Mind-body problem-Chomsky/Nagel

[Mr.Samsa]

'Philosophical' differences may not exist, but that doesn't imply one can explain behavior from observance.

Nobody has suggested we can. Given this response, and few others down below, I think you seem to be under the impression that the behaviorism under discussion here is logical/analytical behaviorism. That form of behaviorism has nothing to do with what is used in science, and I think is mostly dead in philosophy circles as it's a pretty hollow position to try to hold.

I know the terms can get confusing, and more so when behaviorists themselves mix them up, but radical behaviorism is really the only kind that is still relevant and is still alive. It is the analytical behaviorists who argue that we can infer mental states from public behavior - but no other behaviorist accepts this position. Skinner rightly points out that it's obviously absurd to suggest that 'being sad' is frowning and saying, "I'm sad".

The statement you quoted is simply a position of science - the idea that theories of behavior and mind must include observational elements. This doesn't mean that we should only study external behaviors, or that the mind has to be observable, but that if we are to make claims about the mind which are not empirical, then we have to have logical support for doing so.

Actually, I don't know what to do with the above sentence. It's mostly meaningless except for that it seems to claim that behaviorism ascribes to materialism.

Behaviorism is a philosophy of science, so it holds no real ontological position. It ascribes to methodological naturalism solely because that's what is necessary to do science, but beyond that no further assumptions are made. Of course, people and behaviorists themselves can make extra claims about what they believe, but they aren't central to behaviorism itself.

Behaviorism is plain wrong from a mathematical point of view. You cannot explain, hope to model, a complex entity from behavior solely, period.

We can understand the internal workings of an individual from studying the outside, and subsequently describing it in 'layman' terms of the inside? Idiotic.

This 'objection' was actually the defining feature of radical behaviorism. The identifier "radical" refers to the idea that inner states cannot be studied or understood by studying the observable/external behavior.

I am from CS, so I don't understand everything. But I can tell you one thing: It is impossible to derive the internal workings/behavior of an entity from studying its behavior, except for essentially stateless entities. It is also impossible to derive the behavior from studying physiology of entities, except for essentially stateless entities. These are hard mathematical facts.

Agreed.

Watson may be right that it is impossible to study the inside, but that doesn't imply that one can derive behavior from the outside. Seems Skinner developed some common sense.

Exactly. Skinner rejected the "behaviorist" position you're attacking.

Meaningless semantics. Everything is behavior for a sufficiently broad definition of behavior, just as everything is cake for a sufficiently broad definition of cake.

Not meaningless at all, actually. By describing everything an organism does as 'behavior', it conceptualises previously 'immaterial' entities as something that can be studied. The term 'behavior' can be changed to whatever you want, but the important part was that everything has a cause and effect.

As mentioned above, this claim is not controversial and people may accuse it of being trivial or meaningless, but this is because everybody accepts this claim now. But this wasn't always so acceptable - in the times of William James and Freud, and to an extent Watson, the idea that inner states can be studied scientifically was something that was unheard of.

As I said before, mathematically one can show that studying the outside isn't sufficient, and that studying the physiology of an entity, is also insufficient. So a mathematician can simply prove Skinner wrong.

(I would say it's even worse. Mathematically, for a sufficiently complex entity, understanding it is impossible from observing behavior, and worse, simple physiology is sufficient to generate incredibly complex behavior, so studying the physiology will tell you almost nothing about behavior. [STRIKE]Behaviorism, from a math point of view, is flawed beyond believe.[/STRIKE])

And the behaviorists agree with you.

Well, behaviorism caved into reality and common sense. What else was there to do?

I'm not sure if "caved in" is the right phrase, as that seems to imply that it ignored evidence or refused to shift from an unreasonable position.

I say nonsense. I'll give it to you that behaviorism only talks about organisms since it cannot explain anything except for essentially the most simple entities, microbes, a mathematical fact. Since practitioners cannot admit that, they therefor proceed to conflate humans with microbes which is an immoral act.

But behavioral psychologists regularly study and explain human behavior, including complex behaviors like language, and even how people converse. Not to mention the applied area of the field, applied behavior analysis, which regularly uses behavioral principles to study, predict and control the behavior of individuals in a wide range of contexts and behaviors, which includes cognitive-behavioral therapy which is a successful treatment for depression.

So I can't understand your position. Are you arguing that behavioral psychologists don't study humans, or are you arguing that all the studies on humans are just made up or something?

As for conflating microbes with humans being "immoral", I don't understand that at all. Under what moral system is such an act immoral? I don't think even religious people would argue that such a position is immoral, they just disagree with it.

 

[bohm2]

Behaviorism is a philosophy of science, so it holds no real ontological position. It ascribes to methodological naturalism solely because that's what is necessary to do science, but beyond that no further assumptions are made. Of course, people and behaviorists themselves can make extra claims about what they believe, but they aren't central to behaviorism itself.

I think this is the claim that is disputed by many rationalists/nativists like Chomsky. Pierre Jacob writes:

Chomsky‟s major input to the cognitive revolution lies in his criticism of the behaviorist confusion between evidence and subject-matter:

I think that there is some significance in the ease and willingness with which modern thinking about man and society accepts the designation “behavioral science”. No sane person has ever doubted that behavior provides much of the evidence for this study — all of the evidence, if we interpret “behavior” in a sufficiently loose sense. But the term “behavioral science” suggests a not-so-subtle shift of emphasis toward the evidence itself and away from the deeper underlying principles and abstract mental structures that might be illuminated by the evidence of behavior. It is as if natural science were to be designated “the science of meter readings”. What in fact would we expect of natural science in a culture that was satisfied to accept this designation for its activities? (Chomsky, 1968, 1972: 65)

The advent of the cognitive revolution was in turn responsible for the shift away from the study of human behavior towards the study of internal mental states and processes that may or not give rise to observable behavior.

Chomsky, Cognitive Science, Naturalism and Internalism
http://hal.inria.fr/docs/00/05/32/33/PDF/ijn_00000027_00.pdf

 

[Mr.Samsa]

I think this is the claim that is disputed by many rationalists/nativists like Chomsky. Pierre Jacob writes:

Chomsky, Cognitive Science, Naturalism and Internalism
http://hal.inria.fr/docs/00/05/32/33/PDF/ijn_00000027_00.pdf

But Jacob makes the exact same mistake that I've been discussing throughout this entire thread. Behavioral science is not a turning away from private events and abstract inner states - it is the direct study of those things. In other words, Skinner did not suggest that we study external behaviors to infer internal workings (i.e. "meter reading"), as he explicitly rejected such a simplistic position. He suggested that we attempt to study the inner workings directly.

This all stems from Chomsky's misunderstanding of the subject matter. The "cognitive revolution" was a rejection of S-R psychology. But psychologists tend to reject the notion that it was a "revolution" because it wasn't like there was any significant paradigm or resistance that they had to overthrow. The groundwork that needed to be done to convince the world of psychology that it needed to study the inner workings of the mind had already been done by Skinner and the radical behaviorists. Look at the work of Tolman and Guthrie, with their discussion of "cognitive maps" and thoughts causing behavior. Nobody at the time rejected the idea that cognition is something that should be studied.

Cognitivism therefore wasn't a "revolution", but just an extension of the ideas that were already in place. The behaviorists had already set out the methodology for studying cognition - the idea that organisms are information processors that are controlled by internal states. The only real transition was in the late 60s with Neisser's text "Cognitive Psychology", where the idea of information processors was taken from the behaviorism methodology and adapted using the metaphor of the computer.

There's a good article on this here: http://www.radford.edu/~tpierce/622%20files/Leahey%20(1992)%20The%20mythical%20revolutions%20of%20american%20psychology.pdf

Conclusion. The coming of cognitive psychology is
best regarded, not as the revolutionary creation of a new
paradigm slaying the older one of behaviorism, but as the
appearance of a new form of behavioralism based on a
new technology, the computer. By the 1950s, mediational
S-R behaviorists were already looking for ways to represent
internal processing of stimuli, and the computer
metaphor provided a better language than mediational
r-s notation did. Moreover, the existence of artificial intelligence—
the manufacture of information-processing
devices behaving intelligently and purposively—bolstered
faith in mediating mental processes by showing they could
be embodied in material devices rather than immaterial
souls (J. Miller, 1983). Information-processing psychology,
no less than any form of historical behaviorism, aims
at the description, prediction, control, and explanation
of behavior, without any special attention being given to
conscious experience (Tulving, 1989). Perhaps during the
feverish days of the 1960s, another, less behavioral, road
might have been taken—but it was not taken, at least not
by the main body of experimental psychologists. The
mainstream of psychology in 1992 remains as firmly behavioralistic
as it was in 1910.

