But it is important to note, once again, that correlation does not imply causation -- no matter how fancy a scientific tool you are using, either fMRI or genetics.
The best data Kandel cites in support of his argument is the improved treatment outcomes that could be "predicted" due to genetics. However, it would not be too surprising if some behavioral or cognitive measure did just as well at predicting the improved outcome.
It's pretty funny that Kandel would ever receive this critique. If you've read any of his work he is relentless at proving causation. His Nobel was earned through painstaking elimination of alternate possible methods for the formation of memory.
His genius is in seeing that A. Predictive power is scientific understanding and B. You can never be to thorough in understanding the chain of events between an event and its predicted outcome.
I highly recommend "In Search of Memory" as a wonderful autobiography, an inspirational demonstration of how science should be done, and a deep but accessible education in basic neuroscience. One of the best and most important books I've ever read.
Genetics are more fundamental than behavior or cognition. The prediction of a biological outcome using genetics tends to be more damning than if the predictor were something else. Genetics carry a lot of weight, you cannot teach them away, nor think them away and still in 2013 you cannot replace them. You are damned by them.
Plus, you cannot genuinely critique a revelation based on an article aimed at the general audience. That is a cheap shot with the potential to mislead self.
from a month or so ago with lots of links to full text articles about human behavior genetics from leading researchers on the subject. Genetics is a more complicated topic than most people suppose.
I don't think the article mentions causation in any way, and if so the journalist is probably overemphasizing the point.
The fMRI studies do not imply causation, but they are evaluated as a tool to predict successful therapies which is very valuable. No need for a cause and effect dualism yet.
I hope one day we'll have a complete model for understanding depression, but for now, it's encouraging enough to know that there are different kinds that are more/less responsive to different treatments.
Also, nice to see CBT get some props. It's really effective. And as someone who often proselytizes about the wonders of therapy, I'd like to get rid of the quacks who practice "black magic" therapy i.e. a non-evidence-based mish-mash approach. They are not helping the bias against mental health services in the US.
“Psychotherapy is an undefined technique applied to unspecified problems with unpredictable outcome. For this technique we recommend rigorous training.” -Horowitz
All therapy is "black magic" in the sense that no-one quite understands exactly how it works. CBT gets props because it's one of the few therapies that has a manual and can actually be studied scientifically.
I'm sorry but I prefer to speak with a therapist who doesn't have to consult a manual for every response he/she makes. Therapy is ultimately a human relationship, and just as messy and unpredictable as any other relationship - and subject to all the contingencies thereofe.
Just because something cannot be studied scientifically does not mean it is not effective. Evidence for the limited long term effectiveness (high relapse rates, ect) of shallow and manualized interventions like CBT is growing by the day. Compare that with more messy therapies that dig deeper (like psychodynamic) whose outcomes actually improve the longer the subjects are tracked.
The vast majority of the effectiveness of therapy is due to hope (placebo effect) and the relationship. Focusing on the "technique" is distracting and leads to a view of therapy as something that is done to you, as opposed to something that is done with you. Lasting change comes from taking personal responsibility, not from having special "scientific" techniques done to you.
That para-professionals with 2 weeks of training have outcomes as good or sometimes better than those with Ph.d's in clinical psychology is all one needs to know when deciding whether or not to categorize psychotherapy as a science.
CBT gets props because its empirical results have shown to be more effective than the traditional talk therapy approach. It's been around for a long time, has been studied enough for us to be confident that it makes a real difference, and I consider it as medicine for thoughts. Of course there's no way for us to quantify its precise method of action, so it's not that different than making the observation that the bark of a certain tree seems to reduce pain and inflammation, but if the burden of proof is to show direct causation from our thoughts and CBT to a better outlook on life, we might as well be asking about the nature of human consciousness.
CBT is a framework more than it is a manual, and not nearly as robotic as you've described. All therapy requires a human relationship to be successful. You have to trust your therapist to believe in the things they say, and that belief is a prerequisite for putting in the effort to reshape your thoughts and attitudes, but what's so robotic about a therapist asking you to explore the origin of a preconceived notion? About them asking you to be mindful of those type of thoughts? On coaching you to intercept those negative thought patterns and replace them with improved ones?
I think chalking up the efficacy of therapy to hope and placebo is a bit disingenuous. Yes, belief in change is a precursor to real change -- no therapist can help someone who isn't willing to believe or even want to change -- and in a way, the therapist's role is to guide you in your own endogenous healing process. However, there needs to be some rigor in the way they guide you to break bad patterns and replace them with better ones.
