New research paper about Saami reindeer herders and gift-giving

I’m enormously pleased to say that I have a new bit of research published. The paper is about gift-giving in a group of reindeer herders living in Norway. It’s the culmination of nearly two and a half years’ worth of work – from our initial planning, though some months of hanging out in the (surprisingly warm) arctic, to endless rounds of statistics and writing and revising and statistics and *facepalming* and statistics and *headdesking* and statistics and revising.

The paper itself is published in a journal called Behavioral Ecology but because of funding constraints we weren’t able to make it open access – to my great chagrin (but you can download a PDF for free here or read it online here, or feel free to email me if you want a copy).

What I want to talk about in this post is a bit about my experiences of doing the research (also see my posts from when I was in Norway talking to the reindeer herders) and show you some of our results in an interactive fashion you can’t currently do in scientific articles.

Into the field (of memories)

After many months of reading about Saami reindeer husbandry and the evolution of cooperation, scrabbling for funding and having my questionnaires torn apart (in a good way) by various colleagues, I thought I was ready to interact with humans in the real world and jumped on a couple of aeroplanes to the far north.

Telling people you hang out with indigenous reindeer herders deep into the Arctic Circle sounds cool down the pub but I was struck by just how un-exotic it was. True, reindeer husbandry is a lifestyle far removed from my own, the landscape is vast and can be hard, and the clouds look strange up there, but at the same time there was a certain everyday-ness within the experience – replacing tubes of toothpaste at the supermarket, cooking pasta (Norway is expensive…) or bumping into familiar faces at the local café.

A lot of the unusual perhaps goes unnoticed when you’re at work and all you want is for people to talk to you. And not many people talked to me. A few weeks before arriving, the Norwegian Institute for Nature Research (known as NINA) published a report suggesting that the high reindeer mortality rate in our study area was due not to predation but rather too many reindeer on not enough pasture. Many herders took this to mean they were being blamed for their livestock dying.

The relationship between Saami herders and outsiders – particularly researchers and governments – was already strained (through years of mistrust on both sides). As a true outsider – a Brit, a student, an anthropologist who’d shaved his head and grown his beard to indulgent, extravagant, almost colossal proportions – maybe I was met with greater distrust and no small amount of curiosity. Regardless we managed to interview 30 reindeer herders. Without the generosity of these folks, or the support of Jon Mikkel, there would be no study to be wittering on about.

Some interactive results

One of the downsides to scientific publishing as it stands (beyond the ridiculous cost of publishing open access) is that the output is very much based on an old-school reading-things-on-paper kind of model. This is especially apparent in the figures, printed as pictures rather than what they really are: visual representations of data – scientific objects in their own right.

Researchers are combating this to an extent by giving away the data and sometimes the code that produced their results (our data is on Dryad and our code is on GitHub) but still I can’t help but think we’re missing a trick. Web technologies keep advancing in astonishing ways and we really should be trying to take advantage of that. Just look at some of the cool interactive visualisations like the New York Times’s network of Oscars contenders or FiveThirtyEight’s predictions for the last UK general election (ignore that it was way off) or the Explained Visually project or SciDev.Net’s stunning piece about urbanisation.

It would be unfair to expect researchers to keep up with the cutting-edge of web development – we already have to be project managers, teachers, communicators, designers, writers, programmers, statisticians and administrators, among other things – but I think we should take some inspiration from these to make our research interactive and engaging. Sites like plotly or WordPress plugins like Visualizer (which I use below) make this kind of thing so very straightforward.

So in that spirit, I present some very simple interactive visualisations of results from our paper.

Who were the herders?

The two pie charts below show the ages and sex ratio of people who took part in our study. Hover over them for more details.

What did they do?

The network below shows all herders in the district. Circles represent licensed herd owners, coloured by siida membership. Hover over the circles for more information. Links between people are gifts – thicker lines are larger gifts. Play with the layout and filter buttons.

  • James Baker

    On ‘Web technologies keep advancing in astonishing ways and we really should
    be trying to take advantage of that. Just look at some of the cool
    interactive visualisations like the New York Times’s network of Oscars contenders or FiveThirtyEight’s predictions for the last UK general election (ignore that it was way off) or the Explained Visually project or SciDev.Net’s stunning piece about urbanisation.’

    The problem, as far as I see, is that those web technologies are shifting sand that are hugely problematic to preserve (see the wonderful, challenging work of my colleagues at the UK Web Archive http://britishlibrary.typepad.co.uk/webarchive/ – especially the more techy posts). So in a way those dull flat viz, underpinned by data deposit, are more sustainable – which is as, if not more, important to scholarly that flashy, viz, bang. Not to decry what you’ve done (you know I enjoy the fact that you summarise and make more accessible your research outputs in blog form), just to say that the Dryad, Zenodo, Git deposit route has clear virtues.

    On which, on a recent project I was a CI on https://github.com/UCL-dataspring we used IPython Notebook and Bokeh to get some reproducible interactivity going (download and open in-browser the html at https://github.com/UCL-dataspring/visualisations/tree/master/diseases for example). I even wrote some rather simple/dumb R https://github.com/UCL-dataspring/visualisations/blob/master/find_all_figures/micro/imagesinbook_bypage_bysize.R. Your R is much better than mine… (though I figured out how to use `facet_grid` properly the other day and am now all over it like a cheap suit)