About Rationally Speaking

Rationally Speaking is a blog maintained by Prof. Massimo Pigliucci, a philosopher at the City University of New York. The blog reflects the Enlightenment figure Marquis de Condorcet's idea of what a public intellectual (yes, we know, that's such a bad word) ought to be: someone who devotes himself to "the tracking down of prejudices in the hiding places where priests, the schools, the government, and all long-established institutions had gathered and protected them." You're welcome. Please notice that the contents of this blog can be reprinted under the standard Creative Commons license.

Wednesday, April 28, 2010

Essential or Incidental?

I have a friend who decided a while ago that she doesn't want kids. She's got pretty sound reasons for not wanting to procreate, and she's always seemed happy with that decision. But recently she told me that she felt a pang of doubt when she saw a family sitting out on their patio, barbecuing and looking really cozy and happy together.

She confessed to me, "Now I'm starting to worry that I could be making the wrong choice, because I found the idea of being in that scene so appealing." My response: "Okay, imagine that same scene again, but instead of being there surrounded by your hypothetical children and husband, you're surrounded by your close friends. Now how do you feel about it?" She replied, "Oh. Actually, that feels just as good."

The thought experiment had revealed that she was reacting positively to the "Idyllic afternoon in the company of people you love" aspect of the scene, not specifically to the "Idyllic afternoon in the company of family" aspect. The latter is a subset of the former, of course, but she had erroneously assumed that the "family" component was a necessary condition of the scene's appeal to her, when in fact it was incidental. And it only became clear which components were necessary, and which incidental, by isolating them via thought experiment.

What I thought was interesting about this case is that it was a real-life occurrence of a problem researchers face all the time: you observe a phenomenon with a bunch of attributes -- say, A, B, and C -- and it's not clear which attributes are essential to the phenomenon, and which are irrelevant. Is A necessary and sufficient to produce the phenomenon, with B and C being totally irrelevant? Is the interaction between A and B the necessary cause? Would any one of the three attributes on its own be sufficient?

For example, if you observe that vegans tend to live longer than average, is that effect due to the lack of meat in their diets? The lack of all animal products? The higher amounts of plants in their diets? Is their veganism in itself completely irrelevant and merely correlated with some other feature that's responsible for the benefit, like higher levels of exercise? Or maybe the increased lifespan is caused by the combination of more plants and more exercise, and neither factor would be sufficient on its own?

If you're attuned to this problem of ambiguous causality, as good researchers usually are, you can try to isolate potential causes and study how the outcome changes when you change one factor at a time. But in your daily life, it's much easier to jump to conclusions without even realizing you're doing so, like my friend did.

The happy-family case study deserves particular mention, I think, because it's a special case of ambiguous causality, in which you form an overly-specific hypothesis instead of the correct, more general one. I touched on this phenomenon in an earlier post, in which I described a study where people were given the series of numbers "2, 4, 6" and asked to guess the rule that those numbers were following. Most people hypothesized something like "increasing even numbers" or "increasing by intervals of two", when in fact the correct rule was simply "increasing numbers." The evenness, and the intervals of two, were both incidental, the same way the family-ness of the scene my friend observed was incidental to the feelings it induced in her.

For a less contrived example, consider acupuncture. People found that sticking needles in specific "pressure points" prescribed by traditional Chinese medicinal theory helped alleviate pain, and they took that as validation of the pressure point theory. But when researchers tried sticking needles in other locations on subjects' bodies, they got the same effect. As far as science has been able to tell so far, the "pressure points" were incidental to the pain relief; the "sticking with needles" was the only essential component.

I have a suspicion that this fallacy might also come into play when people consider the effect of religion in their life. I've heard people describe how belonging to their church makes them happier, or makes them better people. But is the religiousness of the community essential to its beneficial effect? Or would any tight-knit community produce the same result? It's not necessarily obvious until you experiment with removing the component you thought was crucial -- although fortunately, as in the case with my friend, sometimes a thought experiment does the trick, and you can disentangle essential from incidental from the comfort of your armchair.

