Discover more from The Works in Progress Newsletter
Notes on Progress: The eye of the tiger
What makes tigers different from one another?
Notes on Progress are pieces that are a bit too short to run on Works in Progress. You can opt out of Notes here.
I’m reading this book called Dersu the Trapper, written by the late C19th geographer and explorer Vladimir Arsenyev and recommended to me by my friend and collaborator @inpurier, Yao Zheng. At the end of one chapter, Arsenyev quotes:
To the Lord Tiger who dwelleth in the Forest and the Mountains. In ancient days in the time of the Khan dynasty, he saved the state. To-day his Spirit bring happiness to man
‘Lord’ applies to tigers (also perhaps lions, elephants) in a way that it simply doesn’t to other mammals of equivalent or larger mass – hippos for example. We are uplifted by the dappled beauty of the tiger, their power, their sinuous yet purposeful gait.
Yao and I have collaborated on academic work once or twice; he introduced me a few years ago to a comparative psychologist, Professor Yanjie Su whose work is well known within China. She had a dataset on the personality of the Siberian Tiger (also called the Amur tiger) that she was keen on someone having a look at. We did.
It took a while for Professor Su to shake loose the data file which had become tangled with the peppermints, packets of tissues, worn to a stub lipstick, chipped protractors that accumulate at the bottom of every file-drawer. As you may know. We were eager to see the data because tigers!
Siberian tigers are the world’s largest cats. They live near the border between Russia and China in snowy forested mountainous territory. A male can have a home turf that’s around 2,000 square km. Cubs can spend up to two years with their mother. This suggests that it takes a while to learn to be a tiger. They have massive crania. This is also consistent with needing a lot under the hood to be an effective tiger. The tiger encephalization quotient (ratio of observed to predicted brain size given a mammal’s mass) is bigger than that of the lion, leopard, or jaguar. This doesn’t prove they are intelligent, but it is consistent with their being smart cats. They eat deer, bear, boar and fish. Their stripes, like our fingerprints, are unique.
Professor Su emailed the data to Yao and me. I asked Abdel Abdellaoui to collaborate with us. He is a meticulous scholar; skilled and very careful which is exactly what one wants in a team-mate. A dreamy dataset is an invitation to articulate a research question crisply – what do you want to know? Can these data tell you? If so, how? There are often several different kinds of tests or analytical strategies one could implement. Later, reviewers may suggest approaches you have not thought of. At its best, collaboration is a small hive mind.
We wrote the manuscript during the Year of the Tiger which we all thought was frightfully auspicious until the pandemic made everything glacial yet not attractively milk-blue shiny and mountainous. It was the year of the Water Rabbit before the reviews were in and done. And what kind of lagomorph is a water rabbit anyway?
One reviewer asked that the introduction be re-written because much of it was manifestly irrelevant and colloquial. For example, we quoted from Sooyong Park’s gripping book in which he describes a mother tiger noticing a camera wire:
She let out an indescribable, blood-curdling growl, and the attack began. She caught the zoom cable that was dragged out with the lens when it fell off and yanked at it … [she] clawed off the camouflage shrubs.
Reviewer was correct that this is irrelevant. But we wanted to clothe the bones of the data with observations from the wild. We didn’t comply with the re-write; my co-authors bit their tongues against their better judgment. We hoped that readers who would glaze over at the Intraclass Correlations Coefficients, Procrustes Rotations and factanal would enjoy engaging with our key points. So, we kept the colloquial language and the irrelevance. The methods and results are still clear, we provide the data and the script, so there was no scientific trade-off. It’s a cool paper and I’m proud of our work. Tigers are poster children (poster Lords) for ‘the charismatic animal’; our research probed whether tiger personality has structure. This takes a bit of explaining.
Concerning personality, the best-studied animal is looking at you in the mirror. The Big Kahuna in personality research is the five-factor model. The five-factor model says that although my love is like a red, red rose, incomparable, unique, and individual in every way, it is also true, broadly speaking, that he could be described with reference to where he lies on each of five dimensions or spectra called: Openness to Experience, Conscientiousness, Extraversion, Agreeableness and Neuroticism. This is what’s meant by pattern or structure in personality. You can think of each of the five dimensions as being a vertical line anchored at the bottom with ‘very low’ and ‘very high’ at the top end. You could perhaps introspect about whether you are low, average, high on each of these traits. Where you hit each of these five ‘notes’ is your own tuning. The claim is not that these five dimensions by themselves capture all there is to know about your personality. But these five dimensions do pick up quite a lot of the stable differences in personality among us. The five-factor model says that personality is reasonably well described by our individual tuning in these five broad chords – and it recognises that we all have ornamentation and top notes.
