The Moral Life of Babies
May 10, 2010 9:58 AM   Subscribe

"A growing body of evidence suggests that humans have a rudimentary moral sense from the very start of life... Some sense of good and evil seems to be bred in the bone... [But] the sense of right and wrong that [babies] naturally possess diverges in important ways from what we adults would want it to be."
posted by AceRock (84 comments total) 16 users marked this as a favorite
 
I read this article when it was linked to in the Babies Babies Babies Babies thread. There are some really provocative insights in there but the "there must be some influence from God because evolution clearly cannot explain why we would be nice to each other" moral they tacked on the end is nearly the definition of specious.
posted by idiopath at 10:01 AM on May 10, 2010 [6 favorites]


It's funny that even the NY Times has anointed The Onion as the de-facto standard of satire. Has Die Zweibel really become that much of a standardized cultural signpost?

Anyway, the actual content of the article was interesting if unsurprising (to me at least). Babies have about the same level of morality as moneys, which is to say they have a lot, but it's a little vicious and not exactly thought out for the long-term in all cases.
posted by GuyZero at 10:05 AM on May 10, 2010


Actually they point out a way that babies differ from monkeys that I found particularly interesting. If a monkey gets unfairly treated the other monkeys will ignore or mistreat the downtrodden monkey, while babies will show sympathy to someone treated unfairly.
posted by idiopath at 10:07 AM on May 10, 2010 [1 favorite]


"there must be some influence from God because evolution clearly cannot explain why we would be nice to each other"

I didn't get that at all - it seemed more like they were trying to say "babies have a limited moral foundation but we don't know where that comes from." Just because it is still unexplained doesn't mean that there is an automatic jump to divine intervention.
posted by infinitefloatingbrains at 10:11 AM on May 10, 2010 [1 favorite]


I have been accused (and stand guilty as charged) of having a fairly pessimistic view of our species. But one fact tempers the ugliness: when you meet a baby for the first time, and she is sitting in her high chair, enjoying her Cheerios, she will almost always (with very few exceptions in my experience) offer you, a complete stranger, some of her food. "Hey, these are good! You hungry?" That's some pretty sweet morality right there.
posted by barrett caulk at 10:12 AM on May 10, 2010 [76 favorites]


Why don't they think that there is any social conditioning by year one?
posted by cmoj at 10:16 AM on May 10, 2010 [10 favorites]


hmmm, like all these cognitive psych stories, the experimental flaws stick out in the first few moments of reading:

It’s a challenge to study the cognitive abilities of any creature that lacks language . . . when babies see something that they find interesting or surprising, they tend to look at it longer than they would at something they find uninteresting or expected. And when given a choice between two things to look at, babies usually opt to look at the more pleasing thing. You can use “looking time,” then, as a rough but reliable proxy for what captures babies’ attention: what babies are surprised by or what babies like.

needless to say, the number of assmptions that lead to the idea that somehow you can measure what people think based on some external thing is very common in these cognitive psych studies.

Since we cannot have reliable 2-way communication with these children (1) we do not know if a baby finds something interesting, uninteresting, or expected; (2) we cannot prove the theory that "And when given a choice between two things to look at, babies usually opt to look at the more pleasing thing"

This is pseudo-science, pure and simple. If you cannot reliable measure something, then no amount of application of the scientific method will create scientific results.
posted by Ironmouth at 10:20 AM on May 10, 2010 [9 favorites]


. . .the "there must be some influence from God because evolution clearly cannot explain why we would be nice to each other" moral they tacked on the end is nearly the definition of specious.

Clearly, you didn't read the whole article. From the conclusion:
The notion at the core of any mature morality is that of impartiality. If you are asked to justify your actions, and you say, “Because I wanted to,” this is just an expression of selfish desire. But explanations like “It was my turn” or “It’s my fair share” are potentially moral, because they imply that anyone else in the same situation could have done the same. This is the sort of argument that could be convincing to a neutral observer and is at the foundation of standards of justice and law. The philosopher Peter Singer has pointed out that this notion of impartiality can be found in religious and philosophical systems of morality, from the golden rule in Christianity to the teachings of Confucius to the political philosopher John Rawls’s landmark theory of justice. This is an insight that emerges within communities of intelligent, deliberating and negotiating beings, and it can override our parochial impulses.
If anything, the article rejects your interpretation entirely. Bagging on religion seems to be de regeur around here, but seriously, RTFA.
posted by valkyryn at 10:21 AM on May 10, 2010 [6 favorites]


infinitefloatingbrains: "I didn't get that at all - it seemed more like they were trying to say "babies have a limited moral foundation but we don't know where that comes from.""


A version of this article appeared in print on May 9, 2010, on page MM44 of the Sunday Magazine.

If I am remembering correctly, the other version I read ended with the Dinesh D’Souza citation, and did not contrast it with Dawkins' view. But I could have misread or misremembered.
posted by idiopath at 10:21 AM on May 10, 2010


An inherent moral sense is about as difficult to predict or explain evolutionarily as an inherent ability to see colors.
posted by DU at 10:24 AM on May 10, 2010 [1 favorite]


… she will almost always (with very few exceptions in my experience) offer you, a complete stranger, some of her food. "Hey, these are good! You hungry?" That's some pretty sweet morality right there.

It's a lot less cute when it's an adult chucking cereal at your head.
posted by wreckingball at 10:27 AM on May 10, 2010


Ok, the other version had a similar conclusion. Sadly I probably just closed the window when the author left the subject of the study itself and moved on to the "big conclusions" that seem de rigour for pop-sci so I missed his actual conclusion.
posted by idiopath at 10:29 AM on May 10, 2010


Ironmouth: "This is pseudo-science, pure and simple. If you cannot reliable measure something, then no amount of application of the scientific method will create scientific results."

Self describing speech and memory are actions too, and not necessarily much more reliable than following someone's gaze.
posted by idiopath at 10:31 AM on May 10, 2010 [3 favorites]


The author and researcher, Paul Bloom, also teaches the Introduction to Psychology course at Open Yale. His lectures have many more fascinating examples of his cognitive work with babies.
posted by rocket88 at 10:33 AM on May 10, 2010 [3 favorites]


Ironmouth: "This is pseudo-science, pure and simple. If you cannot reliable measure something, then no amount of application of the scientific method will create scientific results."

Self describing speech and memory are actions too, and not necessarily much more reliable than following someone's gaze.


Hence why any cognitive psych is pseudo science, including attractiveness studies based on showing volunteers photos.

But just look at the incredible biases the writer lays out:

Moral ideas seem to involve much more than mere compassion. Morality, for instance, is closely related to notions of praise and blame: we want to reward what we see as good and punish what we see as bad. Morality is also closely connected to the ideal of impartiality — if it’s immoral for you to do something to me, then, all else being equal, it is immoral for me to do the same thing to you. In addition, moral principles are different from other types of rules or laws: they cannot, for instance, be overruled solely by virtue of authority. (Even a 4-year-old knows not only that unprovoked hitting is wrong but also that it would continue to be wrong even if a teacher said that it was O.K.) And we tend to associate morality with the possibility of free and rational choice; people choose to do good or evil. To hold someone responsible for an act means that we believe that he could have chosen to act otherwise.

All of this is pre-experimental bias about what "morality" is or is not. Any time you have a squishy word or concept like "morality" the fact that a word cannot be defined like weight, velocity, mass, chemical composition or the like immediately destroys any chance for actual "scientific" work to be done on it. This is not to say we can't learn from studying it, but we should not attach the word "science" to such studies. Until we have the brain's code cracked and can read the individual messages going from neuron to neuron, there cannot be true "science" in this area.
posted by Ironmouth at 10:39 AM on May 10, 2010


I, for one, welcome our teeny tiny cutey-pie moral authority overlords. With Cheerios!
posted by Mister_A at 10:44 AM on May 10, 2010 [4 favorites]


. Any time you have a squishy word or concept like "morality" the fact that a word cannot be defined like weight, velocity, mass, chemical composition or the like immediately destroys any chance for actual "scientific" work to be done on it.

As an example to the rest of us, could you define "mass"? Please keep in mind that using the word "matter" will be considered circular and "that which falls out of m = F/a" will be considered isomorphic to an operational definition of "morality".
posted by DU at 10:45 AM on May 10, 2010 [8 favorites]


It sounds not unlike Jung's theory of the collective unconscious.
posted by hermitosis at 10:48 AM on May 10, 2010


Ironmouth, your personal definition of science seems to limit the use of the label to physics only. There are many phenomena that we can observe, but can't measure. Just because we haven't defined an SI unit of morality that we can measure to three significant digits doesn't mean that scientists can't observe it and draw conclusions from it.
If you disagree with those conclusions, that's fine, and I'm sure Dr. Bloom and the other Psychology Ph.Ds at Yale will be interested in reading your next paper.
posted by rocket88 at 10:53 AM on May 10, 2010 [2 favorites]


Ironmouth, I think you are conflating the author's pseudo-science grasping at theories that explain the resuls of the basic science.
There was hard science done here. Babies were shown specific actions (good and bad behavior) and their later reaction to the entities that performed that behavior. What was found was that babies between 6 months and 1 year had a strong predilection towards rewarding nice behavior and punishing bad behavior. How is that not science? In the article the author even mentions that this behavior only strictly applies to alert, non-fussy babies, as sleepy/fussy babies weren't considered for the experiments. So with certainty you can say that alert non-fussy babies rewarded nice behavior and punished mean behavior a very high percentage of the time (the article doesn't state the actual percentage, but did indicate it was substantial).

They even took the model further to observe second order affects. Do babies did have an innate sense of justice (I only use that word for brevity) that extends to social interactions of the nice and mean? If so, then the babies should like those that help the nice, and not like those that are mean to the nice (the latter is really not new data, though). Likewise, they should not like those that help the mean, and like those that are mean to the mean. And the observed data fit that model. Again, I'm not seeing where the science was missing.

I completely agree that the broader implications that the author tried to draw out in the article had nothing to do with Science. But just because Stephen Hawking doesn't think we should seek out aliens, doesn't invalidate Astrophysics as a science.
posted by forforf at 10:56 AM on May 10, 2010


Ironmouth: from what I understand of the article, they measure a reaction (length of babies' attention, the babies' preferences) to morality plays of increasing complexity. Their results seem to show that that measurement is correlated to "morally positive" outputs and characters. We can discuss the signification of the measurement itself (for all we know, the babies are salivating and planning to eat the good guys, hence the longer attention), but the correlation remains. Even if does not prove that the babies "approve" of the positive output, it still shows that they can tell the positive from the negative or at least read a signal in the morality plays. Given that babies are able to vocally express moral preferences just a few months later, perhaps it's not such a large jump to think of attention span as an early indicator.
posted by elgilito at 10:58 AM on May 10, 2010


Is there anything that looks at whether this "baby morality" goes away once they start talking and start watching Disney movies and need constant clarification on the morals of every single character and action? And when they need help learning how to share?
posted by amethysts at 11:07 AM on May 10, 2010


This is a worthwhile article, but a lack of understanding about the philosophical issues involved hamstrings the author's considerations. I don't think Paul Bloom is aware of the battleground concepts involved when people wonder if morality is innate or learned. That is: the idea that babies are born with a capacity for moral thought is really the same as the idea that morality is learned, at least on the scale that it's been argued about in the past. None of the studies Bloom mentions even come close to demonstrating some Kantian categorical imperative inherently inborn in the minds of babies, or some specific immanent moral code that all human beings are born knowing.

The idea of some innate morality stems from a difficulty which people face when they want to argue for an absolute and specific moral code that prescribes in detail how human beings should live, a code against which we can praise or blame all of humanity. If we can really blame people for not following this specific code, then we must remove the excuse given them by their potential ignorance of the law; that is, if people are really at fault for breaking the moral code, this moral code must be so obvious to all human beings at all times that any infraction comes about as a result of a choice to do evil. Strict moralists would like to say: "this person has done something wrong, and she has no excuse, for all of us know that what she has done is wrong." If the moral code isn't something innate, then it becomes more difficult to say that, at least in a universal sense.

It's worth mentioning that deciding that there isn't an innate moral code – that is, agreeing that morality is learned – doesn't mean that universal morality is right out. It only means that praise and blame become more difficult and questionable, because morality is thus an obscure thing which requires some wisdom. Socrates showed one direction this line of reasoning might take when he argued that no one does evil willingly – that all human beings who do evil things do so under the mistaken impression that those evil things are actually good in some way.
posted by koeselitz at 11:14 AM on May 10, 2010 [1 favorite]


Mencius was the Confucian famous for arguing about an inherent moral nature that tends to the good, with the locus classicus being 2A6, related to his particular take on ren that has our full humanity in our social selves.
posted by Abiezer at 11:17 AM on May 10, 2010 [1 favorite]


amethysts: "And when they need help learning how to share?"

Even adults don't do such a good job of introspection. Many of us understand morality quite well but only as long as we are not one of the parties involved.
posted by idiopath at 11:19 AM on May 10, 2010 [1 favorite]


It's a lot less cute when it's an adult chucking cereal at your head.

Depends on the cereal, and the distance between you and the adult.
posted by davejay at 11:24 AM on May 10, 2010


We don't know the mechanism by which this biological predisposition functions, but we know very well why it evolved : kin selection.
posted by jeffburdges at 11:25 AM on May 10, 2010


In the Moral Roots of Liberals and Conservatives (TED talk), Jonathan Haidt argues that humans are born with a sort of moral blueprint, or fundamental structures, on which our morality are based. In particular, he lists 5: care, fairness, loyalty, respect purity. I would love to see research attempting to test for these in babies and young children.
posted by christonabike at 11:30 AM on May 10, 2010


me: “I don't think Paul Bloom is aware of the battleground concepts involved when people wonder if morality is innate or learned.”

The World Famous: “I'm going to go ahead and guess that he is aware of those concepts, but that he just decided not to write about them in that article. Though perhaps I'm giving him too much credit, based on the fact that he's a Psychology professor at Yale, and I assume that anyone who does what's necessary to reach that position probably knows about that sort of basic stuff.”

Well... I don't know. I'm not trying to drag him down or anything – I think it's a great article, very interesting, and I got a lot out of it. It's only that when he says that "some sense of good and evil seems to be bred in the bone," he certainly doesn't mean that what some people, Kant for example, would like to believe about an innate moral sense is true, And from that, I assumed (maybe unfairly) that he thought he was talking about what people often mean by "innate moral sense," whereas actually his account seems to agree with those of us who believe that morality is a learned thing.

I just don't think this is "basic stuff;" at least it's complicated and hard to follow for me, anyway. But all of this might be completely external to the article, which was in itself very interesting and thoughtful.
posted by koeselitz at 11:34 AM on May 10, 2010


cmoj: “Why don't they think that there is any social conditioning by year one?”

That seems like it's the real difficulty with trying to do focused studies in this area. To really determine whether babies have some innate moral sense, and to test the extent of that sense, you'd have to isolate a set of babies entirely for a couple of years, making sure they had no social conditioning whatsoever, and then test to see if they seemed to have that moral sense.

And that's not likely to happen any time soon. I hope.
posted by koeselitz at 11:38 AM on May 10, 2010 [1 favorite]


If Paul Bloom is who I remember him to be, the guy is pretty much at the top of his field. I think it's awfully presumptuous to believe that he doesn't know what he's talking about. Especially when those claiming he's full of fluff aren't even trained or working in his field.
posted by five fresh fish at 11:38 AM on May 10, 2010


He's not full of fluff. If anybody says he is, they're wrong.
posted by koeselitz at 11:39 AM on May 10, 2010


Ironmouth: Any time you have a squishy word or concept like "morality" the fact that a word cannot be defined like weight, velocity, mass, chemical composition or the like immediately destroys any chance for actual "scientific" work to be done on it.

Of course, you are quoting from an article written for the NYT. Behaviors like visual fixation on a stimulus are certainly measurable, with a reliability easily on par with detecting the decay rate of nutrinos. Furthermore, arguing that we can't investigate behavior without understanding the code of neurons begs the question if how the heck are we going to decrypt that code without depending on correlating behavior to neural activity.

koeselitz: That is: the idea that babies are born with a capacity for moral thought is really the same as the idea that morality is learned, at least on the scale that it's been argued about in the past. None of the studies Bloom mentions even come close to demonstrating some Kantian categorical imperative inherently inborn in the minds of babies, or some specific immanent moral code that all human beings are born knowing.

Well, this strikes me as a bit of a straw man, at least as applied to the field of developmental psychology. They don't argue that infants are born with an adult grasp of the nuances of language and the ability to write epic poetry either. What they do argue is that some of the building blocks of language learning appear to develop in the first few months without anything that we recognize as formal instruction.

Of course, showing that infants might (pending replication of course) understand theft and punishment at a surprisingly young age is a meagre basis for a universal moral code, and I don't think anyone is arguing otherwise. I sincerely doubt that Bloom would contradict to the evidence that we should ignore all the other evidence that critical forms of moral reasoning don't fully develop until adulthood. All this does is push the time frame for the early development of moral reasoning back a few years.
posted by KirkJobSluder at 11:47 AM on May 10, 2010


Look, I'll reframe what I'm saying, since it seems to have set off people who are in a rush to 'defend' Paul Bloom against criticism, real or imagined:

There are a lot of people who might read the opening of Bloom's article and say: "ah ha! I knew it! This proves what I've said all along – that the absolute universal moral code is innate, and that when people violate it, we can universally blame them for violating a code that is obvious to all!" There are other people who might read this and say: "this isn't fair! He's trying to say that some psychological studies prop up the religious argument that there is some absolutely universal moral code that is innate in all of our minds, but it's preposterous to argue that way!" (idiopath above had that sense of this article when he first saw it, although he seems to have felt differently on reading it again.)

I am only saying: the studies Paul Bloom is talking about don't fit that debate in the way those people would like them to, and the conclusion that he's trying to prop up absolute morality is a false one. And what's more in the face of those debates I have a feeling he himself would laugh and say that those people had somewhat missed the point.

When I say that maybe he's "not aware of the battleground concepts," I'm saying (badly, and in far too strident language) that some people will likely read this article and take from it conclusions that Paul Bloom certainly doesn't intend. That's not Paul Bloom's fault, of course. In fact, that might be what he means to indicate when he mentions how much the morality observed in babies is different from "what we adults want [morality] to be." In which case it would be silly and moreover quite wrong to say that he "isn't aware" of anything here.

I don't think having a Ph D or a teaching position at Yale mean much at all. I've met people with both who were clearly idiots. Paul Bloom is clearly not an idiot, however, and I'm sorry if I've made it sound like I thought he was. He probably deserves the benefit of the doubt in this case, so I hope I can amend my earlier statement to say that I only think people might be jumping to unfortunate conclusions if they assume that this article will allow them to make grand pronouncements about absolute and universal moral codes.
posted by koeselitz at 11:55 AM on May 10, 2010


An inherent moral sense is about as difficult to predict or explain evolutionarily as an inherent ability to see colors.

This isn't even vaguely true. There's plenty of literature on the evolution of human vision, the ability to see colors confers great advantages over monochromatic vision, and there's also plenty of literature on the evolutionary advantages of altruism and cooperation (two things which are generally viewed as tied up with morality), going all the way back to Peter Kropotkin's Mutual Aid as a Factor of Evolution.
posted by Pope Guilty at 11:57 AM on May 10, 2010


koeselitz: ""this isn't fair! He's trying to say that some psychological studies prop up the religious argument that there is some absolutely universal moral code that is innate in all of our minds, but it's preposterous to argue that way!" (idiopath above had that sense of this article when he first saw it, although he seems to have felt differently on reading it again.)"

My initial misreading (I think) was thinking I saw the "evolution won't provide us with morality and that is why we need religion" argument toward the end - coming more from the Catholic original sin side of things than the Protestant innate religious sense side.

Pope Guilty: "This isn't even vaguely true. There's plenty of literature on the evolution of human vision, the ability to see colors confers great advantages over monochromatic vision, and there's also plenty of literature on the evolutionary advantages of altruism and cooperation (two things which are generally viewed as tied up with morality)"

I read his "just as difficult" as being a synonym to "not especially difficult, similar in complexity to other standard questions in the field".
posted by idiopath at 12:04 PM on May 10, 2010


What I see out of it is that babies are inclined to "help" the achievement of simple goals (remember the dropped chalk experiment?) and inclined to perceive those entities helping those who are struggling as being preferred versus those who are hindering. This is not morality.

The red ball could be trying to push a vat of poison up into the city reservoir. Babies don't know any better. That isn't morality — that's an inborn impulse to infer motion towards a goal and to view that as preferable. It has an evolutionary advantage, clearly. Toddlers would help other toddlers out of a mud pit. It loosely maps to morality, until the hindering entity is trying to stop someone from stealing a TV (in so far as you regard private property as being "good").

Voice of God? In that case, God must have a great phone line to bees and other social insects.
posted by adipocere at 12:11 PM on May 10, 2010 [4 favorites]


I'll be the first to admit that due to my lifelong efforts to avoid babies, I'm not the best judge. But my limited experience has led me to believe that babies have about the same moral compass as a cat;

Food. Sleep. Attention. All of these things need to be available on command and at a moments notice. Anything to the contrary is bad and needs to be vociferously denounced.
posted by quin at 12:19 PM on May 10, 2010


Also, for clarification's sake, this is what I think the author is saying regarding religion, morality, culture, and evolution:

He makes the claim that if babies had a fully developed innate morality, that that would be evidence of a loving God as intelligent designer. He goes on to claim that since the type of moral sense that babies possess is rudimentary, that it therefore can be explained as an accidental byproduct of an evolved adaptation. Finally he claims that our more advanced sense of right and wrong is cultural rather than divine or biological in origin.

I would object that an innate moral sense would not be proof of a loving God, and the absence of that innate moral sense is not incompatible with mainstream theology (ie. Catholic original sin). The only difference between the Catholic interpretation of morality and his own is who is presumed to have taught it to humans first (evolution of ideas in culture or divine introduction via prophets and guidance of philosophers by the Holy Spirit).
posted by idiopath at 12:22 PM on May 10, 2010


I should have put "accidental byproduct of an evolved adaptation" in quotes, because they are the authors words and as I understand it imply a fundamental misunderstanding of natural selection.
posted by idiopath at 12:24 PM on May 10, 2010


Metafilter: The red ball could be trying to push a vat of poison up into the city reservoir. Babies don't know any better.
posted by Skot at 12:27 PM on May 10, 2010 [1 favorite]


What is the take-home from this in re fighting 100 babies at once?
posted by everichon at 12:34 PM on May 10, 2010


koeselitz: Ok, I misunderstood you.

idiopath: Yes, I do agree that Bloom is out of his depth when he uses flaws in infant morality to critique the TAG.
posted by KirkJobSluder at 12:35 PM on May 10, 2010


I'm not sure I follow you. How is Toddler A, who helps another toddler out of a mud pit, more likely to reproduce than Toddler B, who does not?

A breeding population of toddlers (uh... or something) that displayed that characteristic would be more likely to survive (as a group) and pass that trait on.
posted by brundlefly at 12:52 PM on May 10, 2010 [1 favorite]


KirkJobSluder: "I do agree that Bloom is out of his depth when he uses flaws in infant morality to critique the TAG."

For sake of clarity, what do you mean by "the TAG"? That particular TLA is ungooglable.
posted by idiopath at 1:20 PM on May 10, 2010


I wouldn't bother engaging Ironmouth on issues of cognitive psychology. He persists in misunderstanding both cognitive psychological research methods and the philosophy of science, while accusing researchers of practicing pseudoscience. See extensive discussion in a previous thread.
posted by Philosopher Dirtbike at 1:21 PM on May 10, 2010


The World Famous: "More likely than what?"

Another group, relatively isolated from the first, lacking that trait. Differences between groups can be selected for to the degree that the groups remain separate.
posted by idiopath at 1:25 PM on May 10, 2010 [1 favorite]


idiopath: The Transcendental Argument for God or at least one variation of it. Basically, it goes that if there is a universal sense of morality/logic, there must be a universal morality/logic, and therefore, there must be a universal something defining that morality/logic. Call it God. I'm probably mangling it to hell because I never found it to be be very compelling. But if you are going to take it seriously, it deserves a bit more of a response than what Bloom provides here.
posted by KirkJobSluder at 1:32 PM on May 10, 2010 [1 favorite]


Self describing speech and memory are actions too, and not necessarily much more reliable than following someone's gaze.

Yes. Exactly. And you'll find a host of similarly-specious "research" in sociology and psychology that assumes you can rely on people's responses to surveys and questionnaires and interviews to accurately tell you what they're thinking or what they've been doing.
posted by straight at 1:35 PM on May 10, 2010 [1 favorite]


I have some qualms with this study, for certain. Like cmoj suggested, clearly there has already been a large amount of social conditioning taking place before the babies possess the motor control to even complete the task. To assume that this behavior is innate is somewhat spurious.

needless to say, the number of assmptions that lead to the idea that somehow you can measure what people think based on some external thing is very common in these cognitive psych studies.

The assumption that we can measure what someone thinks based on "some external thing" is not one that we experimental psychologists generally make. Introspectionism was essentially supplanted by behaviorism, a long ass time ago. Conceiving of the mind as a black box, however, is equally problematic. What is measured in the study is preferential attention. Yes, I think it is fair to make the assumption that direction of gaze toward a particular stimulus and maintenance of that gaze is a legitimate measure of attention. Tell me what else attention is if not the allocation of cognitive resources in order to engage in more focused processing.

Looking time as an index of attention in work with infants is well-established and can shed a great deal of light on the way that we come to understand the world as our brains and bodies grow and age.

Since we cannot have reliable 2-way communication with these children (1) we do not know if a baby finds something interesting, uninteresting, or expected; (2) we cannot prove the theory that "And when given a choice between two things to look at, babies usually opt to look at the more pleasing thing"

If you want to indite behavioral science in general, be my guest, but I do not think it even remotely reasonable to suggest that because we cannot ask a child what she's thinking that we cannot use inferential methods to determine stimuli toward which resources are preferentially directed. Self-report carries its own host of complications. You can ask people why they made a decision all day long, and get plenty of nice stories, but the fact of the matter is, subjective introspection cannot get at the underlying cognitive mechanisms of a given process in an objective, measurable way.

Is the rich history of work with animal models equally invalid because we cannot ask the animal what it thinks? Studies of animal decision making and cognition are abundant, and it has been a long time since the notion that only stimulus and response are worth considering.

What's more, infant attention is moderated by a host of variables. Preferential looking may be driven by stimuli that are "more pleasing," or "more familiar," or "more novel," or whatever. It doesn't matter what subjective state is assigned to the behavior. It is important to speculate about the underlying state that drives preferential looking in order to generate testable hypotheses about it, but the only strong conclusion rigid cognitive psychologists draw about preferential looking for a given stimulus is that more time is spent, you know, looking at it (or "visually attending to it").

That study has some flaws, for certain. I am not entirely comfortable with some of the logical leaps the authors make. What's more, a "universal moral grammar" is every bit as ridiculous as universal grammar in the linguistic sense. It still has its proponents, but they are clinging to a sinking ship, in my opinion. However, casting aspersions on "these cognitive psych studies," lumping decades of work in a massive field together, and calling into question the validity of non-self report data is a bit of a logical leap as well. The implication that we can learn more about how humans process information by asking them then by watching them has been demonstrated time and time again to be empirically incorrect.
posted by solipsophistocracy at 1:56 PM on May 10, 2010 [5 favorites]


And you'll find a host of similarly-specious "research" in sociology and psychology that assumes you can rely on people's responses to surveys and questionnaires and interviews to accurately tell you what they're thinking or what they've been doing.

No kidding. Self-report measures provide another data point, for sure. To me though, it don't mean a thing if it ain't got that brain.
posted by solipsophistocracy at 1:57 PM on May 10, 2010


I wouldn't bother engaging Ironmouth on issues of cognitive psychology. He persists in misunderstanding both cognitive psychological research methods and the philosophy of science, while accusing researchers of practicing pseudoscience. See extensive discussion in a previous thread.

oh. whoops. Guess maybe that was a waste of typing. Yeah, upon alternate-thread re-not-preview, you can just go ahead and assume that I am one of those people who does not practice science as you see it. I'd be interested to know, though, what it is you do think constitutes "science," and what your basis for that judgment is.

posted by solipsophistocracy at 2:00 PM on May 10, 2010


The World Famous: Well, that's the problem in dealing with the evolutionary biology of behavior. My personal opinion is that without either some hard numbers on the quantitative contributions of genetics on the variance within populations or more closely-related species within our local clade, that we really can't say much of anything about the kinds of selective pressures for/against a specific behavior.

And would it be specifically selecting for moral reasoning in infants, or is moral reasoning in infants a necessary part of boot-strapping the cognitive process in adults? There is a strong argument to be made that people who don't get the necessary cognitive and neurological development in the first five years face substantial difficulties trying to learn those skills in adolescence and adulthood.

solipsophistocracy: That study has some flaws, for certain.

I feel the need to interject that we've not seen the study. What we've seen is a NYT Magazine opinion piece by a secondary author of the study. Anyone making conclusions about what the actual study does or does not say is talking out of their ass.
posted by KirkJobSluder at 2:23 PM on May 10, 2010 [1 favorite]


That was a bit too harsh. But it's really annoying when people go off on the methods or conclusions of a study based on a popular-press write-up.
posted by KirkJobSluder at 2:42 PM on May 10, 2010 [3 favorites]



Actually they point out a way that babies differ from monkeys that I found particularly interesting. If a monkey gets unfairly treated the other monkeys will ignore or mistreat the downtrodden monkey, while babies will show sympathy to someone treated unfairly.
posted by idiopath


idiopath, you fail to mention that de Wall also pointed out that chimpanzees “will approach a victim of attack, put an arm around her and gently pat her back or groom her.”
posted by chance at 3:39 PM on May 10, 2010


yeah, but chimps are not monkeys
posted by idiopath at 3:45 PM on May 10, 2010 [1 favorite]


Anyone making conclusions about what the actual study does or does not say is talking out of their ass.

Agreed. I am in a field quite closely related to this one, and have, in fact, read (what I am fairly certain) is the "revise and resubmit" draft of the study in question, or at least a study in the same line of research. It's not really on the level to disseminate papers that aren't yet "in press," and since the version I saw got a "revise and resubmit," perhaps they have fixed some of the stuff I have beef with, although some of my issues are on a purely conceptual level.

Please accept my apologies for just up and offering my opinion like that. When I saw that this NYT article (which has received no small amount of attention/discussion at my dept.) had been posted, I just assumed that the paper had been accepted and was likely already available online. My bad.

But it's really annoying when people go off on the methods or conclusions of a study based on a popular-press write-up.

Thanks for the retraction. I totally agree.

If you're interested, you can learn more about this line of research:

Bloom's website

Wynn's website (mentioned in the article)

Kiley Hamlin's bio contains some links of interest on the Yale InfantLab's "People" page of their website, too.
posted by solipsophistocracy at 5:03 PM on May 10, 2010


Evolution gradually gave us the rational capacity to control our environment, rather than simply survive it by the odds. Having empathy is proof of our control. An evolved being cares deeply about stability because it allowed us to evolve.
posted by Brian B. at 5:11 PM on May 10, 2010


There are a lot of people who might read the opening of Bloom's article and say: "ah ha! I knew it! …I am only saying: the studies Paul Bloom is talking about don't fit that debate in the way those people would like them to…

So you're arguing about people that might say this thing that you disagree with, and who are not here posting to MeFi. Call me crazy, but I believe that's called "strawmanning."
posted by five fresh fish at 6:08 PM on May 10, 2010


I immediately wondered if… the baby was projecting into the central puppet (who gave the ball away) their own selfish want to have the ball.

In that case, the baby would project into the final puppet, which ultimately got to possess the ball with finality.
posted by five fresh fish at 6:15 PM on May 10, 2010


What was found was that babies between 6 months and 1 year had a strong predilection towards rewarding nice behavior and punishing bad behavior.

What's the scientific definiton of nice
posted by Ironmouth at 10:30 PM on May 10, 2010


People who work with kids a lot and/or write articles for the New York Times so that lay folks can get a sense of what's going on in the laboratory have a tendency to modulate their language based on their audience. I am most upset that this study isn't available for your erudite analysis, because I somehow doubt that a dude on the internet's interpretation of the press release version of what was actually measured in the study is what the study actually purports to measure.

If you really have beef, why don't you check out some of the studies in this line of research? They're readily available on the websites posted a few comments up.
posted by solipsophistocracy at 10:43 PM on May 10, 2010


I immediately wondered if… the baby was projecting into the central puppet (who gave the ball away) their own selfish want to have the ball.

In that case, the baby would project into the final puppet, which ultimately got to possess the ball with finality.


I wonder whether projection is not the basis for morality. Baby, for its whole life, has been the one to whom things have been done. Both good things but also in the majority (from a baby perspective) unfair, mean and unpleasant things! I wouldn't be surprised if that wouldn't cause baby to identify with the one from whom the ball was unfairly stolen, and if that were not the seed from which a system of morality was born.
posted by Omnomnom at 2:31 AM on May 11, 2010


Something they didn't elaborate on, but I found interesting: for the experiments involving coloured shapes with faces, the babies react differently if the shapes don't have faces - they don't view it as a social interaction. But this means that six-month old babies can tell that a cartoon square/triangle/etc with a face is supposed to represent a person. I thought that was fascinating. I wonder at what age babies can understand representations of non-human faces? Or does that ability just come at the same time as recognising human faces?
posted by badmoonrising at 5:22 AM on May 11, 2010


What's the scientific definiton of nice

Here are some reasonable definitions of various terms that would enable one to categorize behavior. Note: I bet in the study the categorization and definitions were much more rigorous.

Nice: Work with, assist, help.
Mean: Obstruct, hinder, prevent

Justice: Reward entities that exhibit nice behavior, Punish entities that exhibit mean behavior
Reward: Provide something desirable or withhold something undesirable
Punish: Impose something undesirable, or withhold something desirable
Desirable: Something enjoyed or wanted by the entity
Undesirable: Something not enjoyed or not wanted by the entity

Just as the classification of living things into a taxonomy preceded (and provided the hints) for evolution theory, so might classification of behaviors provide clues that could lead to a broader behavioral theories. Sniping at classification methodologies as non-scientific is mystifying to me.
posted by forforf at 6:23 AM on May 11, 2010


I guess I do think of chemistry, physics and astronomy as being the only real "science." And the reason I think it is important to think that way is that people not really thinking about stuff will take the definitions of what is "moral" behavior as having been scientifically discovered and supported in the exact same way our understanding of the escape velocity of the earth is supported by evidence and experiment. And its not.

So much of this stuff is clearly pseudo-science. Its people taking their own biases and pressing them on to double-blind studies and doing all sorts of shady things. Its exactly why "creation science" exists.

As for evolution, well, frankly, it is two parts. The first is somewhat unprovable--that current forms are the descendants of unlike forms which changed over time in response to the environment. The second is the well-documented change in organisms' genetic code over generations via reaction to environmental conditions, such as what happens when you don't take the whole bottle of penicillin. The first is frankly, in some sense, unprovable without a time machine. It is an educated supposition based on the sound science of the second observations. However, aliens planting all of the evidence and landing us here would produce the same result. Evolution is merely the most parsiminous explanation of biodiversity. It isn't the same as say, astrophysics, which, if the math is correct, uses current conditions to find the only possible past universe which could have produced those conditions.

I think it is very, very important to break down the differences between what it is we know and what it is we suppose and not to treat those things as the same. Because when some "science" gets unproven, people tend to think the rest of it is bunk, which is not true.

Applying this to the case here, we see a huge leap. The author tries to dance around the fact that he is, indeed, trying to tell us what he cannot--that he can read babies' minds:

This experiment was designed to explore babies’ expectations about social interactions, not their moral capacities per se. But if you look at the movies, it’s clear that, at least to adult eyes, there is some latent moral content to the situation: the triangle is kind of a jerk; the square is a sweetheart. So we set out to investigate whether babies make the same judgments about the characters that adults do. Forget about how babies expect the ball to act toward the other characters; what do babies themselves think about the square and the triangle? Do they prefer the good guy and dislike the bad guy?

They keep wanting to force their own adult construction of "morality" on to the babies, at the very moment they continue to deny that it is actually possible.

We'll know what's going on in babies' heads the minute we can read their minds via the signals that their neurons send to one another. Until then, all we have is this flawed, hopeful, made-up shit.
posted by Ironmouth at 9:07 AM on May 11, 2010


People who work with kids a lot and/or write articles for the New York Times so that lay folks can get a sense of what's going on in the laboratory have a tendency to modulate their language based on their audience. I am most upset that this study isn't available for your erudite analysis, because I somehow doubt that a dude on the internet's interpretation of the press release version of what was actually measured in the study is what the study actually purports to measure.

If you really have beef, why don't you check out some of the studies in this line of research? They're readily available on the websites posted a few comments up.


Here's the problem--my beef isn't with the studies. My beef is with the article. It suggests a certainty that is not there. It suggests you can read babies' minds by measuring how long they look at things. These are unprovable assumptions. And they are not labelled as such in this article. The fact that the author is a scientist does not excuse him. More importantly, I do think it does open a window into the untested assumptions discussed. For more of the same kind of junk, read Daniel Gilbert's Stumbling on Happiness. Its the same pseudo-science packed into a NYT best seller. And he's a Harvard psychologist. The core of his approach is that self-reporting is accurate and can therefore constitute the basis for scientific measurement. Bollocks.
posted by Ironmouth at 9:13 AM on May 11, 2010


Ironmouth: I guess I do think of chemistry, physics and astronomy as being the only real "science."

Except that all sciences use inferential conclusions based on data. My sister is a chemist for a company that uses data on a spectrometer to make inferences about athletic and therapeutic drug compliance. Has she actually observed people not in compliance? No. That's an inference made by employers and law enforcement based on data from metabolites.

Have physicists ever seen a top quark? No. They've made an inference that a top quark was created based on data about the energies of detected decay products.

Do astronomers know that Eta Carinae is due for a supernova in the next few million years? No. It's an inferential claim made from millions of data points and mathematical theories of stellar evolution.

For that matter, you can't prove heliocentrism either, as it's physically impossible to position an observer in a location that can capture the entire rotation of the Earth around the Sun. Heliocentrism is an inference based on multiple different types of data.

As for evolution, well, frankly, it is two parts. The first is somewhat unprovable--that current forms are the descendants of unlike forms which changed over time in response to the environment. The second is the well-documented change in organisms' genetic code over generations via reaction to environmental conditions, such as what happens when you don't take the whole bottle of penicillin. The first is frankly, in some sense, unprovable without a time machine. It is an educated supposition based on the sound science of the second observations. However, aliens planting all of the evidence and landing us here would produce the same result. Evolution is merely the most parsiminous explanation of biodiversity. It isn't the same as say, astrophysics, which, if the math is correct, uses current conditions to find the only possible past universe which could have produced those conditions.

Which is a blatant double-standard because you're applying last-tuesdayism (the counterthesis that everything was created last tuesday in the same shape as now) to biology and not astrophysics.

The Big Bang theory is an inference put together from multiple different lines of evidence (the CBE, the expansion of distant galaxies, and ratios of elements in intergalactic gas). Evolution is an inference put together from multiple different lines of evidence (fossil cladistic relationships, genetic cladistic relationships, quantitative genetics, observed selection and speciation in current time, the success of epidemiological theories.) The question is why are you applying one standard to astrophysics and not evolution, given the inherently inferential and speculative nature of both?
posted by KirkJobSluder at 10:10 AM on May 11, 2010 [4 favorites]


The core of his approach is that self-reporting is accurate and can therefore constitute the basis for scientific measurement. Bollocks.

And yet, self-reporting has about the same reliability rate as many of the early experiments in physics and astronomy.
posted by KirkJobSluder at 10:15 AM on May 11, 2010


So much of this stuff is clearly pseudo-science. Its people taking their own biases and pressing them on to double-blind studies and doing all sorts of shady things. Its exactly why "creation science" exists.

... However, aliens planting all of the evidence and landing us here would produce the same result. ...

Actually, this is the exactly how creation science operates, both in regards to evolution and astrophysics. Since we can't prove uniformitarianism or disprove some big event that makes the multiple lines of evidence pointing to deep time misleading, Biblical theories deserve equal treatment to evidence-backed theories.

But thankfully, science doesn't work on proof, it works on showing that hypotheses are so unlikely as to be practically impossible. The hypothesis that an alien species:
1) altered the chemistry of our planet to be consistent with billions of years of oxygen-generating ecosystems
2) planted meters-thick fossil beds hundreds of meters thick all over the planet
3) falsified tens of thousands of fossils
4) falsified existing chemical metabolisms
5) falsified multiple genetic sequences,
6) falsified anatomical features

in millions of species across every ecosystem in a way to be consistent with quantitative genetic models of descent with modification just to uplift a few monkeys is so improbable that it's not a working framework.
posted by KirkJobSluder at 10:57 AM on May 11, 2010 [2 favorites]


More importantly, I do think it does open a window into the untested assumptions discussed. For more of the same kind of junk, read Daniel Gilbert's Stumbling on Happiness. Its the same pseudo-science packed into a NYT best seller. And he's a Harvard psychologist. The core of his approach is that self-reporting is accurate and can therefore constitute the basis for scientific measurement. Bollocks.

Wrong. Self-reporting is measurable and can experiments can be designed to study it and scientists know what interventions affect it, in a predictable way. Not only that, self-reported happiness is exactly what they're trying to study (and increase), so it only makes sense to use that variable.

Dozens and dozens of experiments have been designed and conducted, with reproducible results. You know, like in science and stuff. It's amazing that you think Dan Gilbert, Martin Seligman and others in that field haven't considered these issues.
posted by r_nebblesworthII at 11:04 AM on May 11, 2010 [1 favorite]


Sniping at classification methodologies as non-scientific is mystifying to me.

Dude, you are not helping. Tight science does not operationalize "nice" like that and then call it a classification methodology.

Because when some "science" gets unproven, people tend to think the rest of it is bunk, which is not true.

I cannot fathom how the chance that an uneducated misinterpretation of a damn NYT article would cause someone to reject all science as bunk, but even though you claim that's what you're afraid of, isn't that exactly what's happened to you?

Science is essentially a process of disproving hypotheses. That's what it's all about. If uneducated yahoos fail to understand that and choose to think that the disproving of a hypotheses or the sound razing of bunk science means that all science is wrong, then quite frankly, they probably lack the capacity to contribute to or understand the scientific process anyway. I don't have time to try to correct every fool ass notion in the world. On the contrary, I seek to provide evidence of what's actually going on, present it in such a way that other people who have spent the time and effort to understand it with the same frame of reference can readily interpret my findings and then start the process again based on their measured response.

Here's the problem--my beef isn't with the studies. My beef is with the article.

That NYT times article is clearly not a scientific publication. Take issue with an attempt to communicate a position to a lay audience if you like. That sort of column is not targeted at extreme skeptics. Duh. If you want to be skeptical, then go read the fucking science.

You are in no way presenting your argument as if your problem is with the NYT article. You stated that chemistry, physics, and astronomy are the only sciences you think are "real." This is a serious claim that essentially suggests to me that you think my life's work is meaningless.

How about biophysics? Electrophysiology? Neuroscience? Can you seriously assert that the motherfucking Nernst equation is not science?

If we were to have an argument about the law, and I read a poorly written 9th grade civics textbook and proceeded to base my entire argument on it that essentially boiled down to an indictment of the entire legal system, would you take me seriously for a split second? Keep in mind, I have not read nor am I basing my argument upon anything other than this textbook. I cannot imagine that you would have much respect for that position.

If your beef is actually with the article, frame your argument as such. If you want to attack science, then don't beat up his retarded little brother, scientific journalism. Come play with the big kids or go the fuck home.
posted by solipsophistocracy at 11:23 AM on May 11, 2010 [1 favorite]


Wrong. Self-reporting is measurable

I never said it wasn't. What I said was that it is impossible to prove that it accurately represents what the experimenters say it does.

The huge problem here is that the experiments rely on the assumption of outside actors to provide context to the experiments:

To find out, we tested 8-month-olds by first showing them a character who acted as a helper (for instance, helping a puppet trying to open a box) and then presenting a scene in which this helper was the target of a good action by one puppet and a bad action by another puppet. Then we got the babies to choose between these two puppets. That is, they had to choose between a puppet who rewarded a good guy versus a puppet who punished a good guy. Likewise, we showed them a character who acted as a hinderer (for example, keeping a puppet from opening a box) and then had them choose between a puppet who rewarded the bad guy versus one who punished the bad guy.

The "good guy" and the "bad guy" are set by the moral judgments of the researchers. There is no measurement of whether or not the child actually thinks one of the puppets is "bad" or "good." That is imposed from the outside. Even worse, there is an assumption that reaching for the puppet equals perferring or liking the personality of the puppet. While our own experience and self-awareness would like to tell us that this is true (and it very well might be), the experiment tells us no such thing. Take away the assumptions and you have an experiment that tells us nothing.

The problem is compounded by our own highly-developmed emotional system which is highly invested in the idea of a fair world, where people can be easily divided into good and bad and that can be understood based on our own emotion. In such a world "good actors" are rewarded and the "bad actors" are punished. These biases seem to seep into these cognitive-psych experiments from the beginning--indeed it is natural they would because the things we most want to know involve questions like this.

This is very much unlike the Eta Carinae or Top Quark examples. The prediction that Eta Carinae will go supernova in the near future is based on the fact that observable data from the star mimic those of a star that went supernova after experiencing an event much like the one that Eta Carinae went through in the 1840's where it appeared to act like a supernova and became the second-brightest star in the sky despite being 5-7 kpc away. In other words, there are no inferences made about the content of the data. We don't make a judgment call about what the variability rate means other than to compare it to another variable star that displayed similar characteristics and then went supernova.

With the babies in the above experiment, there is an inference about what the content of the data is in the first place. The inference is that a grabbing motion towards one or the other means the baby wants X. We have no way of measuring that. For the situations to be analogous, we'd have to have an internal system that has feelings about the variablity of light coming from stars and an emotional reaction to certain variability profiles which migh bias our judgment of what happens next.

I'm not saying these researchers are wrong in their estimation of what happens in a baby's mind. I'm saying their experiments, as described in the article, do not prove what they say they do.
posted by Ironmouth at 11:32 AM on May 11, 2010


Can you seriously assert that the motherfucking Nernst equation is not science?

That's hard as nails science. You're measuring physical characterisics. There's no doubt that much of biology, biophysics, and the like is rock solid hard science based on phsyics and chemistry.

But when you talk about fuzzy categories like "morality," "bad," and "good," you are not in the realm of science. There are too many assumptions built in, and too much potential for bias. When they start reading the bits and bytes of neuro transmission and break down the actual code of what is being thought, then we can call it science.
posted by Ironmouth at 11:38 AM on May 11, 2010 [1 favorite]


Ironmouth: The huge problem here is that the experiments rely on the assumption of outside actors to provide context to the experiments:

Except that this is always the case in regards to experiments without exception. You can't conduct an experiment (not the only form of data used in science BTW) without making theoretical assumptions about the variables recorded.

Even worse, there is an assumption that reaching for the puppet equals perferring or liking the personality of the puppet.

Sure, and it's probable that they actually tested the assumption that the babies reach out for things they like in other, less arguable choices. And almost certainly the designation of good/bad is based on other theoretical work regarding moral decision-making, likely constructs that are reliable for other age groups, as would be work that toddlers see puppets as social rather than inanimate.

All of that prior work would be substantially cited and detailed in the peer-reviewed work. Without examining that prior work, you have no basis on which you can make the claims that these assumptions are ungrounded.

The prediction that Eta Carinae will go supernova in the near future is based on the fact that observable data from the star mimic those of a star that went supernova after experiencing an event much like the one that Eta Carinae went through in the 1840's where it appeared to act like a supernova and became the second-brightest star in the sky despite being 5-7 kpc away. In other words, there are no inferences made about the content of the data. We don't make a judgment call about what the variability rate means other than to compare it to another variable star that displayed similar characteristics and then went supernova.

Many astronomers do make that judgement call, which is based on a large number of assumptions about the data 1) 20th and 21st century astronomical observations can be consistently compared with 19th century data, 2) observations are diagnostic of the internal dynamics of the star, 3) observations correspond with theoretical models.

So we have three parts here.
Data: Eta Carinae brightened using such and such a measurement method (which would have been carefully validated by prior work)
Theory: The brightness of Eta Carinae corresponds to pre-supernova stellar evolution.
Inference: Eta Carinae might explode in the next few million years.

The baby research has three parts.
Data: Babies prefer helpful puppets to hindering puppets using such and such a measurement method (which would be carefully validated by prior work)
Theory: Babies are capable of rudimentary moral cognition.
Inference: Babies have a preference for good over bad.

Which, granted looking at moral cognition is a bit more tricky than looking at arithmetic. But the key issue here is that there was a significant and measurable difference. If all actions taken by puppets were treated with equal behavior, then the researchers would fail to reject the null hypothesis.
posted by KirkJobSluder at 12:04 PM on May 11, 2010


You're measuring physical characterisics.

And behavior isn't physical?
posted by KirkJobSluder at 12:11 PM on May 11, 2010 [1 favorite]


When they start reading the bits and bytes of neuro transmission...

Just caught this. While our inferences of human cognition based on behavior, are certainly fuzzy and theoretical, we certainly do know enough to reject most mind-as-computer models in favor of something more along the lines of mind-as-interconnected-organs.

The problem is that we really can't say much about how those organs might work without correlating inputs to outputs. It's like saying that genetics was a pseudoscience until we fully understood the central dogma of molecular biology or sequenced the human genome.

Anyone who says that chemistry and physics doesn't rely on similar inferential constructs really doesn't understand those disciplines.
posted by KirkJobSluder at 12:34 PM on May 11, 2010 [1 favorite]


The problem is compounded by our own highly-developmed emotional system which is highly invested in the idea of a fair world, where people can be easily divided into good and bad and that can be understood based on our own emotion.

This sort of question is highly applicable to the lines of research pursued by my various advisors and mentors. I am in the process of developing a research program based on a synthetic approach that relies on findings from a number of subfields.

I am of the belief that measuring fluctuating hormone levels, the social context in which a decision is made, neural activity, and reaction time with respect to behavioral decisions in the Ultimatum Game are a, dareIsayit, fair way to investigate the biological underpinnings of people's sense of fairness.

Self-report data about emotional states is useful in this sort of investigation as well, not because the way that people introspect about their decisions and the affective components thereof can reliably inform us about the cognitive, systems, and cellular level processes in and of itself, but because measuring the relationship between people's subjective explanations of their own behaviors/decisions and the biological modulation of behavior or decision-making can inform us about the extent to which our sense of a fair world is based on mathematical calculations at the neural level.

Fear is an emotion that has received considerable attention in laboratories with approaches that range from molecular, "Nernsty" kinds of measures to those more like my own, in which we study the broader neural response to the perception of fear in others as it changes across development. We can study fear in rodents, and because the differences between human brains and rat brains is primarily one of quantity rather than quality (and I am not speaking here to a more abstract comparison between "what it is like to be a rat" and "what it is like to be a human," but to a more rigid comparison between the properties of electric meat contained in within different sorts of mammalian skulls. This is more difficult with the more abstract subjective states, like a moral sensibility about fairness, but we are making progress toward being able to address higher-order emotions as well.

By defining situations in which people are forced to make decisions about fairness, we do not aim to boil down an intricate system to a sound-byte. No one is legitimately asking "ooh, is that the fairness spot, in the orbitofrontal cortex? You know, just below the cytoarchitecturally defined Brodmann's Area 10 to include rostral portions of BA 12 but not BA 11. Yes, there is crappy science out there in which people do make ridiculous claims like that, (though it was far more prevalent in the early days of fMRI [see the infamous 'Love Ventricle']) but that is precisely why science is a process of disproving hypotheses.

Do you choose to reject the clearly unfair offer of an $8/$2 split? How about $6/$4? An economist might suggest that you should accept any offer, because rational self interest dictates that $1 is better than $0, and who the fuck cares what the other guy gets? People don't do that though. They punish unfair offers by deciding that no one gets paid, although this tendency to punish unfair offers can be changed by disrupting the electrical activity of the dorsolateral prefrontal cortex (on the right, but not the left). This behavioral tendency to reject unfair offers and the modulation of that tendency (which requires a synthesis of physics [gotta use the magnet to deliver the electrical pulse] and anatomy/physiology [gotta know where in the brain you're putting that pulse]) is readily observable, and easily measurable, if you have the proper tools and background. The neural and endocrinological correlates of decisions like this are becoming better understood every day. We do ask people: "why did you reject that offer?" What they say is not by any means the end of the road for us. It is another variable, but it is not what allows us to learn more about how brains encode things like fairness, or their logical precursors.

A clear, mathematical split between fair (equal payoff for both parties) and unfair (anything else) offers may not be what you mean when you say fair, but it is difficult for me to accept the idea that this measure of equality vs. inequality is any less real than something like mass or a redox reaction. Before someone up and empirically demonstrated that reliable and valid measurement of mass is readily replicable in a variety of circumstances (and that replicability is really the lynch-pin here), it was just some thing some dude thought. In fact, it still is. If someone else comes along and demonstrates conditions under which our measurement of mass is inaccurate, and this demonstration can be repeated, we will have to seriously rethink some shit. All good scientists are prepared to deal with this contingency. Now, realistically, do we think that things like the way we index mass are honestly likely to be demonstrably falsified? No, of course not. We are comfortable with that assumption because it has been verified over and over again. People don't waste their time trying to disprove shit like that, because there are plenty of equally real phenomena that we have yet do demonstrate over and over again. The only way mass is "real," however, is in the context in which it is measured. There is no inherent "massness."

Observing the judgment of an event as fair or unfair by humans is every fucking bit as real as observing the position of celestial bodies to make inferences about motion that allow for calculations of planetary mass. Physics has a couple hundred years on us, so perhaps they're well beyond the planetary motion of the solar system stage and can start to investigate more complex issues about the synchronous motion of the universe.

The study of human emotion and its role in decision making, however, is not currently prepared to make the kind of coherent statements that can be distilled into an NYT column with no content that is not perhaps somewhat misleading to someone who thinks about it in more than a cursory way but less than a rigid empirical one.

We still have to talk about it though. If people are not aware that things like this are being studied and of the importance of continuing such lines of research, then we will not be able to communicate the broader importance of our work so that we can secure the funding we need to do it and pique the curiosity people who might be well suited to helping us in our pursuit of knowledge that this kind of science even exists. That NYT article is not there to stand up to hardcore scrutiny. It is there so that people can say "oh, wow. You can even study that kind of shit? I had no idea..." Some of those people will help direct the necessary resources our way, because to investigate things with a level of specificity we can stand behing is not cheap. Some of those people will think about and apply what we find to ultimately fucked up ends. Some people will take what we find and figure out a way to use it to help people with profound problems. Some people will see an NYT article and begin to think about these kinds of issues in a reasoned but speculative way and make a career out of it. And an extremely small number of people will go on to think about these things in a rigid, controlled way, developing an approach that is well-grounded in empirical findings but theoretically motivated.

Is it honestly your stance that we should not make an attempt to pursue these kinds of questions or inspire others to do so? Do you think that what I do is unscientific just because it involves an interdisciplinary approach based explicitly on things that you do regard as science? Sure, it makes for a more complex problem, but I don't shy away from complexity or throw my hands up and say "Well, if it's going to be even remotely fuzzy at any level, then we should just give up, regardless of how solid each other level is."

I hate bad science reporting, too. I would speculate that I hate it even more than you do, if something like hatred of a practice so abstract as "bad science reporting" is currently quantifiable (but I don't really have a clue about that, so perhaps we can just agree to hate it equally).

This column is not bad scientific journalism, however. There is plenty of that, and plenty of it that has been thoroughly ripped up right here on the Blue by various experts in a wide range of fields. This column is written by one of the primary motivators of this line of research. It is what he thinks is important to communicate to a lay audience. The man is talking about a straight up behavioral tendency to reach for one object or another. He describes this as a preference, and for the purposes of explanation, summarizes it as reaching for the "nice" or "mean" puppet. However, IN THE SAME DAMN PARAGRAPH he addresses the sort of variables he controls for and indicates that he did not submit this research to a peer-reviewed journal with operationalizations for "niceness." Science is complicated, and the dude is breaking it down so that people who don't have extensive education can see what he's talking about on a more rudimentary level.

If you have an extensive education (which I have a sneaking suspicion that you most certainly do), then do you honestly think that article is designed to stand up to your rigid argumentative standards? Even the peer-reviewed articles that this column is about are not designed to stand up to the kind of argumentative deconstruction you were trained for. I have much respect for the extent to which your systematic approach can be applied to so many different kinds of situations. Thinking like an attorney is neat in that it allows you to adopt a set of cognitive techniques by which you can rationally assess a huge number disparate issues as they apply to the practical resolution of human conflict. I understand that many times, lawyers have to (or get to) become temporary experts in whatever matter is at the center of the issue with which they are currently dealing, if not to the extent that they must communicate this expertise directly then at least to the extent that they can confer with more highly specialized experts in such a way as to convey that information to total non-experts. That is an interesting process, and the kind of training it takes to be that flexible in your ability to address whatever issue might come up is impressive and useful for what you do. I definitely strive to be adaptable, and take an interdisciplinary, multi-methodological approach. Nonetheless, I am motivated by a limited set of theoretical questions that I can ask with a limited set of practical techniques.

I sacrifice that sort of flexibility so that I can show people specific things instead of tell them about whatever thing is necessary. Explaining something or persuading/convincing someone that your explanation is reasonable is important to everyone. When a scientist explains things to the readership of the New York Times, he is telling them. When he explains it to the scientific community, he is showing them. Do not conflate the two. If you want to see what he's talking about, go look. If you don't like what he is saying and you want to call bullshit, well that is right on and you are encouraged to do that. However, the way to call bullshit in science is to roll up your fuckin' sleeves, demonstrate that his hypotheses are not empirically supported, demonstrate it again, and then demonstrate it again. Demonstrate support for your own hypotheses while you're at it. Once you have shown a shitload of people (who are all equally likely to call bullshit on your demonstrations as well), then you get to this magical place, where you can just tell people things and they will listen. That doesn't mean they will (or that they in any way should) just accept your argument from authority, but once you have earned a certain degree of respect as a researcher, you can start to make speculative claims so that other people will test them. That's what this guy is doing. He is telling a broad audience about some technical work in the hopes that it will cultivate interest and disseminate knowledge.

I think it is very, very important to break down the differences between what it is we know and what it is we suppose and not to treat those things as the same. Because when some "science" gets unproven, people tend to think the rest of it is bunk, which is not true.

The assumptions that you are comfortable with and the assumptions that I am comfortable with are very different, but when smart people who are good at rational argument deconstruct things they do not fully understand in a compelling way, that is what makes people think science is bunk. Not some guy talking about his line of research to a broader audience.

It is hard enough to convince most people that science has merit. Even if the only science you think has merit is physics/chemistry/astronomy (and sometimes biology? I guess it's like 'y'...), we don't need any help turning people off to science.

If you want to grab a bucket and a shovel, come on down and party, but don't bring a backhoe and don't poop in our sandbox.
posted by solipsophistocracy at 2:25 PM on May 11, 2010 [1 favorite]


Don't quit your day job, Ironmouth.
posted by five fresh fish at 7:22 PM on May 11, 2010


Sure, and it's probable that they actually tested the assumption that the babies reach out for things they like in other, less arguable choices

How? How? Really? They cannot tell what a baby likes or dislikes. They have to work on an assumption. Whereas the luminosity of a star is a raw measurement.

Seriously, you must acknowledge the assumption here.
posted by Ironmouth at 5:44 AM on May 12, 2010


Ironmouth: How? How? Really? They cannot tell what a baby likes or dislikes. They have to work on an assumption. Whereas the luminosity of a star is a raw measurement.

No, the luminosity of a star is not a raw measurement because there's no practical way to capture the energy output of stars directly. Luminosity is an inferred value based on theoretical assumptions. If the distance to the star is known you estimate luminosity from apparent brightness (the raw data) using the inverse square law (a theoretical framework). If the distance is unknown, you correlate (a theoretical framework) spectrographic features (raw data) to known stars.

Of course, astronomers have had to unravel chicken and egg issues in regards to apparent brightness, luminosity, and distance for some time. And all of those steps requires a mess of theoretical assumptions. Originally, apparent brightness was subjectively eyeballed in comparison to reference stars. Then, you had to deal with the issues of calibrating photographic media over a range of exposure and development conditions, before finally, we have measurements that are critically dependent on our understanding of QM WRT semiconductors. Translating CCD readings to apparent brightness requires even more assumptions.

While mechanical methods of measurement are certainly more reliable than human observation, they are not necessarily more valid and they are certainly not raw data. The use of any measurement device is an abstraction that has to be theoretically and empirically supported. Which is why your claims that assumptions and conclusions made about Eta Carinae are fallacious. With radically different means of measurement in 1840, 1950, and 2010 (human comparison, chemical photography, and CCDs), there's a mess of theoretical baggage in saying that an event from 1840 is similar to one that's happening today.

This is methodology 101 here. If you don't understand this, you can't claim to understand science as a practice and discipline.

Seriously, you must acknowledge the assumption here.

Certainly, I've acknowledged all along that their methods involve assumptions. The difference between you and me is that I understand that theory-grounded assumptions are an inherent part of the hypothesis-testing process, and without reviewing the prior research regarding the reliability and validity of studies regarding infant cognition, I can't make claims those assumptions are inherently unwarranted.
posted by KirkJobSluder at 7:37 AM on May 12, 2010 [2 favorites]


How? How? Really? They cannot tell what a baby likes or dislikes.

They cannot tell what a baby likes or dislikes. You are on the level on that front. "Like" is a direct self-appraisal, right? Only you can report on what state you perceive in yourself. Functionally though, if the baby repeatedly reaches for the same thing, it is fair to say that the baby preferentially reaches for it, right? The probability that the baby will reach for something is calculable. Is that fair to say?

They have to work on an assumption. Whereas the luminosity of a star is a raw measurement.

Seriously, you must acknowledge the assumption here.


How the flying fuck can you even begin to think that the luminosity of a star is a "raw thing." What the hell does "raw" begin to even mean?

You can't begin to measure things like the amount of radiation emitted by a celestial body over time without a metric fuckload of assumptions. You can't measure without assumptions.

The amount of willful ignorance and ill-founded criticism you exhibit is deplorable. Your mystical devotion to whatever bizarre conception of "physics" that exists in your mind is in and of itself a grievous insult to physicists the world over.

Please stop denigrating the name of science with your idiocy.
posted by solipsophistocracy at 7:39 AM on May 12, 2010 [1 favorite]


solipsophistocracy: They cannot tell what a baby likes or dislikes. You are on the level on that front. "Like" is a direct self-appraisal, right? Only you can report on what state you perceive in yourself. Functionally though, if the baby repeatedly reaches for the same thing, it is fair to say that the baby preferentially reaches for it, right?

While I'd say that moaning about a difference between "preferential reaching" and "like" is largely a matter of semantic pissing. Perhaps its a relevant discussion to have in the context of the highly-technical language of an academic paper. But casually speaking it strikes me as on the same level of bitching about whether the statement "Jupiter is big" refers to mass or radius.

Yes, yes, there's significant issues involved with making inferences about internals by observing externals. That's not a problem that's unique to psychology though, and physics is a poor model for an alternative. Cosmology depends on the assumption we can infer a larger picture from tiny snapshots and evidence out of context. And particle physics depends on making inferences of black-box collisions based on decay products.
posted by KirkJobSluder at 7:59 AM on May 12, 2010


I pretty much agree, actually. Operationalization is crucial. I think if you interact with the baby, you can probably tell damn well what it likes. If you use a rigid measure, that doesn't meant that you can't assess whether or not it works in a subjective way, and communicate that casually. I agree that you can restrict the rigor to the article, but I was busy slamming my head against a wall, and foolishly. I was just trying to figure out whether Ironmouth was applying a sort of logic that I am remotely familiar with. I foolishly thought that perhaps we could reach some area where we could agree, but that was stupid, because I think any agreement in an argumentative context is a mutually accepted assumption, which we all know is BAD.
posted by solipsophistocracy at 12:46 PM on May 12, 2010


« Older This is not your father's shack.   |   Escape from Somalia Newer »


This thread has been archived and is closed to new comments