 

[bohm2]

But Jacob makes the exact same mistake that I've been discussing throughout this entire thread. Behavioral science is not a turning away from private events and abstract inner states - it is the direct study of those things. In other words, Skinner did not suggest that we study external behaviors to infer internal workings (i.e. "meter reading"), as he explicitly rejected such a simplistic position. He suggested that we attempt to study the inner workings directly.

Let's forget all the reviews and assume these guys are wasted on drugs and just look at Skinner's papers. Have you read any of Skinner's works, in particular his "Science and Human Behavior" including the section "Why Organisms Behave"? Do you still feel like that part I bolded above, is consistent with his works?

Science and Human Behavior
http://www.bfskinner.org/BFSkinner/Society_files/Science_and_Human_Behavior.pdf

 

[apeiron]

Let's forget all the reviews and assume these guys are wasted on drugs and just look at Skinner's papers. Have you read any of Skinner's works, in particular his "Science and Human Behavior" including the section "Why Organisms Behave"? Do you still feel like that part I bolded above, is consistent with his works?

Skinner's position is completely reasonable because he carefully describes the limits of his project. But this is also why it does not scratch the mind~body problem. It carefully just does not go there.

As he says, his is a science devoted to the control of behaviour. And this then left open to cognitive science the question about the architecture of cognition.

The second link is useless in the control of behavior unless we can manipulate it. At the moment, we have no way of directly altering neural processes at appropriate moments in the life of a behaving organism, nor has any way been discovered to alter a psychic process. We usually set up the second link through the first: we make an animal thirsty, in either the physiological or the psychic sense, by depriving it of water, feeding it salt, and so on. In that case, the second link obviously does not permit us to dispense with the first. Even if some new technical discovery were to enable us to set up or change the second link directly, we should still have to deal with those enormous areas in which human behavior is controlled through manipulation of the first link. A technique of operating upon the second link would increase our control of behavior, but the techniques which have already been developed would still remain to be analyzed.

So in contrast to the public bashings that Skinner gave cogsci, and Chomsky gave Behaviourism, everything Skinner says here is good commonsense. There is no reason the two fields can't live alongside each other, and even complement each other.

Skinner correctly identifies the "enemy" as folk psychology and Freudianism (Freud being guilty of dressing up folk psychology as science). But even here his tone is reasoned rather than polemic.

What I don't accept is Mr Samsa's sweeping claim that Radical Behaviourism created the jargon, the causal concepts, the methods, the observation data, which then transferred seamlessly to become later the study of "the second link". And that this expansion of the field was so smooth, cognitive scientist are behaviourists, and behaviourist are cognitive scientists.

All mind scientists should have studied both to some level. But they are also different paradigms in terms of concepts, aims and methods.

Calling it empiricism vs rationalism is rather simplistic - I am tempted to call this folk philosophy. But there is something of this dichotomy in the division of the two fields. One asks what can we be most certain about if we are investigating the mind in terms of observables. The other is asking the same question in terms of general architectural or systematic principles.

Are cogsci and behaviourism opposed, or are they complementary? I of course would argue that they are both. Breaking things apart creates clarity, re-connecting them creates understanding.

And here Skinner is being very reasonable. He says I am not studying the intervening mental processes because I don't see how I could control them. So I am creating a methodology that puts all the attention on the contextual factors that I can control. But equally, these mental processes clearly exist, and someone else could study them.

Did Behaviourism pave the way for this study? Well yes it did in the sense that it did a deflationary job on folk psychology. Perhaps it also split off the field of "contextual factors" leaving cogsci to be purely a study of "the architecture of reason" - but in fact to me that was a bad thing so far as cogsci went, because it went far too far in becoming a study of the disembodied mind.

But I don't see that Behaviourism provided anything that helped very much in shaping specific cogsci hypotheses about the general architecture of cognition. Again, if you ask what kind of processing concepts were talked about in Behaviourism, you come back to simple associative chaining as the implicit architecture of thought.

I accept now, having read his critique of Thorndike (p67) that Skinner again carefully ruled out making any interpretations about cognitive architecture. It was clear to him that mind was much more than simple associative learning and so a science of behaviour should not start imposing simplistic explanations on intervening variables. If it couldn't talk about them in a way backed up by theory/data, then it should just remain silent on the matter.

This is sound policy. But again, it is why Behaviourism is seen as putting the question of mind off-limits (to a methodology) and so why a different methodology might arise to fill that gap.

Now I judge all this from a multidisciplinary perspective. Cogsci was in its way as deliberately limited in scope as Behaviourism. And there is a real problem of how much knowledge can be transferred in either direction between these domains. They certainly broke things apart in a specific way, but did they ever come to complement each other much?

In the 1970s - and I was only a hot-head teenager of course - my quick judgement was that these two fields, either jointly or separately, were not cutting it. To understand the mind, you had to bring in evolution, development, semiotics, anthropology, neurology, systems science.

If you are starting out now, of course, mind science is much more interdisciplinary. There is evolutionary psychology, cognitive neuroscience, comparative cognition, social anthropology, biosemiotics, etc. The fusions are happening.

But where is Behaviourism in all this? Perhaps it is pervasive as Mr Samsa argues because all these richer fields are indeed situating cognition in its various contexts - evolutionary, developmental, social, neurological. These give the empirical facts that constrain the otherwise dangerously unfettered imaginings of the rationalist tradition.

Or if we just define Behaviourism as the science of the control of behaviour (that avoids attempting to model intervening variables), then it is not much part of any of these fields. Although it persists, and even flourishes, as its own field of applied science.

 

[Q_Goest]

The Bishop article is worth reading: http://rsfs.royalsocietypublishing.org/content/early/2011/09/02/rsfs.2011.0065.full

Did you notice that the author, Robert Bishop, was a professor at Wheaton? I would be suspicious that his intent in writing about "downward causation" is to support his christian beliefs.

 

[apeiron]

Did you notice that the author, Robert Bishop, was a professor at Wheaton? I would be suspicious that his intent in writing about "downward causation" is to support his christian beliefs.

Yes, I certainly noticed that. But thank you for the ad hominen anyway. If you can show that Bishop has a motivation that distorts his account of the science (which is frequently an issue with intelligent design, etc) then please highlight it for us.

Systems science - like quantum mechanics - does attract fellow travellers. If you are religious or in other ways dualist/mysterian, you are naturally drawn to the kinds of science that seem to be asking the same questions, offering possibly similar answers.

It is just the same with Darwinian evolution, Newtonian determinism, or other arch-reductionist theories. They attract their fellow travellers too. Neo-liberal economists, eugenists, etc.

Science is always getting co-opted to support prejudices about the way the world should work.

So here you yourself seem to be making the argument, if an approach to science can be used to give credence to a belief system I don't like, then I don't want to believe that science either.

I don't see that as valid. I would rather you made an attempt to understand the science and address it directly.

Bishop is one of a dozen papers in a Royal Society special issue on interdisciplinary approaches. Do you really think if Bishop was a religious crank posing as a scientist they would publish him?

So your "suspicions" don't even seem to be very far thought through. My suspicion is you saw "religious college" and thought oh goody, I don't even have to have an argument against an argument I don't want to believe. Instead, I can offer a prejudice to counter a prejudice.

But of course, you may have actually read the paper and so have some critique of it as science...

 

[bohm2]

The second link is useless in the control of behavior unless we can manipulate it. At the moment, we have no way of directly altering neural processes at appropriate moments in the life of a behaving organism, nor has any way been discovered to alter a psychic process.

Weren't psychotropics available back then that could somewhat alter a psychic process? Not that they were very good. And how successful are behavioural therapies in comparison to person-centred therapies/psychotropic drugs/other methods? I mean, is the scientic studies for efficacy much more so than other less behaviorally-guided methods?

 

[Mr.Samsa]

Let's forget all the reviews and assume these guys are wasted on drugs and just look at Skinner's papers. Have you read any of Skinner's works, in particular his "Science and Human Behavior" including the section "Why Organisms Behave"? Do you still feel like that part I bolded above, is consistent with his works?

Science and Human Behavior
http://www.bfskinner.org/BFSkinner/Society_files/Science_and_Human_Behavior.pdf

I've read extensively on the history of behaviorism, including Skinner's work, and yes it certainly is consistent with his work. The section of "Why Organisms Behave" chapter, where Skinner discusses how inner variables like the subject being "afraid" of an experimental condition, or purposely wants to disprove the experimenter, is quite good for demonstrating how Skinner thinks that "[o]ther variables may, of course, affect the result". This is expanded on in the chapter "The Individual as a Whole", where he discusses thinking and private events.

I think it's important to distinguish between what he refers to as explanatory fictions, and inner states. Skinner slammed explanatory fictions, where the explanation that appeals to some unobservable inner state is inferred directly from the behavior making it redundant, and people often confuse this with the idea that Skinner rejects inner states - probably compounded by the fact that this was the position of the methodological behaviorists. His discussion on inner states largely revolves around how useful they are in scientific explanations, and when it is feasible or possible to include then into scientific explanations.

Skinner's position is completely reasonable because he carefully describes the limits of his project. But this is also why it does not scratch the mind~body problem. It carefully just does not go there.

As he says, his is a science devoted to the control of behaviour. And this then left open to cognitive science the question about the architecture of cognition.

Not quite - behavioral science is focused on the explanation, control and prediction of behavior, where "behavior" includes cognition.

What I don't accept is Mr Samsa's sweeping claim that Radical Behaviourism created the jargon, the causal concepts, the methods, the observation data, which then transferred seamlessly to become later the study of "the second link". And that this expansion of the field was so smooth, cognitive scientist are behaviourists, and behaviourist are cognitive scientists.

Read the behaviorist research of the time and look at the cognitive research. For example, look at Tolman's concept of "cognitive maps" - how does that differ from the cognitive concept of mental maps and schemata? It doesn't at all, because the cognitive research in that area stemmed directly from Tolman's work.

Look at the methodology that cognitivists use when studying cognition - do they assume that thoughts have causes and effects? Do they treat self-reported data as distinct from the phenomenon under study? Do they base their hypothetical constructs on empirical and observable evidence? Yes, yes, and yes. These are all major contributions to psychology that behaviorism brought along.

Just look at that list of most influential works in cognitive psychology that was presented earlier, where there were about 5-10 behaviorists on that list. Why would a "revolution" be significantly shaped and influenced by the paradigm it's attempting to overthrow?

All mind scientists should have studied both to some level. But they are also different paradigms in terms of concepts, aims and methods.

That's certainly too far. An argument can be made that cognitive psychology and behavioral psychology are distinct fields, but no argument can be made that they are different paradigms. One of the defining features of opposing paradigms is that they are incommensurable, but that clearly isn't true in this case. Look at the major fields of cognitive psychology; thinking, memory, psychophysics, etc, and where did some of the major breakthroughs come from? Behaviorists.

Calling it empiricism vs rationalism is rather simplistic - I am tempted to call this folk philosophy. But there is something of this dichotomy in the division of the two fields. One asks what can we be most certain about if we are investigating the mind in terms of observables. The other is asking the same question in terms of general architectural or systematic principles.

The two are the same thing. General architectural and systematic principles are based on what we can be certain about when investigating the mind in terms of observables. Keep in mind that the "observable correlates" that behaviorism discusses does not mean that the mind has to be observable. It means that our logical inferences and hypothetical constructs must be grounded in reality to some degree. Otherwise we are simply speculating and guessing.

Are cogsci and behaviourism opposed, or are they complementary? I of course would argue that they are both. Breaking things apart creates clarity, re-connecting them creates understanding.

And here Skinner is being very reasonable. He says I am not studying the intervening mental processes because I don't see how I could control them. So I am creating a methodology that puts all the attention on the contextual factors that I can control. But equally, these mental processes clearly exist, and someone else could study them.

Well he went further than that. He said that intervening mental processes can be studied and control, and then laid out the methodology for experimenters to do so. He, for the most part, wasn't personally interested in studying mental processes, but other behaviorists obviously took up his methodology and did so.

But I don't see that Behaviourism provided anything that helped very much in shaping specific cogsci hypotheses about the general architecture of cognition. Again, if you ask what kind of processing concepts were talked about in Behaviourism, you come back to simple associative chaining as the implicit architecture of thought.

I accept now, having read his critique of Thorndike (p67) that Skinner again carefully ruled out making any interpretations about cognitive architecture. It was clear to him that mind was much more than simple associative learning and so a science of behaviour should not start imposing simplistic explanations on intervening variables. If it couldn't talk about them in a way backed up by theory/data, then it should just remain silent on the matter.

This is sound policy. But again, it is why Behaviourism is seen as putting the question of mind off-limits (to a methodology) and so why a different methodology might arise to fill that gap.

But that's what cognitive psychology does as well, it adopted that policy directly from Skinner. Cognitive psychology does not talk about intervening variables that can't be backed by theory or data. Such work is rejected from science and is appropriately labelled "pseudoscience".

Now I judge all this from a multidisciplinary perspective. Cogsci was in its way as deliberately limited in scope as Behaviourism. And there is a real problem of how much knowledge can be transferred in either direction between these domains. They certainly broke things apart in a specific way, but did they ever come to complement each other much?

Massive amounts of information is passed between the two - as I've mentioned, cognitive psychologists and behavioral psychologists routinely work together and publish in the same journals. Look at the work on memory and the behaviorists' discovery of memory decay and interference, look at signal detection and the behaviorists' inclusion of the discriminability parameter, etc.

Go to any university with a cognitive and behavioral psych department, and have a look around. As well as using the exact same equipment to studying the same things, you'll also find that the two groups of researchers will often pop into each others labs, either working directly together, or at least discussing things with each other.

The idea that cognitive psychologists study something different from behavioral psychologists, or in a different way using different methods, is just so foreign to me.

In the 1970s - and I was only a hot-head teenager of course - my quick judgement was that these two fields, either jointly or separately, were not cutting it. To understand the mind, you had to bring in evolution, development, semiotics, anthropology, neurology, systems science.

If you are starting out now, of course, mind science is much more interdisciplinary. There is evolutionary psychology, cognitive neuroscience, comparative cognition, social anthropology, biosemiotics, etc. The fusions are happening.

But where is Behaviourism in all this? Perhaps it is pervasive as Mr Samsa argues because all these richer fields are indeed situating cognition in its various contexts - evolutionary, developmental, social, neurological. These give the empirical facts that constrain the otherwise dangerously unfettered imaginings of the rationalist tradition.

You seriously don't know where behaviorism is in the fields of evolutionary and cognitive psych, neuroscience, comparative cognition, development, social fields, etc?

Evolutionary psychologists use behaviorist techniques all the time because to determine whether a behavior has an innate element or not, they need to rule out learning as a possible cause. For example, the New Caledonian crow example I gave before, which is one of the shining jewels of the evolutionary psychology world. Cognitive psych, I've already pointed out various areas where it's important (including the entire field itself) but for a concrete example look at psychophysics which is entirely behavioristic. Neuroscience, we've been discussing already the breakthroughs that learning theory has provided both directly (breakthroughs in the discovery of neural processes like in-vitro reinforcement) and indirectly (the use of behaviorist techniques to test theories, e.g. conditioned fear paradigms). Comparative cognition - you're just taking the piss now. This area is dominated by behaviorists. And development and social areas, there's "social learning theory" which has had a significant impact in the area, not to mention the fact that behaviorist methods are used to test various theories (e.g. testing whether babies look at faces due to an innate feature, or through reinforcement contingencies).

Or if we just define Behaviourism as the science of the control of behaviour (that avoids attempting to model intervening variables), then it is not much part of any of these fields. Although it persists, and even flourishes, as its own field of applied science.

Behavioral psychology persists, and flourishes, as an interdisciplinary field of experimental and applied science. The majority of work done in behavioral psychology is experimental studying how organisms behave and think, and the applied work is a happy side-effect of this research.

 

[Mr.Samsa]

Weren't psychotropics available back then that could somewhat alter a psychic process? Not that they were very good.

Most of Skinner's work regarding the building of the foundation for behaviorism was done between the 30s and 70s (and realistically, after the 40s-50s, the field was not under his personal control as other researchers had begun to shape it in various ways) - for example, the book you linked to above was published in 1953. So during his time, psychotropics were crude at best. Yes, they could change neural processes but not with the precision that's needed to base a science of thought and behavior on. As the field progressed though, and things like fMRIs etc were invented, Skinner began discussing the importance of looking at changing neural processes (leading to works like Edelman's neural darwinism).

And how successful are behavioural therapies in comparison to person-centred therapies/psychotropic drugs/other methods? I mean, is the scientic studies for efficacy much more so than other less behaviorally-guided methods?

It changes depending on what you're specifically looking at, and it's not like a behavioral therapy will be best for all conditions all the time (as behavioral therapies can only help with conditions that can be modified by learning). With that said, behavioral therapies have proven to be hugely successful in a number of areas, like Cognitive-behavioral therapy which is one of the leading treatments for depression and anxiety (often in conjunction with medication). For phobias, there is systematic desensitisation which (as far as I know) is unparalleled. And it has various behavioral techniques for other problems, like eating disorders, self-injurous behaviors, and is currently the only treatment for autism (and arguably, it can be a "cure" in the sense that behavioral therapy can often help a child reach the point where they no longer meet the requirements for a diagnosis of autism). And, of course, one of the advantages of behavioral techniques is that they aren't limited to people with learning or mental disorders, and has been successfully applied to school settings to improve a variety of things, like the children's enjoyment of school and general happiness, their on-task behavior, their academic scores and later success in life, etc.

 

[apeiron]

Weren't psychotropics available back then that could somewhat alter a psychic process? Not that they were very good. And how successful are behavioural therapies in comparison to person-centred therapies/psychotropic drugs/other methods? I mean, is the scientic studies for efficacy much more so than other less behaviorally-guided methods?

Making a broad generalisation, most chemical or medical approaches to controlling/repairing the mind are quite unbelieveable a-theoretic. Drugs, lobotomies, electric shock - you are talking about "science" that is often about the level of kicking a TV set to make it work. With the difference that the brain is a self-organising system and so sometimes a scrambling blast of ECT does indeed cause some kind of homeostatic reset.

Cognitive talking therapies and densitisation therapies at least put theory and practice in reaching distance of each other.

You know how it still is for drugs. For people with mild problems, how much of any efficacy is placebo or nature doing the healing? For those with serious problems, either the treatment is still the metaphoric kick to the TV set or a way to make the symptoms go away to the extent that society is not bothered by them.

The record of other methods in Skinner's day - either medical or psychoanalytic - was in fact so bad, so lacking in a credible theoretic basis, that this would be one of the reasons to see Behaviourism as a great leap forward for science.

So no, the efficacy of the theory as applied science is in fact a prime justification of Behaviourism. And it didn't go round claiming to be able to cure organic problems like schizophrenia. Another proof that it knew more about what it was doing.

Well, there was the "refrigerator mother" hypothesis of childhood autism/schizophrenia - a contextual explanation for behaviour. But this came out of Freudian psychiatry rather than Behaviourism. And was utter nonsense of course.

Behaviourism could probably be blamed for aversion therapy - as still apparently used to cure people of homosexuality. But again, that is mostly psychiatrists mis-applying psychological theory to organic traits.

So on the whole, the record for Behaviourism seems very good on this score. And yeah, don't get me going on psychiatry and its history of voodoo thinking. Or big pharma's record on responsible medical research. The doctoring establishment still has a lot to live down.

 

[apeiron]

Read the behaviorist research of the time and look at the cognitive research. For example, look at Tolman's concept of "cognitive maps" - how does that differ from the cognitive concept of mental maps and schemata? It doesn't at all, because the cognitive research in that area stemmed directly from Tolman's work.

We keep coming back to your intent to claim all psychological schools are examples of behaviourism, making the term so elastic it is no longer useful in tracing the history of ideas.

Tolman is seen as a proto-cognitivist...

Tolman's purposive behaviorism was not as widely received in its day as other psychological theories. This was largely due to the fact that many did not consider its foundation to being in line with behaviorism at all, which was the dominating force in psychology at the time. However, the insistence on studying implicit mental concepts as opposed to looking solely at explicit behavior was an idea that opened the door to the school of cognitive psychology.

http://en.wikipedia.org/wiki/Purposive_behaviorism

And there were many more sub-schools like dynamic psychology...

Woodworth introduced and popularized the expression Stimulus-Organism-Response (S-O-R) to describe his functionalist approach to psychology and to stress its difference from the strictly Stimulus-Response (S-R) approach of the behaviorists in his 1929 second edition of Psychology[5]. He later published the theory in Dynamic psychology (1918) and Dynamics of Behavior (1958). Within his modified S-O-R formula, Woodworth noted that the stimulus elicits a different effect or response depending on the state of the organism. The “O” (for organism) mediates the relationship between the stimulus and the response.

http://en.wikipedia.org/wiki/Robert_S._Woodworth

Yes, I am interested in the similarities and the differences that mark the history. And there were many "proto-cognitivists" who were reacting to the patent oversimplification of the Watson and Thorndike style S-R approach. Skinner reacted by reaffirming he was not modelling the "O", whereas other contemporaries wanted to - and that led them to the kinds of experiments that showed there was something there worth studying in its own right.

You may want to sweep every one into the one bag so as to make the transition from behaviourism to cognitive science appear seamless and non-revolutionary. But maybe because in the 1970s I felt that there was recruitment campaign going on - "join us, we are the true science", "no join us, the others are all old fools" - I see the history through different eyes.

 

[bohm2]

With that said, behavioral therapies have proven to be hugely successful in a number of areas, like Cognitive-behavioral therapy which is one of the leading treatments for depression and anxiety (often in conjunction with medication). For phobias, there is systematic desensitisation which (as far as I know) is unparalleled. And it has various behavioral techniques for other problems, like eating disorders, self-injurous behaviors, and is currently the only treatment for autism (and arguably, it can be a "cure" in the sense that behavioral therapy can often help a child reach the point where they no longer meet the requirements for a diagnosis of autism). And, of course, one of the advantages of behavioral techniques is that they aren't limited to people with learning or mental disorders, and has been successfully applied to school settings to improve a variety of things, like the children's enjoyment of school and general happiness, their on-task behavior, their academic scores and later success in life, etc.

I had CBT on two different occassions. The last institutionally-based one was about ~1 year ago, although you're supposed to practise till forever. I have both anxiety (GAD and performance anxiety) and sensory issues (possibly due to mild form of ASD?) and I've also used many different medications. It's really hard to make a call on which was better for me. They all helped a bit but not enough. I also did an internship (both in medicine and pharmacy) at a large psychiatric hospital and was apparently using CBT, first as an intern and then as a patient). The funny thing is that I didn't even know at the time. I agree about systemic desenitization. I found gradual exposure therapy to be the most useful non-drug approach. But without medication, I'm sure I'd be screwed. I'm still doing that gradual exposure stuff now, I guess.

I'm still confused though. I think Chomsky's nativist/rationalist stance is pretty clear as I pointed out in a previous post where environmental cues/influence are considered to play a very minor role (equivalent to the development of other organs) whereas I'm guessing behaviourists, Piaget (empirical constructivists) and semioticians put a much greater empasis on environmental influence. So what is it that differentiates these 3 different perspectives? Is there a major difference with respect to the importance of environmental/social/cultural influence on behaviour/cognitive development/language/thought between these 3 perspectives or is it just "window dressing"?

 

[apeiron]

Evolutionary psychologists use behaviorist techniques all the time because to determine whether a behavior has an innate element or not, they need to rule out learning as a possible cause. For example, the New Caledonian crow example I gave before, which is one of the shining jewels of the evolutionary psychology world. Cognitive psych, I've already pointed out various areas where it's important (including the entire field itself) but for a concrete example look at psychophysics which is entirely behavioristic. Neuroscience, we've been discussing already the breakthroughs that learning theory has provided both directly (breakthroughs in the discovery of neural processes like in-vitro reinforcement) and indirectly (the use of behaviorist techniques to test theories, e.g. conditioned fear paradigms). Comparative cognition - you're just taking the piss now. This area is dominated by behaviorists. And development and social areas, there's "social learning theory" which has had a significant impact in the area, not to mention the fact that behaviorist methods are used to test various theories (e.g. testing whether babies look at faces due to an innate feature, or through reinforcement contingencies).

You are not getting it are you? You keep making reference to the experimental methods of behaviourism being part of the kit-bag of tools used by cognitive science to test its theories.

Good old behaviourial analysis can be used to rule out the simple stuff to leave you then with the hard bit that a more complicated cognitive architecture theory must explain.

 

[Mr.Samsa]

We keep coming back to your intent to claim all psychological schools are examples of behaviourism, making the term so elastic it is no longer useful in tracing the history of ideas.

Tolman is seen as a proto-cognitivist...

Yes he's seen as a proto-cognitivist now because he was a behaviorist studying what Skinner termed "private behaviors" - i.e. cognition. He was a behaviorist though, there was no difference between his research, methods, philosophy, beliefs, than that of any other behaviorist at the time.

The fact that one of the major behaviorists at the time is now termed a "proto-cognitivist" is evidence that the divide between behavioral and cognitive psych is not so wide.

And there were many more sub-schools like dynamic psychology...

Tolman's behaviorism was inconsistent with the behaviorism at the time (i.e. methodological behaviorism). We've been discussing throughout this thread, and I've presented numerous lines of evidence, that Skinner and the radical behaviorists did not accept just studying explicit observable behavior - so how can Tolman's philosophy be at odds with the behaviorists on the grounds that he did not agree with simply studying the explicit observable behavior?

Yes, I am interested in the similarities and the differences that mark the history. And there were many "proto-cognitivists" who were reacting to the patent oversimplification of the Watson and Thorndike style S-R approach. Skinner reacted by reaffirming he was not modelling the "O", whereas other contemporaries wanted to - and that led them to the kinds of experiments that showed there was something there worth studying in its own right.

Not quite true. Skinner's model was dependent entirely on the organism. What he was modelling was the organism. However, rather than having the organism as a discrete entity in a causal chain, he thought that the divide between the organism and environment was more vague, and that at times the organism itself was an environmental variable. This is why he held, more or less, to a Stimulus-Response-Stimulus approach - where an initial cue generates a response which produces a contingent consequence. The whole equation is the "organism", as it describes the feedback process that occurs within an organism as it interacts with its environment.

You may want to sweep every one into the one bag so as to make the transition from behaviourism to cognitive science appear seamless and non-revolutionary. But maybe because in the 1970s I felt that there was recruitment campaign going on - "join us, we are the true science", "no join us, the others are all old fools" - I see the history through different eyes.

In the 70s? The "cognitive revolution" was well over by then, according to all accounts. It began in the late 40s with the symposium where Chomsky gave a talk on syntactic structures, and was fully complete by the time Neisser formulated the computational theory of mind. If students at the time were battling amongst themselves, it doesn't seem to have affected the opinions of psychologists at the time - nobody described it as a revolution, and nobody was aware of any particular uprising or overthrowing. It wasn't until the 80s when Baars described it as a "revolution" that subsequent students began to call it such. By that time, most were under the impression that behaviorism was a blank slate theory of behavior that ignored thoughts and feelings - so of course such a ridiculous paradigm needed to be replaced! Even Baars, writing just a few decades after the action supposedly took place, had little understanding of what behaviorism was and is. And now we have writers like Pinker continuing the misunderstanding and misrepresentation.

I had CBT on two different occassions. The last institutionally-based one was about ~1 year ago, although you're supposed to practise till forever. I have both anxiety (GAD and performance anxiety) and sensory issues (possibly due to mild form of ASD?) and I've also used many different medications. It's really hard to make a call on which was better for me. They all helped a bit but not enough. I also did an internship (both in medicine and pharmacy) at a large psychiatric hospital and was apparently using CBT, first as an intern and then as a patient). The funny thing is that I didn't even know at the time. I agree about systemic desenitization. I found gradual exposure therapy to be the most useful non-drug approach. But without medication, I'm sure I'd be screwed. I'm still doing that gradual exposure stuff now, I guess.

Yeah it's not always a case of "one or the other", as sometimes problems are caused by a biological issue that can only be solved by medication or even surgery, and behavioral approaches can only help to a degree. For example, suppose someone has an extreme problem with aggression - teaching some meditation or relaxation methods isn't going to help if they have a problem with their pituitary gland pumping out adrenaline.

I'm still confused though. I think Chomsky's nativist/rationalist stance is pretty clear as I pointed out in a previous post where environmental cues/influence are considered to play a very minor role (equivalent to the development of other organs) whereas I'm guessing behaviourists, Piaget (empirical constructivists) and semioticians put a much greater empasis on environmental influence. So what is it that differentiates these 3 different perspectives? Is there a major difference with respect to the importance of environmental/social/cultural influence on behaviour/cognitive development/language/thought between these 3 perspectives or is it just "window dressing"?

Mostly the difference is just in degree. Chomsky largely argues for innateness, he believes that there are specific modules in the brain that control things like language generation. Skinner argued that there is a significant amount of learning that occurs during language acquisition, and he outlined exactly how this process occurs when we learn words and grammar, etc. But he still argued that there were important brain structures that were necessary for learning language, he was just less certain of the idea that there were structures for highly specific aspects of language. Piaget's ideas I'm less sure on, to be honest. As far as I know, he wasn't particularly interested in the details of language acquisition, and instead he came up with the stages of development in a child - and he argued that children need to reach these particular stages before certain levels of language development can be reached.

For Chomsky and Skinner, the general difference can be explained in the sense of domain-specific and domain-general brain processes. Both believed that the brain played a vital role, but Skinner argued that a lot of the linguistic elements that Chomsky argued had to be innate because they were too complex to be learnt, were in fact learnt.

You are not getting it are you? You keep making reference to the experimental methods of behaviourism being part of the kit-bag of tools used by cognitive science to test its theories.

Good old behaviourial analysis can be used to rule out the simple stuff to leave you then with the hard bit that a more complicated cognitive architecture theory must explain.

"Simple stuff"? :rofl:

But I think you missed where I pointed out that a number of evolutionary psychologists are behaviorists, not only studying the "simple stuff", but developing theories to explain the complex world of behavior and thought. For example, the work on empathy and altruism in the area (and in comparative cognition) is predominantly done by behaviorists. Surely those areas are relatively complex?

And as for the section on neuroscience, that was an example of an indirect influence of behaviorism in that area. If you don't like it, then you can just ignore it and focus on the more direct influences and works.

By the way, what area are you in? I assume you're not a psychologist as your perspective on the issue seems to be more from a pop-science perspective which, whilst detailed, seems to fall into the traps of looking in from the outside.

 

[apeiron]

In the 70s? The "cognitive revolution" was well over by then, according to all accounts. It began in the late 40s with the symposium where Chomsky gave a talk on syntactic structures, and was fully complete by the time Neisser formulated the computational theory of mind. If students at the time were battling amongst themselves, it doesn't seem to have affected the opinions of psychologists at the time - nobody described it as a revolution, and nobody was aware of any particular uprising or overthrowing. It wasn't until the 80s when Baars described it as a "revolution" that subsequent students began to call it such. By that time, most were under the impression that behaviorism was a blank slate theory of behavior that ignored thoughts and feelings - so of course such a ridiculous paradigm needed to be replaced! Even Baars, writing just a few decades after the action supposedly took place, had little understanding of what behaviorism was and is. And now we have writers like Pinker continuing the misunderstanding and misrepresentation.

As I say, I was there. So I watched things unfold in real-time.

Personally I saw a big difference between early cogsci - Neisser, Broadbent and others who were about "information processing" - and late 1970s/early 80s when symbolic processing, schematas, modularity and other strong AI ideas were fashionable.

Neural connectionism was also part of the early wave that got submerged during this symbolic processing period. Then it bounced back a decade later.

Baars' revolution was really the consciousness studies one. This was people in the late 80s/early 1990s saying cogsci was still not working on theories of mind - subjective experience - and so another revolution must be proclaimed.

There was a semi-Vygtoskian revolution that nearly happened in the late 1980s - social constructionist psychology - but this got swamped by evolutionary psychology, effectively the cogsci crowd adopting modern synthesis Darwinism.

Things are always happening. Often they go away for a while then re-emerge as the pendulum swings - as that Leahey paper correctly says. It is not actually one paradigm replacing another, but an ebb and flow between polar views.

But there is always the fighting talk because people really do seek to define themselves tribally. Social psychology even has a theory for it - boundary maintenance.

The philosophy of science historian then has the job of disentangling the ideas actually at stake. And the view is quite different depending on whether you want to be a lumper or a splitter.

I agree that I was taking the usual turf warfare/patch protection view of the history because it is colourful and memorable - it is what people do, what they find engaging. But if you want to take a more dispassionate and considered view, then I can be interested in that too.

As I've said, there are some deep differences here. You do have the materialist's description of causality. And you do have the structuralist's. You can treat this as an either/or polarity - one to be viewed as fundamental or primary, even if the other also exists. Or you can take a systems perspective which seeks to fit both extremes into the one causal model.

 

[MarcoD]

I know the terms can get confusing, and more so when behaviorists themselves mix them up, but radical behaviorism is really the only kind that is still relevant and is still alive. It is the analytical behaviorists who argue that we can infer mental states from public behavior - but no other behaviorist accepts this position. Skinner rightly points out that it's obviously absurd to suggest that 'being sad' is frowning and saying, "I'm sad".

Yah, it looks that I got confused by terminology since the psychologist use confusing terminology. Which is a problem, right? If you call something 'radical behaviorism,' I expect it to mean that - disregarding all history. From my point of view, if all definitions are scrambled to mean the opposite, or not quite the opposite, any statement on the topic becomes vacuous. (Or stated differently, you end up doing politics, not science. One could state, of course, that science has become politics only, but that's a different debate.)

The statement you quoted is simply a position of science - the idea that theories of behavior and mind must include observational elements. This doesn't mean that we should only study external behaviors, or that the mind has to be observable, but that if we are to make claims about the mind which are not empirical, then we have to have logical support for doing so.

Which is what I mean, I can see that debate raging among psychologist, but when in practice everybody agrees that you need to study both the white box and the black box to validate theories, it looks to me like a meaningless high-brow debate among scientists.

Behaviorism is a philosophy of science, so it holds no real ontological position. It ascribes to methodological naturalism solely because that's what is necessary to do science, but beyond that no further assumptions are made. Of course, people and behaviorists themselves can make extra claims about what they believe, but they aren't central to behaviorism itself.

Which is what I mean with a high-brow vacuous statement. Of course all psychology must adhere to basic scientific principles - both behaviorism or cognitive science. What else are they going to do? Bang drums in the hope of discovering new psychological theories? I am tempted to disregard these comments completely, except for that they probably have a historical connotation w.r.t. older psychology like Freud's. But then say that, not that psychology should adhere to scientific principle.

This 'objection' was actually the defining feature of radical behaviorism. The identifier "radical" refers to the idea that inner states cannot be studied or understood by studying the observable/external behavior.

Noted, I would say that's not radical but the opposite, hence my confusion.

Not meaningless at all, actually. By describing everything an organism does as 'behavior', it conceptualises previously 'immaterial' entities as something that can be studied. The term 'behavior' can be changed to whatever you want, but the important part was that everything has a cause and effect.

Which is something nobody can disagree with.

But I now read, and skipped, your other comments, and I can see where the debate stemmed from now, thanks for clearing that up.

So I can't understand your position. Are you arguing that behavioral psychologists don't study humans, or are you arguing that all the studies on humans are just made up or something?

Uhm, this was -I think- regarding Chomsky's view that behaviorism is empiricist, not a rationalism, therefor it doesn't study anything. I don't think anybody in the psychological field either knows or cares since they are doing a mixed approach anyway. If I get nasty, I would propose that they are doing a mixed approach since they have now conflated all terms beyond any meaning.

(Personally, I can see the Freud vs Behaviorism as a debate on whether one is going to do qualitative vs quantitative analysis, and yeah, it stems to reason that regarding humans one needs to do both.)

As for conflating microbes with humans being "immoral", I don't understand that at all. Under what moral system is such an act immoral? I don't think even religious people would argue that such a position is immoral, they just disagree with it.

I am not religious and I don't think religious people are more moral than other people. Sometimes, I would say even less so.

I already explained my position: People are not microbes, conflating the terms leads to treating people as microbes, therefor I find it an immoral act to overly use the term organism when (partly) addressing people.

I would also propose that it is unscientific since people, from a systematic complexity view, cannot be explained with the same simple models as the behavior of rats can be. It therefor leads to a view on humans which is overly simplistic, which should be avoided.

I agree it's a personal opinion.

 

[Mr.Samsa]

Yah, it looks that I got confused by terminology since the psychologist use confusing terminology. Which is a problem, right? If you call something 'radical behaviorism,' I expect it to mean that - disregarding all history. From my point of view, if all definitions are scrambled to mean the opposite, or not quite the opposite, any statement on the topic becomes vacuous. (Or stated differently, you end up doing politics, not science.)

It's mostly "confusing" to you because you don't know the history of the terms and the time in which it was coined. At the time, when methodological behaviorism was at its height, Skinner proposed something which changed the fundamental nature of behaviorism itself - he proposed that we use behaviorist methodology to study the mind. This was radical. This is what radical means.

Which is what I mean, I can see that debate raging among psychologist, but when in practice everybody agrees that you need to study both the white box and the black box to validate theories, it looks to me like a meaningless high-brow debate among scientists.

Not quite, because there is no debate ('raging' or otherwise) among scientists on this topic. The idea that behaviorism is controversial or debatable is like suggesting that there is controversy or debate over the theory of evolution, or global warming. No scientist really debates those issues, only the public do (and usually when they're misinformed).

Which is what I mean with a high-brow vacuous statement. Of course all psychology must adhere to basic scientific principles - both behaviorism or cognitive science. What else are they going to do? Bang drums in the hope of discovering new psychological theories? I am tempted to disregard these comments completely, except for that they probably have a historical connotation w.r.t. older psychology like Freud's. But then say that, not that psychology should adhere to scientific principle.

My comment that you're responding to here was a response to the claim that behaviorism assumes an ontological position. I respond by pointing out that it doesn't assume an ontological position (i.e. materialism), and instead it is simply a philosophy of science that only assumes what is necessary to do science (i.e. methodological naturalism).

Some philosophies of science do assume ontological positions, so it's important to point out that behaviorism does not.

Noted, I would say that's not radical but the opposite, hence my confusion.

It's radical because it completely flipped the fundamental assumption of behaviorism at the time. Methodological behaviorism was based, more or less, on the claim that we cannot and should not study the inner states of organisms - Skinner overthrew this central assumption of behaviorism. You really can't get more radical than that.

Uhm, this was -I think- regarding Chomsky's view that behaviorism is empiricist, not a rationalism, therefor it doesn't study anything. I don't think anybody in the psychological field either knows or cares since they are doing a mixed approach anyway. If I get nasty, I would propose that they are doing a mixed approach since they have now conflated all terms beyond any meaning.

I'm still not quite sure what you're arguing though. So all of those studies on human behavior (in Chomsky's opinion) aren't real, or is the suggestion that they are meaningless or something?

(Personally, I can see the Freud vs Behaviorism as a debate on whether one is going to do qualitative vs quantitative analysis, and yeah, it stems to reason that regarding humans one needs to do both.)

Yeah scientists agree that both approaches are important.

I am not religious and I don't think religious people are more moral than other people. Sometimes, I would say even less so.

I agree, but my point was that the religious are recognised as strongly believing in the idea of humans being special, or "god's creatures" etc, and so if even they don't believe it's immoral to compare humans to animals, then I can't think of any moral system that could reasonably argue such a position.

I already explained my position: People are not microbes, conflating the terms leads to treating people as microbes, therefor I find it an immoral act to overly use the term organism when (partly) addressing people.

But it's just an accurate label. Do you object to humans falling under the term "animals" as well? Or "mammals"? "Apes"?

I don't accept that using accurate terminology to refer to humans and other organisms leads to us treating humans like microbes - and I don't even know exactly what that means. Is there a trend of people being attacked in the street by masked men spraying them down with antibacterials?

I would also propose that it is unscientific since people, from a systematic complexity view, cannot be explained with the same simple models as the behavior of rats can be. It therefor leads to a view on humans which is overly simplistic, which should be avoided.

I agree it's a personal opinion.

The last part isn't a personal opinion though, it's a comment on a scientific issue. Given that the same laws of behaviors apply equally to all organisms, it is no longer your "opinion" and is simply "wrong".

 

[MarcoD]

God man, when looking at it, I only see a (mostly) meaningless historical political debate with ill-defined questions, concepts, and -consequently- lousy answers. We'll disagree on this.

The last part isn't a personal opinion though, it's a comment on a scientific issue. Given that the same laws of behaviors apply equally to all organisms, it is no longer your "opinion" and is simply "wrong".

Tss. Define behavior and define the laws you would like to apply and we can have a discussion. I see no reason from your posts to believe that that statement is anywhere near the truth, but again, since we didn't define either behavior or laws, it's a pointless discussion anyway.

So I'll keep on believing what you think is "wrong," thank you.

 

[Mr.Samsa]

God man, when looking at it, I only see a meaningless historical political debate with ill-defined questions, concepts, and -consequently- lousy answers. We'll disagree on this.

Everything in the debate has been defined perfectly. If you're struggling to understand something, then just ask, but be aware that your ignorance of the issue does not mean that the terms are ill-defined.

Tss. Define behavior and define the laws you would like to apply and we can have a discussion. I see no reason from your posts to believe that that statement is anywhere near the truth, but again, since we didn't define either behavior or laws, it's a pointless discussion anyway.

So I'll keep on believing what you think is "wrong," thank you.

"Behavior" involves basically anything an organism does, and it can be defined more concretely when there's something in particular that we want to study. For example, let's take "choice behavior" - that is, the act of selecting one alternative over another. Herrnstein's simplistic matching law describes the proportional distribution of responses as being a function of the proportional distribution of reinforcement. All organisms currently tested, from humans, to monkeys, pigeons, rats, fruit flies, octopuses, slugs, etc respond in the same way to the matching law, and we can predict what the subject will choose will equal accuracy (regardless of what species we are testing). The matching law has been refined over the years, giving us Baum's generalised matching law and Davison's contingency discriminability model, and now we consistently predict the behavior of all organisms in controlled conditions.

The interesting part is that all behavior is essentially choice behavior, as whenever we perform one action we necessarily do so at the expense of some other action (as it's impossible to perform two incompatible behaviors at the same time). From here we can then predict a range of complex behaviors, including language, altruism, empathy, self-control, perception, etc etc.

 

[MarcoD]

Everything in the debate has been defined perfectly. If you're struggling to understand something, then just ask, but be aware that your ignorance of the issue does not mean that the terms are ill-defined.

Bull. A statement which in essence boils down to "we are doing science," which implies the "and therefor you are not," is purely politics and implies that even the question was not understood, defined wrong, and people were stuck in political games. (This is interesting historically to some, but sorry, not to me.)

"Behavior" involves basically anything an organism does, and it can be defined more concretely when there's something in particular that we want to study. For example, let's take "choice behavior" - that is, the act of selecting one alternative over another. Herrnstein's simplistic matching law describes the proportional distribution of responses as being a function of the proportional distribution of reinforcement. All organisms currently tested, from humans, to monkeys, pigeons, rats, fruit flies, octopuses, slugs, etc respond in the same way to the matching law, and we can predict what the subject will choose will equal accuracy (regardless of what species we are testing). The matching law has been refined over the years, giving us Baum's generalised matching law and Davison's contingency discriminability model, and now we consistently predict the behavior of all organisms in controlled conditions.

The interesting part is that all behavior is essentially choice behavior, as whenever we perform one action we necessarily do so at the expense of some other action (as it's impossible to perform two incompatible behaviors at the same time). From here we can then predict a range of complex behaviors, including language, altruism, empathy, self-control, perception, etc etc.

Mathematically, it cannot be true, so A) I wonder about the experiments which were performed (to make it true), and B) you didn't prove you can generalize a simple law to all human behavior (which you won't be able to since one can prove the opposite).

(I guess I should clear up my position. In CS, the behavior of a system is clearly defined as the, either discrete or continuous, events occurring on a time scale in response to other events. If people are anywhere near simple inference machines, their behavior cannot be understood or predicted (it is impossible to construct a model from observing behavior alone). This all derives from basic testing theory.

Which explains Chomsky's position too. (Formal) linguistics is the basis of logic, logic is the basis of CS, CS predicts that quantitative studies cannot explain complex systems, only qualitative studies can, and even then it is nearly impossible. So Chomsky's position in my view is a rather badly defined defense against quantitative studies in linguistics/psychology on the basis of philosophical arguments on the scientific method, whereas in my view it would have sufficed to use technical arguments from testing theory.)

(I guess I was rude on this one, but I have a personal thing with psychology.)

 

[apeiron]

The interesting part is that all behavior is essentially choice behavior, as whenever we perform one action we necessarily do so at the expense of some other action (as it's impossible to perform two incompatible behaviors at the same time). From here we can then predict a range of complex behaviors, including language, altruism, empathy, self-control, perception, etc etc.

At last something that I can recognise as a positive statement of a current agenda! Do you have any references in mind that go into this philosophic point in more detail?

I certainly agree with the essence of this view. It tightly ties what an organism does to what it knows, or expects. And thus it would be a way to synthesise the behaviourist and cogsci legacies. Both are about the same thing if both are about intelligent choice-making.

This has lots of implications as you say. Behaviourism becomes not just about what an organism does, but perhaps even more so, about the behaviours that are absent - what it is actively chosing not to do.

The problem of course is observing what does not happen. It is possible, but not so easy.

And also models of the architecture of choice are going to sound pretty cognitive. They are naturally the domain of information processing models it would appear.

But first, what do you consider as recent research that is attempting to account for complex behaviour in this way?

 

[Mr.Samsa]

Bull. A statement which in essence boils down to "we are doing science," which implies the "and therefor you are not," is purely politics and implies that even the question was not understood, defined wrong, and people were stuck in political games. (This is interesting historically to some, but sorry, not to me.)

I'm not sure where you are getting this from, but yes there are important criteria that need to be met in order to be said to be doing science - behaviorism says that psychology should attempt to meet these criteria. If an approach to psychology ignores some or all of these criteria, then they aren't doing "science" as properly defined.

This isn't necessarily a problem, as science isn't the arbiter of truth and some questions cannot even be addressed by science. But if we're discussing empirical questions, then a field has to be scientific in order to have something to say on the matter.

Mathematically, it cannot be true, so A) I wonder about the experiments which were performed (to make it true), and B) you didn't prove you can generalize a simple law to all human behavior (which you won't be able to since one can prove the opposite).

(I guess I should clear up my position. In CS, the behavior of a system is clearly defined as the, either discrete or continuous, events occurring on a time scale in response to other events. If people are anywhere near simple inference machines, their behavior cannot be understood or predicted (it is impossible to construct a model from observing behavior alone). This all derives from basic testing theory.

Given that we have evidence of human behavior being predicted, then I suggest there is something wrong with your math. And I never said that we could generalise a simple law to all human behavior in the absolute sense you're suggesting. We're discussing science, not omnipotence, so of course I can't make a claim like that. However, from the evidence we do have, this simply law can account for every volitional human behavior in every situation it has currently been tested in. It's of course possible that the areas we've tested so far are the extent of its application and every other area of human behavior falls outside its scope, but we won't know this until we find a disqualifying behavior. When we do, we'll adapt our mathematical laws and try again.

Which explains Chomsky's position too. (Formal) linguistics is the basis of logic, logic is the basis of CS, CS predicts that quantitative studies cannot explain complex systems, only qualitative studies can, and even then it is nearly impossible. So Chomsky's position in my view is a rather badly defined defense against quantitative studies in linguistics/psychology on the basis of philosophical arguments on the scientific method, whereas in my view it would have sufficed to use technical arguments from testing theory.)

This sounds like an extremely weird position. Can quantitative studies not explain weather systems, or the global climate? Systems don't get much more complex than that.

At last something that I can recognise as a positive statement of a current agenda! Do you have any references in mind that go into this philosophic point in more detail?

I'm not sure if I know of any that investigate it too thoroughly, as it seems to be more a self-evident fact of logic rather than anything that need to be supported by evidence or that can even be rejected.

This http://www.shapingbehavior.com/images/Matching_Law.pdf gives quite a simple overview of how different areas view this same basic principle, and makes some interesting points, I think.

Choice theory is really where a lot of the emphasis of the field is, and has been for a few decades now - whether people are directly looking into it (e.g. researchers like Baum and Davison), or indirectly needing it to continue their work (e.g. work on self-control and perception).

I certainly agree with the essence of this view. It tightly ties what an organism does to what it knows, or expects. And thus it would be a way to synthesise the behaviourist and cogsci legacies. Both are about the same thing if both are about intelligent choice-making.

This has lots of implications as you say. Behaviourism becomes not just about what an organism does, but perhaps even more so, about the behaviours that are absent - what it is actively chosing not to do.

The problem of course is observing what does not happen. It is possible, but not so easy.

Agreed. I think I mentioned somewhere that Herrnstein explicitly included a variable in his equations to account for "absent behaviors" known as "extraneous reinforcement". What he showed was that by accounting for some variables in the equation, like knowing what the reinforcement ratios are for some alternatives, and/or what the behavioral responses are for some alternatives, we can actually calculate how reinforcing "all other behaviors" are, and thus calculate how likely they are to perform those behaviors at any given point in time.

And also models of the architecture of choice are going to sound pretty cognitive. They are naturally the domain of information processing models it would appear.

But first, what do you consider as recent research that is attempting to account for complex behaviour in this way?

What do you mean by "in this way"? As in, looking at choice behavior in terms of information processing models?

Is this the sort of thing you're looking for: "http://neuroscience.cafe24.com/cgi-bin/ez2000/system/db/board_lab/upload/22/1091171761/1782.pdf"?

Or did you just mean studies using choice theory to explain complex behavior more generally? If so, there's the contingency discriminability model which accounts for issues in perception and signal detection, attention as being a function of the matching law, self-control, altruism and cooperation, and I'm sure there are more but I can't think of them off the top of my head. (Note: I know some of the links I've presented are old, and/or are animal studies not human studies, but the papers I've presented are the seminal papers in the area which explain the logic behind the concept. If you look them up in Google Scholar and click on "Citations", you should be able to find human replication studies. As far as I know, the concept has been demonstrated in humans for all areas I've listed above).

 

[bohm2]

Uhm, this was -I think- regarding Chomsky's view that behaviorism is empiricist, not a rationalism, therefor it doesn't study anything. I don't think anybody in the psychological field either knows or cares since they are doing a mixed approach anyway. If I get nasty, I would propose that they are doing a mixed approach since they have now conflated all terms beyond any meaning.

With respect to Chomsky's criticism of behaviourism, I came across some interesting quotes (not sure about validity/accuracy, however) from one author who is more supportive of Chomsky's position unlike the previous papers we provided:

Chomsky ([1959], pp. 251–2) presents a dilemma for Skinner. Skinner sets out to show that behaviour in general is lawful, i.e. it is under stimulus control. Of course, as our knowledge stands, we really have not a hope of describing such laws under the strict conditions under which the learning theoretic vocabulary is used in the lab. So, the behaviourist must either admit that, as things stand, behaviour can’t be understood as lawful or else he may restrict himself to those areas of behaviour that are lawful, such as the barpressing behaviour of trained rats. Either way, we have no reason to think that learning theory should replace folk psychology. Skinner, however, evades the dilemma by metaphorically extending the technical language of the laboratory to cover any piece of behaviour as required. Yet, this ‘metaphoric reading...is no more scientific than the traditional approaches to the subject matter [viz. folk psychology]’ (Chomsky [1959], p. 552). That is to say, the new vocabulary is merely a misleading paraphrase of familiar modes of description, and does not constitute an insight into the actual causal antecedents of behaviour, let alone an appropriate vocabulary for the natural kinds of human behaviour.

The author argues that Chomsky favours an approach to linguistics/cognitive science that relies on:

abstraction, idealization, conceptual creation, and the positing of unobservables. In other words, science does not have as its target a complete and coherent description of the world as we find it, the world as delineated by our given categories; instead, its aim is to seek highly abstract ‘hidden’ laws and mechanisms that unify otherwise heterogeneous phenomena, in light of which our given categories drop out, at best, as shallow and partial taxonomic artefacts.