In a way, the therapist's challenge is like inception: how can I plant something in someone else's head but make them feel like they came up with it themselves? I think most of us could look at someone with low self esteem and negative thoughts and tell them what's wrong and what they should say or do or believe instead. However, that process needs rigor, and one of the best tools in the toolbox is to increase someone's mindfulness, and one of the best ways to do that is CBT.
I think overall you're a bit dismissive of it as a shallow, robotic approach when really it's a simple, repeatable, controllable frontline intervention. It's not going to work for deep-seated issues, but I think of it as almost being a prerequisite before those deeper issues can be tackled. I am of course speaking from personal experience, in both myself and what I've observed with others.
That hope and the relationship make up a much larger percentage of what causes change than technique is not my personal opinion. It's the result of a trans-theoretical analysis that was done a few years ago.
I don't recall the name of the study, but it's described in detail in "The Heart and Soul of Change".
> Lasting change comes from taking personal responsibility, not from having special "scientific" techniques done to you.
OCB, schizophrenia, PTSS disorders prove you wrong. And your comment about responsibility is ill-advised and bad attitude ; make a favour to people suffering from mental illness: don't try to comfort or coach them.
CBT gets quite a lot of props. It legitimately deserves many of them but probably gets more than it should, especially relative to "black magic" psychotherapies.
The main reason CBT is so popular in the industry is that it was designed from the ground up to be testable. CBT practitioners and CBT-friendly researchers have run a lot of tests, and they have found that CBT is effective.
For a long time, other modalities either weren't tested in this way or were only tested under CBT-friendly conditions. And this is a real black mark against them. But since CBT started eating their lunch from 2001 or so, other schools that could afford to started doing their own tests.
Guess what happened when psychoanalysts started running their own tests? If you guessed that they went home crying and gave everything up for CBT, you would be wrong. Instead, psychoanalytic-friendly researchers have found (empiricially!) that psychoanalytic treatments are effective, except for schizophrenia, just like CBT-friendly researchers found that CBT was effective for everything including schizophrenia (which really should have been a tipoff that something was amiss in the first place TBQFH). In other words, which therapy gets found to be effective depends a lot on which therapy the researcher wants to be found effective.
There are a ton of problems in measuring psychotherapy. Researcher bias is an obvious one. Ensuring that the things you're measuring are actually the things that you're supposed to be measuring (e.g. that the CBT patients are all actually getting by-the-book CBT, psychodynamic are getting psychodynamic, control are getting...what?) is another huge problem.
Across thousands of studies, the biggest variable in effectiveness of psychotherapy seems to be the therapist him- or herself. Some are drastically noticeably effective across a large number of patients and most are not. The effects of treatment modality, assuming you can pin it down at all, are seldom very significant, especially when blinds and control are used.
Other reasons CBT tends to be favored: emphasis on short duration is insurance friendly. Also, the 'cognitive' adjective implies that it has something to do with recent advances in neuroscience, which it doesn't.
(FWIW I've had both CBT and analysis, benefited from both in different ways.)
Agreed on all points. Really when I say "black magic" therapy, I'm being sloppy because what I really mean is that if you're not that good of a therapist, you shouldn't be offering anything but basic, rigorous types of therapy. I wouldn't go so far as to say that a good therapist offers a "human element", but skilled therapists are more observant, have stronger communication skills, have a more intuitive sense on when to challenge their patients, etc., and that allows them to go deeper into murkier waters than others. But to boil all of that down to "the relationship" and the "human element" feels overly reductive. There's still something quantifiable about the qualitative aspects.
The only therapies you can study scientifically (making sure everyone receives the same treatment in the treatment group), are therapies with manuals. However, only about 3% of all therapies have such manuals (CBT being one of them).
Most therapists highly individualize their approach - just because they have the same name doesn't mean they're actually doing anything similar. So claims of "empirical" proof of psychoanalysis is nonsense. The nature of what psychoanalysis is precludes that.
There actually are shorter-form therapies derived from psychoanalysis that do have manuals, and studies purporting to show their effectiveness. The point wasn't that these therapies are all that great - they probably aren't - but that measuring psychotherapy is a methodological ball of tar and prone to confirmation bias.
I tend to think of fMRI based neuroscience as analogous to trying to reverse-engineer how a computer works by measuring the electric fields it generates. A lot could be learned, but only to a degree. Except that there was no evolutionary pressure for understandability on the brain, so it's quite probably a convoluted kludge. This complexity cuts across levels of abstraction, with interplay between the genetic, neural and environmental levels. I suspect the entire thing is too complex for us to understand -- without better mathematical and computational tools. Anti-depressants work for some cases for some reasons, but we might need some paradigm leap to eventually understand why.
