How persistent are intuitive (erroneous) beliefs?

My motivation for posting this blog is simple: I am wondering whether it is possible for humans to ever truly internalize counterintuitive scientific principles like evolutionary theory or Newtonian (let alone Einsteinian) physics.

According to developmental psychologists like Elizabeth Spelke or Susan Carey, and cognitive anthropologists like Pascal Boyer and Dan Sperber, humans are endowed with inference mechanisms that enable them to acquire knowledge of the world (these inference mechanisms are known by several terms, such as core knowledge, conceptual modules or intuitive ontologies). Sometimes these inference mechanisms are at odds with scientific principles. A well-studied example is impetus physics, the view that inanimate objects, in order to be propelled, have to be laden with a force (impetus) by an agent or another object in order to be set in motion. This impetus physics yields a lot of imprecise predictions: for example, over 50% of adults believe that a ball, being launched by a sling, will continue in a curvilinear path, or that a ball dropped by a running person will fall straight down instead of describing a parabolic path. Newtonian physics, in contrast, predicts a parabolic path, a prediction only consistently made by people with a college training in physics (see McCloskey's 1983 review in Scientific American to get an idea).

However, an ingenious experimental procedure by Kohhenikov and Hegarty (2001), Psychonomic Bulletin & Review, 8) shows that even expert physicists are guided by the intuitive impetus physics under some conditions.



In their study, expert physicists had to remember the location of objects 'frozen' after moving along a trajectory. Afterwards, the objects vanished, and they had to recall their position. Typically, they remembered the object somewhat further along its trajectory, along a path best described by impetus physics, not Newtonian physics. So the question this raises for me is: are we ever able to truly internalize scientific and other forms of formal knowledge, or are we somehow always guided by our intuitive beliefs?



Another study that is now available through online first in Cognition also pertains to this puzzle. This study by Kelemen and Rosset (in press) examines the tendency of adults to provide teleological (i.e., goal-directed) explanations. It builds upon earlier work by Kelemen and others which showed that preteenage children have a preference for teleological explanations, they are 'promiscuous teleologists'., meaning that they see the function of a specific object as a necessary and sufficient explanation for that object. For example, when offered the choice to explain why rocks are pointy, they consistently choose teleological accounts (e.g., so that animals can scratch on them when they get itchy) over mechanistic explanations (e.g., bits of rubbish piled up over time). Not only do children prefer such explanations, they also offer them as accounts for why entire organisme exist (e.g., lions are to go in the zoo) or natural kinds (e.g. clouds are for raining). Adults typically don't endorse teleological accounts once they have acquired non-teleological causal mechanistic explanations. Although mountains can be climbed, few people would endorse the claim that mountains are there to climb on.

However, Tania Lombrozo and colleagues (2007, Psychological Science 18) recently found that Alzheimer patients have a re-emerging preference for teleological explanations: they are more prone to believe that water is there, for example, so that animals and plants can drink. So perhaps our promiscuous teleology is not fundamentally replaced by causal mechanistic explanations, but only masked by our explicitly acquired knowledge. This idea is strengthened by the study by Kelemen and Rosset's study, where adults were presented with teleological and non-teleological explanations. When they had to judge these explanations at a glance (speeded judgments), they were again more prone to endorse teleological accounts such as ‘the sun radiates heat because warmth nurtures life’.


  • comment-avatar
    Olivier Morin 23 February 2009 (16:23)

    Thanks a lot, Helen, for this stimulating post! I have two questions: first, you state that “Adults typically don’t endorse teleological accounts once they have acquired non-teleological causal mechanistic explanations”. Really? My own impression of adults (outside academic circles), and particularly of religious adults, is that they are perfectly able to maintain both kinds of explanations at the same time. Aren’t they? My second question is broader: what would be the point of “internalizing” scientific theories like Einsteinian physics or Evolution? My impression is that these theories are just not here to be internalized. Scientific theories are here to be accepted and used when they are useful (and that is not everyday), but they can be used in tht way without being intuitive, merely by being reflexively endorsed.

  • comment-avatar
    Helen De Cruz 23 February 2009 (18:36)

    Thank you, Olivier With regard to your first question, of course teleological and mechanistic explanations are not mutually exclusive. And in many cases teleological explanations seem warranted, e.g. to explain human action (why did the parents install a safety gate at the bottom of the staircase? To prevent their baby from climbing them). In such cases the goal of the action is a necessary and sufficient explanation. But in the cases Kelemen discusses, teleological explanations are unwarranted: animals or natural kinds are not there because they are ‘there for something’. Indeed, even in biology, it is now considered incorrect to say that eyes are there because they help us to see, but rather, they are there because of the [i]effects[/i] they had on the organisms in which they evolved (see e.g., Millikan and Papineau). In such borderline cases, I think it is indeed correct to say that educated adults have two kinds of explanations, teleological and non-teleological, and that these explanations can co-exist. However, once the explicit, mechanistic explanations become unavailable, as in the case of the Alzheimer patients, it seems that humans rely on the intuitive beliefs they had as children. Your second question: as this year marks 150 years of [i]On the Origin of Species[/i], there is a lot of talk about the creationism/evolution debate. In the US, many people do not endorse evolutionary theory (about half of the population), yet everyone has seen something of evolutionary theory in their highschool education. And even those who say that they do endorse evolutionary theory don’t always have a good understanding of the theory. Perhaps intuitive biases like teleology prevent a correct understanding of the theory? The ‘internalization’ of scientific theories might be interesting from the perspective of science education.

