This table is from a 2006 paper, “‘Irrational’ Searchers and IR-Rational Researchers” by Nils Pharo and Kalervo Jarvelin.
It’s a breakdown of a rational model of search engine use.
The paper is about how this model of how we use search engines isn’t very accurate (surprise!). In general, people use heuristics and creatively improvise. Or, in other words, they go to Google and type something in, then tweak it a bit when that doesn’t work.
You already know they do that, of course, and so do we, and you did even in 2006. (No slam against the paper – it needed demonstrating!). We just like this as an example of how complex – and unrealistic – everyday decisions get when you try to prescribe the ‘rational’ way of making them.
Can We Give System 2 Its Due?
Our BrainJuicer Book Clubs started a year ago and have become one of our most popular initiatives – an opportunity for people to get together and talk ideas and issues in a friendly, non-“hard sell” environment. (The free food and wine probably doesn’t hurt, either!) This blog post spins out of recent Book Clubs we ran on Daniel Kahneman’s Thinking, Fast And Slow. Normally in talking about Kahneman’s work we focus on the implications for consumer decisions – here we look at what it implies about our OWN business decisions.
Assume you like to think of yourself as a rational person, and you read about Daniel Kahneman’s work on System 1 and System 2. You accept its premise – that most of our decisions are fast, implicit and intuitive. What do you do? One thing you might conclude is, “Well, I need to use System 1 less and System 2 more.”
But this misunderstands the problem. The issue isn’t that System 2 makes better decisions but isn’t used enough. The issue is that System 2 is weak, and lazy, and even when we consciously try and think slowly and methodically we simply don’t have the tools to do it.
Don’t listen to this guy.
Would you make decisions differently if you made them in a foreign language? That’s the premise of this fascinating WIRED article, reporting on a study that takes one of Daniel Kahneman’s classic experiments and runs it through the scientific equivalent of Google Translate – with surprising results.
We seem to be more rational – or at least, less subject to specific biases – if we consider problems in a foreign language, probably because the increased cognitive effort of doing the translating short-circuits the “system 1” thinking we’d normally employ.
Kahneman’s dual-system ideas are very dear to my heart, so it got me thinking – maybe the way to get rid of all those annoying biases in surveys is to make everyone take them in a second language! Or perhaps not.
Some of my friends at BrainJuicer – a cosmopolitan bunch, with plenty of non-native English speakers – had extra points to make. One is that we’d expect the effect to wear off eventually with fluency – once you start dreaming in a second language you might not need to engage your ‘slow’ brain. Another is that a lot of people pick up English, in particular, by cultural osmosis – pop music and Hollywood – so words and phrases might well still carry strong instinctive associations, even if they’re not in your native tongue.