The current hypothesis of neurotransmitter imbalance always sounded childish to me. A bit like trying to fix a software bug by improving the conductivity of the circuits, or something like that.
I realize that this is aimed at a popular audience, but describing brain structure and function with imprecise words like 'circuits', 'controls', 'connected to' and 'involved in' without being clear as to a specific physical, biologic mechanism completely undermines his point.
Additionally, appealing to one studies which have not been independently reproduced, which are not double blind does little to dissuade me that the bulk of neuroscience is an exercise in confirmation bias.
"Mental illnesses are disorders of the brain that disrupt a person’s thinking, feeling, moods, and ability to relate to others. Just as diabetes is a disorder of the pancreas, mental illnesses are brain disorders... [They] are not a result of personal weakness, lack of character, or poor upbringing."
That's from a 2000 pamphlet written by the National Alliance for the Mentally Ill. Kandal's claim that this is new is BS - it's an old idea, and one not demonstrated by the research he cites. It's simply more evidence that mental illnesses are mediated by biological processes - something we've known for years.
The claim that mental illnesses are "caused" by biology is just as false now as it was then, and these experiments make no suggestion in either direction.
Clearly, neither study permits inferences about cause. Again, just because biological processes are involved does not mean that it is triggered by it. This is the same idea that's been pushed by over-zealous neuro-scientists for years. It's based on the belief that if we just drill down far enough in biology and genetics, we will explain most of the variance in behaviors.
If you don't cherry-pick the evidence, the whole picture suggests the same story that we've known for quite a while - but some in the psych/mental illness field aren't satisfied with it because the truth doesn't fit with their cravings for prestige - with our cultural obsession with the biological being more "real" than the social. Not to mention the financial aspects - one of the reasons we know so little about the social etiology of mental illness is because there's so little funding available for it.
The evidence suggests that environmental factors are just as involved - and in many cases is the primary cause, which plays on biological vulnerabilities that determine how the illness is expressed and through which it becomes self-perpetuating, not only for the individual, but on down through the family through cultural transmission and epigenetics.
Why does this even matter though? Because our perceptions of mental illness influences its outcome - how we view mental illness determines in large part what it is and how it expresses itself. If we believe it's fundamentally a fixed genetic and biological disorder that the individual themselves is helpless to do anything about, then through the nocebo effect it becomes a self fulfilling prophesy. That is one reason why outcomes for schizophrenia are better in sub-saharan Africa than they are in the US.
Scheid, T. L., & Brown, T. N. (Eds.). (2009). A handbook for the study of mental health: social contexts, theories, and systems. Cambridge University Press.http://hmid.basijmed.ir/public/hmid/books/mental%20health/A_.... 104 to end of chapter provides a good overview)
Whitaker, R. (2011). Anatomy of an epidemic: Magic bullets, psychiatric drugs, and the astonishing rise of mental illness in America. Random House Digital, Inc..
On social Expectancy:
Burns, J. K. The Social Determinants of Schizophrenia: An African Journey in Social Ep...
For a long time it was thought that mind and brain were not synonymous.
I do not argue it either way though - my point was that Kandel's claim that mental illnesses are fundamentally biological, that people are not at all responsible and that our brains "control our actions" is not supported by his examples or the evidence.
As TelmoMenezes said above, explaining mental illness entirely through neurological glitching is like explaining a software bug by claiming a circuit board wasn't soldered correctly.
Yes, there can be "hardware" causes, and yes, the "software" does run on the "hardware" as substrate, but the hardware is not a useful model of the software. Neurology is not necessarily an entirely useful model of an individual mind.
The alleged new thing about the research he cites is that it's based on a 'circuit' model rather than a chemical imbalance model. The other new[ish] thing is that strong-version biopsychiatry advocates feel the need to explain themselves at all. It's kind of fun to see one on the defensive for a change.
22 comments
[ 3.3 ms ] story [ 53.4 ms ] threadThe best data Kandel cites in support of his argument is the improved treatment outcomes that could be "predicted" due to genetics. However, it would not be too surprising if some behavioral or cognitive measure did just as well at predicting the improved outcome.
His genius is in seeing that A. Predictive power is scientific understanding and B. You can never be to thorough in understanding the chain of events between an event and its predicted outcome.
I highly recommend "In Search of Memory" as a wonderful autobiography, an inspirational demonstration of how science should be done, and a deep but accessible education in basic neuroscience. One of the best and most important books I've ever read.
Plus, you cannot genuinely critique a revelation based on an article aimed at the general audience. That is a cheap shot with the potential to mislead self.
https://news.ycombinator.com/item?id=6078734
from a month or so ago with lots of links to full text articles about human behavior genetics from leading researchers on the subject. Genetics is a more complicated topic than most people suppose.
http://www.mit.edu/newsoffice/2013/an-easier-way-to-control-...
The fMRI studies do not imply causation, but they are evaluated as a tool to predict successful therapies which is very valuable. No need for a cause and effect dualism yet.
I hope one day we'll have a complete model for understanding depression, but for now, it's encouraging enough to know that there are different kinds that are more/less responsive to different treatments.
Also, nice to see CBT get some props. It's really effective. And as someone who often proselytizes about the wonders of therapy, I'd like to get rid of the quacks who practice "black magic" therapy i.e. a non-evidence-based mish-mash approach. They are not helping the bias against mental health services in the US.
All therapy is "black magic" in the sense that no-one quite understands exactly how it works. CBT gets props because it's one of the few therapies that has a manual and can actually be studied scientifically.
I'm sorry but I prefer to speak with a therapist who doesn't have to consult a manual for every response he/she makes. Therapy is ultimately a human relationship, and just as messy and unpredictable as any other relationship - and subject to all the contingencies thereofe.
Just because something cannot be studied scientifically does not mean it is not effective. Evidence for the limited long term effectiveness (high relapse rates, ect) of shallow and manualized interventions like CBT is growing by the day. Compare that with more messy therapies that dig deeper (like psychodynamic) whose outcomes actually improve the longer the subjects are tracked.
The vast majority of the effectiveness of therapy is due to hope (placebo effect) and the relationship. Focusing on the "technique" is distracting and leads to a view of therapy as something that is done to you, as opposed to something that is done with you. Lasting change comes from taking personal responsibility, not from having special "scientific" techniques done to you.
That para-professionals with 2 weeks of training have outcomes as good or sometimes better than those with Ph.d's in clinical psychology is all one needs to know when deciding whether or not to categorize psychotherapy as a science.
CBT is a framework more than it is a manual, and not nearly as robotic as you've described. All therapy requires a human relationship to be successful. You have to trust your therapist to believe in the things they say, and that belief is a prerequisite for putting in the effort to reshape your thoughts and attitudes, but what's so robotic about a therapist asking you to explore the origin of a preconceived notion? About them asking you to be mindful of those type of thoughts? On coaching you to intercept those negative thought patterns and replace them with improved ones?
I think chalking up the efficacy of therapy to hope and placebo is a bit disingenuous. Yes, belief in change is a precursor to real change -- no therapist can help someone who isn't willing to believe or even want to change -- and in a way, the therapist's role is to guide you in your own endogenous healing process. However, there needs to be some rigor in the way they guide you to break bad patterns and replace them with better ones.
In a way, the therapist's challenge is like inception: how can I plant something in someone else's head but make them feel like they came up with it themselves? I think most of us could look at someone with low self esteem and negative thoughts and tell them what's wrong and what they should say or do or believe instead. However, that process needs rigor, and one of the best tools in the toolbox is to increase someone's mindfulness, and one of the best ways to do that is CBT.
I think overall you're a bit dismissive of it as a shallow, robotic approach when really it's a simple, repeatable, controllable frontline intervention. It's not going to work for deep-seated issues, but I think of it as almost being a prerequisite before those deeper issues can be tackled. I am of course speaking from personal experience, in both myself and what I've observed with others.
I don't recall the name of the study, but it's described in detail in "The Heart and Soul of Change".
OCB, schizophrenia, PTSS disorders prove you wrong. And your comment about responsibility is ill-advised and bad attitude ; make a favour to people suffering from mental illness: don't try to comfort or coach them.
The main reason CBT is so popular in the industry is that it was designed from the ground up to be testable. CBT practitioners and CBT-friendly researchers have run a lot of tests, and they have found that CBT is effective.
For a long time, other modalities either weren't tested in this way or were only tested under CBT-friendly conditions. And this is a real black mark against them. But since CBT started eating their lunch from 2001 or so, other schools that could afford to started doing their own tests.
Guess what happened when psychoanalysts started running their own tests? If you guessed that they went home crying and gave everything up for CBT, you would be wrong. Instead, psychoanalytic-friendly researchers have found (empiricially!) that psychoanalytic treatments are effective, except for schizophrenia, just like CBT-friendly researchers found that CBT was effective for everything including schizophrenia (which really should have been a tipoff that something was amiss in the first place TBQFH). In other words, which therapy gets found to be effective depends a lot on which therapy the researcher wants to be found effective.
There are a ton of problems in measuring psychotherapy. Researcher bias is an obvious one. Ensuring that the things you're measuring are actually the things that you're supposed to be measuring (e.g. that the CBT patients are all actually getting by-the-book CBT, psychodynamic are getting psychodynamic, control are getting...what?) is another huge problem.
Across thousands of studies, the biggest variable in effectiveness of psychotherapy seems to be the therapist him- or herself. Some are drastically noticeably effective across a large number of patients and most are not. The effects of treatment modality, assuming you can pin it down at all, are seldom very significant, especially when blinds and control are used.
Other reasons CBT tends to be favored: emphasis on short duration is insurance friendly. Also, the 'cognitive' adjective implies that it has something to do with recent advances in neuroscience, which it doesn't.
(FWIW I've had both CBT and analysis, benefited from both in different ways.)
Most therapists highly individualize their approach - just because they have the same name doesn't mean they're actually doing anything similar. So claims of "empirical" proof of psychoanalysis is nonsense. The nature of what psychoanalysis is precludes that.
The current hypothesis of neurotransmitter imbalance always sounded childish to me. A bit like trying to fix a software bug by improving the conductivity of the circuits, or something like that.
Additionally, appealing to one studies which have not been independently reproduced, which are not double blind does little to dissuade me that the bulk of neuroscience is an exercise in confirmation bias.
That's from a 2000 pamphlet written by the National Alliance for the Mentally Ill. Kandal's claim that this is new is BS - it's an old idea, and one not demonstrated by the research he cites. It's simply more evidence that mental illnesses are mediated by biological processes - something we've known for years.
The claim that mental illnesses are "caused" by biology is just as false now as it was then, and these experiments make no suggestion in either direction.
Clearly, neither study permits inferences about cause. Again, just because biological processes are involved does not mean that it is triggered by it. This is the same idea that's been pushed by over-zealous neuro-scientists for years. It's based on the belief that if we just drill down far enough in biology and genetics, we will explain most of the variance in behaviors.
If you don't cherry-pick the evidence, the whole picture suggests the same story that we've known for quite a while - but some in the psych/mental illness field aren't satisfied with it because the truth doesn't fit with their cravings for prestige - with our cultural obsession with the biological being more "real" than the social. Not to mention the financial aspects - one of the reasons we know so little about the social etiology of mental illness is because there's so little funding available for it.
The evidence suggests that environmental factors are just as involved - and in many cases is the primary cause, which plays on biological vulnerabilities that determine how the illness is expressed and through which it becomes self-perpetuating, not only for the individual, but on down through the family through cultural transmission and epigenetics.
Why does this even matter though? Because our perceptions of mental illness influences its outcome - how we view mental illness determines in large part what it is and how it expresses itself. If we believe it's fundamentally a fixed genetic and biological disorder that the individual themselves is helpless to do anything about, then through the nocebo effect it becomes a self fulfilling prophesy. That is one reason why outcomes for schizophrenia are better in sub-saharan Africa than they are in the US.
Sources
On Epigenetics: Masterpasqua, F. (2009). Psychology and epigenetics. Review of General Psychology, 13(3), 194. http://pzacad.pitzer.edu/~dmoore/2009_Masterpasqua_Psych&Epi...
On Biological and Environmental Etiology: Bentall, R. (2009). Doctoring the mind: Is our current treatment of mental illness really any good?. NYU Press. http://www.power2u.org/downloads/Richard-Bentall-Broken-Brai...
Scheid, T. L., & Brown, T. N. (Eds.). (2009). A handbook for the study of mental health: social contexts, theories, and systems. Cambridge University Press.http://hmid.basijmed.ir/public/hmid/books/mental%20health/A_.... 104 to end of chapter provides a good overview)
Whitaker, R. (2011). Anatomy of an epidemic: Magic bullets, psychiatric drugs, and the astonishing rise of mental illness in America. Random House Digital, Inc..
On social Expectancy: Burns, J. K. The Social Determinants of Schizophrenia: An African Journey in Social Ep...
What illness has nothing to do with "biology" - robot illnesses?
The idea that there is exactly one cause for any event is ridiculous, this isn't specific to illnesses.
I do not argue it either way though - my point was that Kandel's claim that mental illnesses are fundamentally biological, that people are not at all responsible and that our brains "control our actions" is not supported by his examples or the evidence.
Yes, there can be "hardware" causes, and yes, the "software" does run on the "hardware" as substrate, but the hardware is not a useful model of the software. Neurology is not necessarily an entirely useful model of an individual mind.