  • comment-avatar
    Olivier Morin 24 February 2009 (00:07)

    Thanks for the response, Helen, that made me realize that my questions were a bit too sketchy and not specific enough. Here’s some clarification. For the first point: I meant that unwarranted and false teleological explanations can persist alongside mechanistic explanations. Think for example of this famous example from Evans-Pritchard: we know that the roof tumbled down, and we know that termites had eaten it from the inside, that the wind shook it and that all this was the cause that made it happen. But still, this does not rule out witchcraft. Witchcraft was the real cuse behind the wind, the termites, etc. I am talking about healthy, normal adults endorsing unwarranted teleological explanations that seem (to them) perfectly compatible with mechanistic explanations that they also endorse. As for the second question, again, I was unclear. I meant that people can and do endorse theories that they do not internalise at all. For example, I endorse Relativity Theory, but, alas, so far am I from having internalized it that I could not even explain it to you satisfactorily. Does it stop me from accepting the theory? Not at all! Philosopher of science Gaston Bachelard had a theory of “epistemological obstacles” taught to every philosophy student in France. An epistemological obstacle is a folk psychological intuition that keeps you from understanding a scientific theory. According to Bachelard, a scientist is someone who has overcome every epistemological obstacles and converted his mind to scientific truth. We all need to overcome these obstacles if we are to endorse scientific theories. I think sociologists of science have done an excellent job of disposing of Bachelard’s theory. I also tend to agree with sociologists of science like Bruno Latour, when they say a scientist does not need his personal intuitions to represent in any way the consequences of the theory he is working on – anymore than, say, a Wall-Street trader needs to internalize the ruthlessness and risk-proneness of his job in order to be a good trader. One can be a wonderful theoretical physicist and still be driven by folk intuitions when it comes to catching a ball. Why should we expect things to be different?

  • comment-avatar
    Nicolas Baumard 24 February 2009 (00:36)

    People’s resistance to scientific explanations is indeed very interesting. But are teleology and impetus genuine intuitions or rather (erroneous) folk theories? For instance, let’s go back to McCloskey’s study. In a pen-and-paper experiment, he has shown that people’s answers are compatible with impetus theory. In their everyday life though, they are perfectly able to predict a ball’s trajectory with very good accuracy (think about tennis or football). It seems to me that they have good intuitions (for good adaptive reasons), but that they display bad theories when asked to solve a theoretical problem on a paper. As for teleology, I am also not sure that it is a genuine intuition and not a folk theory. Indeed, in everyday life, you don’t need to reflect upon the purpose of rocks, animals or the meaning of life. You just use rocks, eat or observe animals and live your life. A recent [url=]article [/url]by Greif et al. in Psychological Science makes this point and contradicts Kelemen’s view. Here is the abstract: [i]Thirty-two preschool children were given opportunities to ask questions about unfamiliar artifacts and animals. The children asked ambiguous questions such as “What is it?” about artifacts and animals alike. However, they were more likely to ask about the functions of artifacts, but about category membership, food choices, and typical locations of animals. They never asked questions about either artifacts or animals that would be considered inappropriate by adults. The results indicate that children hold different expectations about the types of information important for categorizing living and artifact kinds.[/i] Spontaneously, children do not seem to be “teleologists” (except with artefacts, again for good reasons). However, when children start thinking about the world, or when someone asks them theoretical questions such as “Why on earth are there some rocks? Why do we live? Why is the sky blue?”, then they make up spontaneous and (often) erroneous theories, like teleology. After all, this is not bad at all: 2400 years ago, teleology also looked like the most promising theory for Aristotle!

  • comment-avatar
    Dan Sperber 8 March 2009 (12:08)

    Than you Helen for this thought provoking post. As suggested in Nicolas Baumard’s comment, it may be important to contrast not just intuitions with scientific theories, but, in three tiers, intuitions with theories, and among theories, improvised theories with culturally transmitted ones, and, among these, folk theories with scholarly theories (with cases in between, sure, but not so as to make these distinctions inappropriate or irrelevant). Intuitions are best studied by eliciting expectations (as, for physics, where do you position yourself to catch a ball – which may at odds with your answer to the question where you should position yourself, given that the ball was thrown from such a place in such a way). Asking questions typically yields reflective, theory-driven answers, more or less improvised, and, if based on transmitted knowledge, more or less sophisticated. The common simplistic contrast between folk and scientific theories (as between folk and scientific psychology) has been a source of great confusion, both in the study of intuitive knowledge and of cultural ideas.