I got all excited when I couldn't find this phenomenon described on Fallacy Files, because one of my life goals is to discover a new fallacy or bias and get the privilege of picking its official name. But after I'd already daydreamed about what I was going to call it, I did a little more googling and discovered it might actually fall under the heading of an obscure fallacy that's already been coined: "Confusion of Essence and Accident," in which you "assume that some accidental feature or property of something essentially belongs to it." Alas, it looks like the world might have to wait a little longer before "Julia Galef's Totally Brilliant New Fallacy" is added to to the logic textbooks.


  1. Unitarian Universalists figured this out a long time ago. You don't need doctrine or magic to have the social and emotional benefits of a church congregation.

  2. Don't forget the additional fallacy that measuring behavior has no effect on it, even if the measurements are taken in the seeming isolation of spreadsheets, far from the social laboratory.

  3. Isnt that a case of good ol' separating the variables?

    In every case where the outcome of an event is determined by many variables as an experimenter you have to make some assumptions on what those variables are and then perform careful experiments where you change only one at a time and see if the result changes.

    If it doesnt change then it means that the assumption you made was incorrect.If it did then you map the variable value-result table.Of course the problem might lie in the fact that there is another variable that affects it that wasnt included in your assumption.It is even possible that changing variable A or B independently doesnt affect the outcome but changing both at the same time does in which case (A,B) would be a valid variable.

    Those are basically the issues that might come up in a situation like this.This is difficult enough for scientists do to properly so its no surprise that people arent particularly good at intuiting it.We are not statistics machines with built in memory.Far from it actually...

    Call it misidentifying the variables if you want.

    But were all variables identified properly in the examples you gave? I think not.

    In your thought experiment you proposed to change the value from kids to friends.Your friend performed it for herself and came to the same conclusion as you did.But it necessarily work out like this? People enjoy the company of people with whom they share strong bonds. The stronger the bond the greater the pleasure. Isnt bond strength a better parameter? How strong is the bond parents have with their children compared to their friends? (i dont know i dont have kids). Is it something that can be predicted reliably? I sure as hell dont know but it is a bit more complicated problem than it seems at first.

    In the number sequence experiment example things are even less clear cut. There are variables that affect the result that lie outside the subject as we all know. The way the question is posed affects the answers. The subjects were asked to identify a rule that COULD produce this sequence. Then they could test several rules by asking questions. But depending on how this problem was posed to them they might have assumed that this rule HAD to produce this sequence. Its not an unnatural assumption to make, unless it has been made explicit that this is not the case. In IQ tests and the like when people are given a sequence they are asked to continue it. That implies that the sequence is unique. On the other hand what you have here is a rule that can produce many different sequences that comply to that rule. You are not looking for a sequence but for a rule and thats not the common problem posed.Simply asking for the rule doesnt rule out (pun intended) the possibility that you re looking for a predetermined sequence.Remember, language is flawed and limited and forces you to make assumptions. I can pose this problem in a way that will make people give the "right" answer and in a way that will make them give the "wrong" one. Did the experimenters measure what they thought they were measuring or did they just set up the subjects? Did they identify the variables?

    As for the church thing.If your friends said that belonging to the church makes them feel good then thats more or less stating a fact.They didnt ever claim that the religiosity necessarily causes that happiness. Maybe thats what they meant of course, i dont know that.

    This is an interesting subject but i am not sure its posed correctly. It might be slightly off focus.


  4. You don't need doctrine or magic to have the social and emotional benefits of a church congregation.

    But the former can sometimes have the effect of intensifying the latter.

  5. Julia,

    Did you consider the possibility that your friend was not honest with herself (and hence nor with you) about the scenario you propounded feeling "just as good"? Mightn't she have actually just been looking to you for support in her fight against her biological desire to mate?

  6. Julia: "I have a suspicion that this fallacy might also come into play when people consider the effect of religion in their life. I've heard people describe how belonging to their church makes them happier, or makes them better people. But is the religiousness of the community essential to its beneficial effect? Or would any tight-knit community produce the same result? It's not necessarily obvious until you experiment with removing the component you thought was crucial -- although fortunately, as in the case with my friend, sometimes a thought experiment does the trick, and you can disentangle essential from incidental from the comfort of your armchair."

