for comfort in local weather and temperatures.

So which of these factors--culture or ecology--has a bigger impact on design? This question, which has been debated among anthropologists for years, was recently answered by two researchers from the University of California at Davis. Their work focused on variations in canoe designs across Polynesia, a series of archipelagos settled in four distinct clumps (Fiji/Samoa/Tonga, Cook/Society/Tuamotos/Austral, Hawaii, and New Zealand). 

 (An aerial photograph of a Polynesian island)

Because the general pattern of settlement throughout Polynesia is known, it was possible to assess whether designs in one region were influenced by designs from ancestral areas. On top of this, previous work has indicated that there were six "interaction spheres" within which the Polynesian peoples traded goods, labor, and ideas. In other words, inhabitants of one group of islands could teach others new and different ways of making canoes or canoe components. By considering these two methods of information transfer--"inheritance" and "learning"--the researchers could investigate the impacts of both old and current cultural influence on canoe design.

Canoes, of course, are used in the water, but very different water conditions can be found throughout the Pacific archipelagos. Islands with protective reefs or enclosed lagoons will have relatively still water, whereas those bordered directly by open ocean will have much choppier waters. Thus, vessels may need to be shallow-keeled in some areas and deeper-keeled elsewhere, or have two hulls instead of one. Some islands may have very different resources available for building, which should also influence canoe design; large islands, for instance, generally have a wider variety of building materials to choose from.

(Modern-built "voyaging canoe" in the traditional Hawaiian style)

In order to evaluate the relative importance of these cultural and ecological factors, the scientists generated a set of 27 different models to predict a total of 65 different canoe traits classified into 6 different categories: hull design, decoration, rigging, paddles, outrigger traits, double-hulled canoe traits. Each different model was capable of explaining different amounts of canoe variation, and models were ranked depending on how much variation they could explain. Therefore, models with higher ranks explain more variation.

Overall, the researchers found that models including both cultural and ecological factors were consistently ranked higher than models including either culture or ecology alone. Specifically, island settlement sequence, island size, and geological type of island seem to be the most important predictors of several Polynesian canoe traits, including hulls, sails, rigging, and outrigging. Occasionally, models that considered only cultural features did rank more highly than culture/ecology models, but in no case was this true of an ecology-only model. The culture-only models were particularly important for paddle and double-hull canoe traits. The authors also considered decorative traits, which generally serve no functional purpose. Perhaps unsurprisingly, these were not influenced by ecological factors, but they were also not particularly well explained by culture. Clearly, Polynesian canoes aren't ready to yield all their secrets just yet.

Cumulatively, the results indicate that neither culture nor ecology is the more important variable; rather, the two interacted to shape canoe design. It would be dangerous to generalize too much from these results, since the study only included one type of technology within a single region. All the same, the models clearly show that, in some cases at least, both environment and human-to-human transmission of information are important drivers of technological design and innovation.


Beheim, B.A. and Bell, A.V. 2011. Inheritance, ecology, and the evolution of the canoes of east Oceania. Proceedings of the Royal Society B 278:3089-3095.

Thanks to the following websites for providing the images used in this post: