Daniel Kahneman’s Thinking, Fast and Slow is making quite a splash (the other day, I saw at Bristol airport that it is currently at the top of the bestseller list for non-fiction -- naturally, it still can’t compete with Fifty Shades of Gray). I haven’t read it yet, but people whose opinion I hold in high esteem tell me that it has been successful in striking the difficult balance between being accessible to a wider audience and scientifically accurate (for the most part at least) at the same time. The book summarizes research on cognitive and reasoning biases of the last decades, a research program in which Kahneman himself has been a major player. The conceptual cornerstone of the book is the (still) popular distinction between System 1 and System 2, the two systems which allegedly run in parallel underpinning all our cognitive processes, and which often conflict with each other.
Now, as I’ve stated a few times before (here for example), I am no fan of System 1/System 2 talk at all (not even of weaker versions, the so-called dual-process theories of cognition), even though I agree that the empirical findings on cognitive biases should be taken very seriously. (I also agree that there is something to the idea of debiasing as suppressing automatic processes.) So I was curious to see how Kahneman himself introduces the System 1/System 2 distinction, and took a quick look at the book (my husband was reading it during our holiday of a few weeks ago, after having gotten it from me as a birthday present – that’s what you get for having a nerdy wife). The first thing that struck me is that, on footnote 20, he lists some of the pioneers of dual-system theories, including Jonathan Evans, Steve Sloman and Keith Stanovich, and adds: “I borrow the terms System 1 and System 2 from early writings of Stanovich and West that greatly influenced my thinking” (he refers to their 2000 BBS article on individual differences in reasoning). But what is puzzling is that Stanovich himself now overtly rejects the conceptualization of the distinction in terms of systems, which unduly suggests reified entities, and now uses the process terminology instead (same with Jonathan Evans).
But perhaps most striking is what Kahneman says in the conclusion of the book:
This book has described the workings of the mind as an uneasy interaction between two fictitious characters: the automatic System 1 and the effortful System 2. You are now quite familiar with the personalities of the two systems and able to anticipate how they might respond in different situations. And of course you also remember that the two systems do not really exist in the brain or anywhere else. “System 1 does X” is a shortcut for “X occurs automatically.” And “System 2 is mobilized to do Y” is a shortcut for “arousal increases, pupils dilate, attention is focused, and activity Y is performed.” I hope you find the language of systems as helpful as I do, and that you have acquired an intuitive sense of how they work without getting confused by the question of whether they exist. Having delivered this necessary warning, I will continue to use the language to the end. (p. 415, emphasis added)
Philosophers will immediately be reminded of debates on scientific fictionalism. Now, while I am not prima facie opposed to scientific fictionalism as such (and if anything at all, I’m probably more of an instrumentalist than anything else), here I think the heavy use of these fictional characters, System 1 and System 2, is quite pernicious. There are good reasons why people like Jonathan Evans and Keith Stanovich have stopped using the system terminology; ultimately, these hypostasized concepts had become cumbersome and misleading for the research program itself. (There have been other important modifications in the original conceptualization of the distinction, such as the idea that biases belong exclusively to type-1 processes and are corrected by type-2 processes, which is no longer endorsed by someone like J. Evans – see this informative review of the research program by Keith Frankish in Philosophy Compass.) It is also puzzling why Kahneman would issue this ‘necessary warning’ only at the very end of the book, by which time the reader has already fully assimilated the 'existence' of these systems. Perhaps he thought that this would be a convenient simplifying device in a work intended for a wider audience, but the truth is that some of Kahneman’s collaborators (e.g. Shane Frederick) continue to use the System 1/ System 2 distinction also in their scientific work.
Either way, I am not convinced; even if seen as fictional entities, I submit that it is best to stop using the misleading System 1/System 2 terminology altogether. This seems to me to be an infelicitous case of factionalism in science, but I’m curious to hear what others think.
Recent Comments