Why?

Chomsky assumes...that folk psychology does not constitute an adequate basis for the explanation of behaviour, verbal or otherwise...Further, Chomsky’s denial of the claim that the behaviourist vocabulary is ‘scientific’ in comparison with the traditional vocabulary does not suggest that the latter is ‘scientific’; the clear implication is that neither is deserving of the epithet. Indeed, since behaviourism differs from ‘traditional mentalism’ ‘only’ in terms of relative obscurity, it would seem that technical mentalism marks a substantial departure from its traditional namesake. So, for reasons independent of the particular failure of behaviourism, folk psychology is not adequate for scientific employment. The failure of both is an instance of the putative failure of every theory of behaviour. The failure of folk psychology in particular follows from the thesis of M-SE (meta-scientific eliminativism): folk theories do not make for science. Of course, this meta-thesis is entirely independent of any claims concerning behaviourism in particular.

Meta-scientific Eliminativism: A Reconsideration of Chomsky’s Review of Skinner’s Verbal
Behavior
http://www.uea.ac.uk/phi/People/Academic/John+Collins#publications

 

[MarcoD]

This sounds like an extremely weird position. Can quantitative studies not explain weather systems, or the global climate? Systems don't get much more complex than that.

Ridiculous. Even for a thing as simple as Collatz's series (if a number is even, divide by two; otherwise, multiply by three and add one; repeat) we are clueless why a series always seems to end with one.

Reducing all systems to trivial systems just so that one can explain something whereas we're already clueless about the simplest of systems, nah, pseudoscience.

 

[MarcoD]

With respect to Chomsky's criticism of behaviourism, I came across some interesting quotes (not sure about validity/accuracy, however) from one author who is more supportive of Chomsky's position unlike the previous papers we provided.

It's interesting to note these discussions, but honestly, I am too simple minded for highbrow discussions. I find them meaningless unless terms follow strict simple rules, and I have the feeling that this particular discussion is just one of name calling. So thanks, but I'll refrain from further comments.

(Though I am now completely on Chomsky's side that (some) psychologists need a course in clear thinking.)

 

[ThomasT]

Lots of quantum experiments involve the counting of certain macroscopic qualitative instrumental behaviors . And this was Skinner's approach in attempting to make psychology into a science of the behavior of organisms.

Afaik, this approach was pretty successful. And it is a science.

Wrt Chomsky's argument that the "body" part of the mind-body problem is ill-defined ... what about the "mind" part, which seems to me to less well defined, and, prospectively less well definable, than the "body" part.

As far as I'm concerned, there's no mind-body problem. What we can manipulate, and experiment upon, and make testable statements about is what we call the physical world, ie., the "body" part of the mind-body problem.

 

[Mr.Samsa]

With respect to Chomsky's criticism of behaviourism, I came across some interesting quotes (not sure about validity/accuracy, however) from one author who is more supportive of Chomsky's position unlike the previous papers we provided:

Skinner sets out to show that behaviour in general is lawful, i.e. it is under stimulus control. Of course, as our knowledge stands, we really have not a hope of describing such laws under the strict conditions under which the learning theoretic vocabulary is used in the lab.

I don't understand this bit - on what grounds does the author make this claim?

Ridiculous. Even for a thing as simple as Collatz's series (if a number is even, divide by two; otherwise, multiply by three and add one; repeat) we are clueless why a series always seems to end with one.

Reducing all systems to trivial systems just so that one can explain something whereas we're already clueless about the simplest of systems, nah, pseudoscience.

I just don't understand your position. Let's make this simple: you sit down to dinner to chat with some friends that you've invited around. Behaviorists, using the matching law, are able to predict with significant accuracy who you will talk to, and how long you'll spend talking to each person.

Is this not predicting a human's behavior? Why is this scientific fact considered "impossible" by your mathematics?

 

[apeiron]

Choice theory is really where a lot of the emphasis of the field is, and has been for a few decades now - whether people are directly looking into it (e.g. researchers like Baum and Davison), or indirectly needing it to continue their work (e.g. work on self-control and perception).

Thanks for the references. The Bill Newsome one in particular does indeed show how a solid Behaviourist finding is now being investigated in terms of its cognitive and neural architecture. While the Seth paper shows the value of "as simple as possible" theories about cognitive processes.

You can see in this work the proper intersection of ideas. There is the empirical aspect - modelling in terms of observables. And this means not just measuring real life behaviour but measuring the behaviour of simulations - actual models of a system architecture.

There is then also the rationalist aspect - the search for general principles to inspire/justify the design of the architecture. The obvious one is an optimality principle. Or free energy minimisation. Herrnstein's matching law proved that optimisation is indeed being employed. And Newsome is investigating the simplest possible neural architectures that can instantiate the necessary "algorithm", doing this by testing the performance of a simulation against the real-world performance of monkeys.

So yes, in this kind of standard mainstream work, the empirical and the rational are being combined. I still see this more as cognitive science, because the idea of proving architectural claims through simulation (as in neural nets, AI and A-life) started out there. But then this is about the behaviour of simulations, so I guess it reflects just as much the basic ethos of Behaviourism.

Then the notion of "choice". Again I see this as an excellent anchoring idea.

People talk about all sorts of things as the focal concept when it comes to mind. Intention, autonomy, adaptive behaviour, prediction, awareness, cognition, processing, intelligence, etc. But "choice" gets right to the heart of things because it so clearly ties what the brain does to the world an organism lives in. You could call mind science the science of choice and that would seem to nail it better than anything.

Behaviourism and Cogsci are taken to focus on opposing extremes - acting and thinking (I realize you dispute this, but I am talking of generalised popular belief that has its grain of truth). And choice stands pretty neatly between acting and thinking, connecting them in a way that makes each meaningful, while respecting also the essential dichotomy.

Some choices take a lot of thinking. Some choices are just (almost mindless/unconscious) acting. A theory of choice would accommodate this spectrum, wrapping it into the one model.

 

[Pythagorean]

MarcoD, I feel you're kind of brow beating with mathematics. You don't actually make any real arguments, you just appeal to mathematics as an authority.

I use mathematics to study neural systems, and I'm not sure where you're coming from exactly. Behaviorism is more about the experimental side. We theoreticians come along and build models with mathematics that match the observations documented by experimentalists. We do recognize there is degeneracy in the system (read Eve Marders work, I can reference if you don't like digging but she has done more than one paper on the subject).

 

[bohm2]

Isn't MarcoD just arguing that human action/thought/cognition, etc. is not lawfully predictable?

 

[Mr.Samsa]

MarcoD, I feel you're kind of brow beating with mathematics. You don't actually make any real arguments, you just appeal to mathematics as an authority.

I use mathematics to study neural systems, and I'm not sure where you're coming from exactly. Behaviorism is more about the experimental side. We theoreticians come along and build models with mathematics that match the observations documented by experimentalists. We do recognize there is degeneracy in the system (read Eve Marders work, I can reference if you don't like digging but she has done more than one paper on the subject).

It seems like a strange line of argument to me and I can't quite understand how it can be valid. It sounds like the old claim that Aristotle believed that insects only have 4 legs due to his particular logical theory, when all he needed to do was catch a fly and count the legs (of course, that story is a myth and Aristotle didn't actually believe that, but it works for illustrative purposes ).

Instead of working out whether something is mathematically possible first, scientists have just gone out, developed models of human behavior, and applied them so that we can predict their behavior in given situations. Since we do this on a daily basis, across many individuals, cultures, situations, etc, then any theory which suggests it is impossible must be wrong. I'm not sure how anyone could argue otherwise.

 

[Pythagorean]

In addition to modeling at the neural level, you can also use operant matching methods. Here's some notes from a computational neuroscience course taught at MIT using the matching law:

http://ocw.mit.edu/courses/brain-an...nce-spring-2004/lecture-notes/lec6_match1.pdf

 

[Mr.Samsa]

In addition to modeling at the neural level, you can also use operant matching methods. Here's some notes from a computational neuroscience course taught at MIT using the matching law:

http://ocw.mit.edu/courses/brain-an...nce-spring-2004/lecture-notes/lec6_match1.pdf

Thanks for the link, Pythagorean. Out of interest, do you know if computational neuroscience has tried to model anything more complicated than strict matching?

 

[Pythagorean]

Thanks for the link, Pythagorean. Out of interest, do you know if computational neuroscience has tried to model anything more complicated than strict matching?

Yeah, for sure. The above was just a lecture for students. In general, modeling at the cellular network level (neural systems) is very complicated when relating it to behavior, this is what I'm most familiar with.

But there are more abstracted approaches; most I've heard from apeiron: Friston's Free Energy Principle of the Brain, bayesian inference networks.

One I found years ago was Mark Gluck's approach: