Mar 31 2010

Magnets and Morality

MIT researchers have published a paper in which they use transcranial magnetic stimulation to alter, or bias, the moral judgments made by test subjects. Many people have e-mailed me this story with the comment that it seems fishy to them. Using magnets and changing morality so simply triggered their skeptical detectors, which is reasonable. But in this case the research seems perfectly legitimate, although some of the reporting has been superficial or dubious.


The whole notion of using magnets as a biological or medical intervention has long been exploited by the dubious magnetic therapy industry. I do agree that there are many fraudulent or quacky magnetic devices out there with unscientific claims. I do not recommend that people buy magnetic insoles for their shoes or strap refrigerator magnets to their joints to relieve pain or promote healing.

However, transcranial magnetic stimulation (TMS) is different. The brain is an exquisitely electrical organ, allowing for electromagnetic interventions to its function. We have recently developed the technology to use strong magnetic fields, precisely tuned and focused, to either stimulate or inhibit the electrical activity in brain tissue.

I recently wrote about a study that showed that such an intervention could potentially help relieve migraine headaches. TMS is also being used to treat depression – a more modern manifestation of the old electric shock treatment (ECT). Implantable stimulators are being used to treat seizures and symptoms of Parkinson’s disease. It is likely in the future we will be using many more electromagnetic interventions for neurological conditions.

So don’t be put off by the whole magnet thing – in this context it is legitimate.

Changing Morality

The other aspect of this story that triggered many people’s BS detectors is the notion of changing moral choices by simply affecting one small part of the brain. This aspect of the study was misrepresented in many of the mainstream reports that I saw. The discussion on the MIT website itself is much better.

The brain region in question, the right temporo- parietal junction (TPJ) is involved with a function known as the theory of mind – the ability to imagine what some other creature is thinking or feeling. In other words, we understand that other people have their own thoughts and feelings similar to our own , and this enables us to consider their possible motivations.

This has obvious survival advantages, especially for such a social creature such as humans.

The theory of mind is also critical to making moral judgments. Most moral judgments are strongly connected to the intention of the person, not just the final outcome. It is the theory of mind, and therefore the TPJ that enables us to consider a person’s intention in arriving at a moral judgment.

For example, the study in question posed to subjects the following scenario: Two people are visiting a chemical plant. Person A asks person B to get them a coffee. Person B does so, and puts a substance in person A’s coffee that is labeled “poison.” It turns out the substance was just sugar, and person A is fine.

This scenario is designed to separate out intention from outcome. Subjects were asked to rate the behavior of person B from a moral perspective, rating it from totally forbidden to permissible. When the TPJ was inhibited by TMS (but not a nearby area) subjects were more likely to find the behavior permissible, because there was no bad outcome. This suggests that their theory of mind was impaired and their judgments therefore were less based upon intention.

This research is a follow up to earlier research where fMRI scans were used to see what brain regions are active when making such moral judgments. This new evidence helps confirm the prior association and show probable causation.


This research shows how fMRI and TMS are being used in a complementary way to determine the functional anatomy of the cortex – even the most abstract intellectual functions. These technologies are still tricky to use, however, and I always emphasize that I would consider all such results preliminary until they are independently replicated.

Also, regarding moral judgments, the researchers emphasize that there is no one “moral center” in the brain, but rather moral judgments involve coordinating multiple different types of processing in the brain and bringing it all together for a final decision. It is literally a balancing of many small assessments and then deciding what the net result is.

What this research shows us is the role of the theory of mind in weighing the intention of a moral actor. We consider not only the outcome of people’s actions, but what they intended. It is fascinating that there is a specific neurological function that underlies these moral judgments.

19 responses so far

19 Responses to “Magnets and Morality”

  1. JurijDon 31 Mar 2010 at 10:16 am

    I just finished reading the article as it also came up high on my “new-research” radar.

    The thing I’d like to point out is that their their findings, though quite interesting, would need replication on a larger group of people. The study consisted of two experiments that tested two different ways of TMS induced RTPJ inhibition (25 min of TMS stimulation before making moral judgments and brief 500ms bursts of TMS stimulation right before the subjects were required to make each moral judgment).

    Experiment 1 was conduced on only 8 subjects and Experiment 2 on 12. While their findings were identical in both experiments the effect size was quite small and only present in one specific type of moral judgment (negative belief & neutral outcome). It is curious that the same effect was not present in the condition negative belief & negative outcome but given their raw data this might be the result of a low subject number.

    What gives me reason to pause is the fact that TMS stimulation did not produce a consistent effect (in the same direction) in all conditions, i.e. the stimulation was not always more prone to produce judgments that certain acts are more morally permissible. There is in fact quite a bit of variability between the experiments and conditions.

    Apart from that the subjects were asked to judge the moral implications on a 7 point scale (0-forbidden, 7 permissible). The statistically significant difference in the above mentioned single condition was about -0.8 points in the TMS condition relative to no stimulation (cca 2 for non-TMS and 2.8 for TMS). This is not a big effect and given the fact that both experiments were carried out on such small numbers of participants its relevance is questionable.

    This finding is interesting but I dont’ think it deserves the wide media hype that is being generated around it as the effect is very small and only present in one specific condition and the trial was made on a very small number of subjects.

  2. superdaveon 31 Mar 2010 at 10:56 am

    Your explanation here is much more coherent and believable than other explanations I have seen.

  3. tmac57on 31 Mar 2010 at 11:09 am

    The question that I have concerning TMS, is what are the possible risks for unwanted permanent changed in the brain?
    This is from the Mayo Clinic web site:

    “Long-term effects unknown”

    “Because transcranial magnetic stimulation involves changes in brain function, unknown long-term adverse health effects are possible. Some studies have shown structural changes in the brain after transcranial magnetic stimulation. The significance of these changes isn’t yet known. Also, the long-term effects of exposure to the strong electromagnetic fields involved remain unknown. “

  4. Steven Novellaon 31 Mar 2010 at 11:34 am

    JuriJD – I completely agree. As I said, these are preliminary, but that point could have been more emphasized as you did.

    I give it some credence because this was a follow up to prior studies suggesting the same connection. But it’s all tentative until we see some more robust follow up.

  5. JurijDon 31 Mar 2010 at 11:59 am


    I agree, I have been involved in research using TMS machines and we’ve had more than one argument about how to phrase the application for the medical ethics oversight board and also the text given to prospective participants to read before applying. Researchers usually do not stress enough the relative unknowns when dealing with this kind of brain stimulation but at the same time one does not want to “scare away” potential subjects by making it appear as there is some great medical risk to them – basically you get a “textbook” conflict of interests.

    In addition a wide and heterogeneous array of stimulus parameters can be used in TMS (the type of stimulus pulse: bi vs. mono phasic, its temporal course, repetition rate, absolute peak strength in terms of magnetic field in Tesla etc.).

    There has been a “consensus” conference on the topic in Italy 2 years ago and based on that a review article was published.:

    Safety, ethical considerations, and application guidelines for the use of transcranial magnetic stimulation in clinical practice and research.

    We’ve used those conclusions when dealing with TMS in terms of limiting exposure and formulating explanatory forms for subjects.

  6. Fred Cunninghamon 31 Mar 2010 at 12:02 pm

    It is interesting that in the latest alumni edition of MIT’s Technology Review there was an article of research being done on autism. They used a similar question about sugar/poison in coffee to determine intent in order to diagnose Asbergers. I doubt that it would be a surprise to you that all their research is pointing to genetic causes.

  7. artfulDon 31 Mar 2010 at 1:38 pm

    Tends to make you see the value of purpose as opposed to its goal rather then the value of the goal as opposed to its purpose.

  8. Steve Pageon 31 Mar 2010 at 3:14 pm

    # tmac57on 31 Mar 2010 at 11:09 am
    The question that I have concerning TMS, is what are the possible risks for unwanted permanent changed in the brain?
    This is from the Mayo Clinic web site:
    “Long-term effects unknown”
    “Because transcranial magnetic stimulation involves changes in brain function, unknown long-term adverse health effects are possible. Some studies have shown structural changes in the brain after transcranial magnetic stimulation. The significance of these changes isn’t yet known. Also, the long-term effects of exposure to the strong electromagnetic fields involved remain unknown. “

  9. Steve Pageon 31 Mar 2010 at 3:24 pm

    Sorry, premature post. 🙁
    tmac, most of these concerns are covered in the Rossi et al (2009) paper that JurijD linked to, but, just to be clear, TMS is extremely safe as long as one sticks to agreed protocols; participants are screened prior to the experiment, and, in our lab, anyone who has a family history of epilepsy, who is taking drugs (e.g. anti-depressants) – anyone who basically shows any indication that we may risk triggering seizure/syncope – is politely informed that we won’t be requiring their participation. As JurijD said, ethical approval is tricky at the best of times, and there are plenty of healthy participants available, so there really is no need to risk jeopardising future ethical approval by testing on an “at-risk” participant.

  10. tmac57on 31 Mar 2010 at 4:50 pm

    JurijD and Steve Page- Thanks for the information. I haven’t seen much of a concern raised before in discussions about TMS, but it looks like there has been a fairly long history of investigation (10 years or more) of its effects. It occurred to me, that if it is being investigated for use in treatment of depression, that there might be the potential for long lasting or permanent psychological effects, some of which could be negative.

  11. JurijDon 31 Mar 2010 at 8:14 pm


    you are of course quite right. whenever we poke around in the body there is the real possibility that our actions will do harm. And what TMS does is to magnetically induce eddy electrical currents in brain tissue that basically “stun” neurons thus obstructing normal function.

    In essence it’s like throwing a wrench into a certain machine section and looking for what stops working when you “disrupt” that particular part.

    Curious (at least to me when I first started dealing with it) is the fact that there do not seem to be any serious side effects to the procedure (if you exclude people who might get a seizure or are taking psychiatric medication).

    I do still hold the belief that the procedure is quite new and that we would be well advised to firmly err on the side of caution when dealing with TMS. The medical profession has enough blunders in its past already and for a lot of them the contemporary view was that “the procedure was not harmful”, only for that to be later shown as quite uninformed (even for the time).

  12. eiskrystalon 01 Apr 2010 at 3:57 am

    If basically it’s not moral to “mess” with people, even if it turns out ok in the end, what does that say about the whole moral god concept?

  13. sonicon 01 Apr 2010 at 4:32 pm

    This is an interesting result, but–
    Small sample, small effect, and I think it is important to note that the subjects in this study are probably not representative of the human population as a whole.

    JurijD- I haven’t seen the original research report- (not currently available to me)-
    Are there really claims of ‘statistical significance?’

    This would be an example of the type of abuse that was pointed out in the recent article by Tom Siegfried and commented on here earlier.

  14. JurijDon 01 Apr 2010 at 8:27 pm


    I wouldn’t go as far as saying there is any king of statistical “abuse” or misrepresentation in the paper. Quite the contrary, I think it’s reasonably well done and their methodology is rigorous as far as one can judge given the length of the paper (the proceedings usually don’t go into much methodological detail due to paper length restrictions)

    The results are statistically significant in both experiments (p<0,05)

    the things I'd pick on are:
    – low subject number
    – inconsistent TMS effects
    – small effect

    however because they performed two separate experiments (on two different groups of people) that basically had the same result their conclusions have more credibility.

    In spite of that this kind of psycho-physiological investigation requires many more participants to average out the inherent inter-subject variability. This could all just be a fluke. If you look at their graphs you can get a better feel for the huge variability involved.

    If they had done the same experiments with say 30 or 40 people, I'd be much more impressed and would probably feel their conclusion of the RTPJ controlling "empathy" (in these kinds of moral dilemmas) as reasonably well established.

    One wonders why only 8/12 subjects were included given the fact that these were all young healthy volunteers. Such low subject numbers are not uncommon in neurophysiology research but usually one only sees this when dealing with some rare disease that makes it really difficult to gather up your experimental group.

    My guess is that the limiting factor were their fMRIs which were done for each participant. These can get expensive and especially time consuming if (as is usually the case) one has to "beg" for research time on the MRI machine or work around a busy clinical schedule.

  15. JurijDon 01 Apr 2010 at 8:35 pm

    P.S. if you still can’t get a hold of the paper I can send you a pdf copy but only if you promise you won’t rat me out to the science publishing anti piracy squad 😉 oh and you’ll have to somehow get your email over to me… which I’m not sure you can do privately here. I do have the same nick on the SGU forums, you can search for me there and send me a PM.


  16. sonicon 02 Apr 2010 at 3:32 am

    Thank-you for your offer- I will decline (neither of us needs any trouble)

    My point is- It seems from what I’ve been able to read that the authors are drawing conclusions about how people make moral choices. These conclusions may or may not be valid, but they cannot be based on a ‘significant statistical’ result from the experiment they describe.

    1) The sample they tested is not representative of the population they are drawing conclusions about. (I doubt that the people being tested are a valid sampling of MIT sophomores, much less humanity). No-no.

    2) Even if we imagine that the people being tested are representative of the whole, the sample size is too small to be meaningful.

    It seems they are using the term ‘statistically significant result’ to justify the conclusions being drawn. This might be an error on my part, but it is clear the result they got has no statistical significance with the overall population of humans.

    Oh, we probably agree here…

  17. DLCon 02 Apr 2010 at 8:07 am

    Two thoughts:
    First, this will have the conspiracy freaks freaking out.
    I can hear it now :”We always knew *They* had mind control technology, so why are *they* now revealing the tip of the mind control iceberg?!”
    Less insane but still troubling is the people who are going to wonder “wait, what is my television set doing to me!?”
    (substitute in the electronic item of your choice )

    But from a more rational perspective, I have to wonder how much
    of this study is akin to other studies which got blown out of proportion, particularly by the popular media, who only want the sensationaliz-able portions. “Magnets can change morals!” cries the headline-writer.

  18. GodHeadon 04 Apr 2010 at 4:32 am

    So when will we have a weapon, that can do this at a distance, to someone’s entire brain?

  19. daedalus2uon 04 Apr 2010 at 9:12 am

    The fields that are needed to do this are gigantic. It takes a couple of Tesla, and that field has to change extremely rapidly, 100 microseconds is a typical timing. This is a magnetic field change of 2/0.0001 or a dB/dt of 20,000 Tesla/second. Waving a permanent magnet (0.2 T for very large and strong rare earth magnets) around gives maybe 0.2/1 or about 0.2 T/s, five orders of magnitude less. TMS is a threshold event. If you don’t exceed the threshold that causes nerves to fire, then nothing happens.

    The field changed caused by rotating in the Earth’s magnetic field is ~0.00005/1 (50 microTesla/1 second). The fields of concern from electric power lines are on the order of microTesla/s with the positive findings of leukemia in the vicinity of power lines occurring at 0.4 microTesla. There is no known or plausible mechanism for 0.4 microTesla to have biological effects.

    Aluminum foil would partially shield the time varying magnetic fields used in TMS. But the currents induced by the field would cause the aluminum foil to melt or vaporize.

Trackback URI | Comments RSS

Leave a Reply

You must be logged in to post a comment.