The five-factor model was arrived at via a cute insight. People talk about each other. Endlessly. We have words for how we carve up the world. The adjectives we use to talk about people were sort of put into a hopper or coin-sorter, figuratively speaking. We used factor analysis to discover hidden commonalities among seemingly disparate adjectives. That’s a thumbnail of how the five-factor model of human personality emerged.
The tiger personality inventory developed by Professor Su and colleagues took a similar approach. They took all the adjectives that can be used to describe someone (in the Chinese language) and put them in a list. They then asked tiger experts to go through that list, retaining only the words that could reasonably be applied to tigers. They used a psychometric system for doing this to achieve agreement among the tiger-scholars before a word was included. Those ~70 words were then discussed with the feeders and veterinarians who would rate the tigers’ personalities. This meant that there was a shared understanding among raters about the meaning and nuance of each adjective.
Each of 152 tigers in the reserve was rated on the 70 adjectives by several raters (with an average of 3, but up to 12 raters per tiger). Those data were then explored to see whether any model (like the human five-factor model) fit them. The five-factor model did not fit. A model of three main dimensions didn’t fit either. The best model was two main dimensions in personality.
We had data acquired in the same way on a second smaller sample of 96 tigers from a second nature reserve. The big test of our two-factor model was to apply the model from the first sample to the second dataset to see if there was any congruence. If not, it would indicate that there was nothing generalizable or useful about our model. But we found that there was moderate convergence. Quantitatively, the congruence was 0.81 overall (where a value higher than 0.9 is considered a match, and any value higher than 0.8 is considered fair dinkum, matchwise). So tiger personality likely does have two key dimensions. In our analyses we called these Factor One and Factor Two. Which was unmemorable. We gave them more descriptive names for our write-up: Majesty and Steadiness because the items/adjectives (listed in Figure 1 in our paper linked at the end of this article) each fitted these terms. These two factors explained 38% of the differences in personality among the individual tigers.
The list of tiger-suitable adjectives was fun to read. I’m not a tiger expert. But. “Far-sighted, methodical, adamant, unapproachable,” gotta love them. Naughtily, we kept one that looks a bit odd: “ripen”. Following discussion with Yao, I learned that it really means mature, but there’s something fruity about ripen. We kept it as it was.
Each tiger sits somewhere on the spectrum of Majesty and each tiger also sits on the dimension of Steadiness. That was something I really stumbled over explaining in some of the short interviews following publication. Typically, a journalist would start with the notion that tigers were Either/Or. But like our extraversion/agreeableness dimensions, we sit on both spectra, we might be high in one, low on the other or high/low on both.
Our dataset included a few health and behavioural measures. We found that tigers high in Majesty mated more often and ate more live prey. Some dimensions on which we all fall somewhere, ‘healthiness’ for example, have a good end and a bad end. But personality is not quite like that. Daniel Nettle’s excellent short book Personality: What makes you the way you are argues that we are all ‘just different’. I highly recommend Daniel’s book, but there are people with simply revolting personalities. It’s to be expected. There are around 7 billion of us. Some people, inevitably given the combinatorial possibilities, will be ‘tuned’ on each of the dimensions to be overwhelmingly unpleasant. Which is why I lament that we don’t more often celebrate the high frequency of reasonable personalities that make many societies pleasant to live in. What causes personality is a different matter from describing it. We have no data on that in our tigers.
A couple of journalists asked ‘what’s the point of your research?’ Well, it ain’t gonna fix malaria. Ain’t gonna create a vast protected wilderness allowing the Lord Amur tiger to dwelleth in the Forest and the Mountains. But for those who are closer to the levers of power than we are, they first must notice before they will care. Any research that makes an endangered species more prominent can help. It is a constructive form of activism without unintended harmful side-effects. We didn’t conduct our analyses as tiger activists; we are scholars. We like to know more about this miraculous, marvellous yet fragile world we live in.
You can read Dr. Rosalind Arden et al’s paper here. She also recommends Sooyong Park’s book, The Great Soul of Siberia. If you enjoyed the piece, please share it.
Thanks for reading The Works in Progress Newsletter! Subscribe for free to receive new posts and support my work.