    But whether social scientists can attribute it to social contact or faith or some combination thereof, the fact still remains that present research (and it is considerable) seems to indicate that religious people are generally more happier than the non-religious segment of the population.[1]

  7. Julia - your "experiment" regarding whether adult friends can satisfy the same social needs as having children is far to simplistic to be of value. After the party, the adults go home. The kids don't. The adult interaction will satisfy many of the social needs. And may be the right choice for some. But the love of a child and having your own children is unique. While kids can be pain in the butts, so to can adults. Experiencing both is probably the most rewarding. But to have the love of your own kids is unbelievable.

  8. I agree with Kostas argument about the series of numbers.
    If the question is: guess the rule that the numbers are following?
    Then any of the answers is correct (incremental numbers, interval of 2, even numbers). A way of showing this is to turn around the question.
    Show me a series of numbers that increase by interval of 2...2-4-6. Is this true? Yes
    Show me a series of numbers that are even and in an increased way...2-4-6.Is this true? Yes
    Show me a series of numbers that go in increased way...2-4-6.Is this true? Yes

    If the question was "Guess correctly the rule that I have in mind for the progression of these numbers" or "Guess the most basic rule that these numbers can follow" then I would agree with your reasoning.
    You actually didn't follow your own advice, and you omitted the essential factor in your question, because you assumed that the people that was going to answer had your assumption in their minds.

    In the thought experiment with your friend, there is also the fact that the 2 possibilities are not mutually exclusive. You can have the happiness with your close friends AND with your kids. So by telling her that the happiness factor is the closeness of the relation, you are actually assuming that:
    1- One is the same than the other.
    2- She is not letting her previous assumptions color the values of each option (as Quinn mentioned). She is already conflicted, she may suffer from confirmation bias (she asked you, because she assumed you were going to give her the answer that she wanted to hear).
    3- That having BOTH possibilities at the same time is not better. Therefore selecting one over the other could be worse.

    A separate evaluation of your reasoning would involve the biologic imperative of reproduction that we as humans can socially repress. If your friend was telling you that she decided only to sleep 2 hours a night because it would give her a more time to do things that she likes, or if she had told you that she was only going to eat 300 calories a day because she wanted to save money for a trip to Europe, you would probably have used the biological imperative that she can't ignore these functions. The reproductive function is optional in our society, but it may coded as importantly or even more than the other 2(lots of people lose sleep and don't eat if they can get some sex!).
    Ignoring our biological function may have long term effects that can't be taken away by wishful philosophical thinking.

    I would be very interested to see if most of the posters that are against your point are either parents or religious.
    My answers would be yes/no
    Confirmation bias for the win!

  9. When a person really does not want to have kids (on purpose) I am curious to know why. More material stuff for oneself? Fear of the future? What is the true reason for wanting to be alone in the world? Regardless of what you think you can replace family or children with one really never will know till one stands over the proverbial edge of the Grand Canyon and surveys it.

    More often than not having children is a lot different and much better than you think it is.

    I am continuously tickled with my own children at how they enjoy their lives, work hard and generally just make us so pleased that they were born. I've said it more than once here, I wish I had three more just like em. :)

    Try not to second guess your future or anyone else's.

  10. I play the same trick on my kids when I eat the last slice of pizza. I say: "Well we have some ice cream. You like ice cream too right?"

    We eat the ice cream and I say to the kids: "See, you actually just wanted something nice to eat and ice cream is nice."

    Of course, I had the extra slice of pizza and the ice cream...

  11. You're friend doesn't sound like she's thought enough about it, frankly.

    Our culture breeds (no pun intended) the supposed need to have children. Its part of 'the plan'.

    Its neither essential or incidental - its (become) mechanical.

    I can't imagine why one would want to have kids, especially in the way one wants to have, say, a career.

  12. To me what this shows is that "Idyllic afternoon in the company of people you love" can be experienced with both family/children and close friends....what it does not show is the qualitative similarity between the two for a given observer.
    The lady is naive to the "children" experience, so she is in no position to draw a qualitative comparison for herself from her observed/imaginary experience.

    There is a potential of inferential fallacy here for some readers...perhaps. The hypo does lead to the literal inference drawn, however, there is a potential for the inference to be misunderstood by a not so careful reader.



Note: Only a member of this blog may post a comment.