Total Posts:10|Showing Posts:1-10
Jump to topic:

Relation between connectionism & behaviourism

Diqiucun_Cunmin
Posts: 2,710
Add as Friend
Challenge to a Debate
Send a Message
10/2/2015 2:08:14 PM
Posted: 1 year ago
Not sure if putting this in science or philosophy will attract more interested members. I picked philosophy though.

Anyway, connectionism seems, prima facie, a strongly empiricist viewpoint, since it appears not to assume a priori knowledge of any kind, unlike symbolic models. It also appears to be behaviourist or at least a successor of the behaviourist tradition, since it places huge emphasis on the response of the neural network to stimuli - much like learning theory.

However, connectionism also seems to strive to be a plausible model of mind/brain at the neurological level. This seems to be different from behaviourism, which seems not to recognise the importance of the brain.
The thing is, I hate relativism. I hate relativism more than I hate everything else, excepting, maybe, fibreglass powerboats... What it overlooks, to put it briefly and crudely, is the fixed structure of human nature. - Jerry Fodor

Don't be a stat cynic:
http://www.debate.org...

Response to conservative views on deforestation:
http://www.debate.org...

Topics I'd like to debate (not debating ATM): http://tinyurl.com...
Diqiucun_Cunmin
Posts: 2,710
Add as Friend
Challenge to a Debate
Send a Message
10/2/2015 2:09:18 PM
Posted: 1 year ago
So, what do you think is the relationship between connectionism and behaviourism? Is connectionism necessarily behaviourist? Are they not as compatible as they seem at face value?
The thing is, I hate relativism. I hate relativism more than I hate everything else, excepting, maybe, fibreglass powerboats... What it overlooks, to put it briefly and crudely, is the fixed structure of human nature. - Jerry Fodor

Don't be a stat cynic:
http://www.debate.org...

Response to conservative views on deforestation:
http://www.debate.org...

Topics I'd like to debate (not debating ATM): http://tinyurl.com...
Devilry
Posts: 489
Add as Friend
Challenge to a Debate
Send a Message
10/2/2015 2:27:11 PM
Posted: 1 year ago
What about I'll-drink-with-you-forever-ism? (Well, besides when I'm banging fit birds.)
: : : At 11/15/2016 6:22:17 PM, Greyparrot wrote:
: That's not racism. Thats economics.
kp98
Posts: 729
Add as Friend
Challenge to a Debate
Send a Message
10/2/2015 2:55:16 PM
Posted: 1 year ago
Broadly, behaviourism concerns itself with only with what is observed. Behaviourists try to do without theorising about how a response (like pushing a lever) gets associated with a stimulus (such as a light flashing) - it merely notes that such an association is observed under certain stated conditions. Behaviourism is an 'anti-theory' of consciousness - it deliberately and assiduously avoids saying anything about consciousness either way, although it may provide data for the theoretically inclined to do so.

On the other hand connectionism is based on the idea that the appearent complexity of conscousness is due to the high degree of inter-conncectedness between the various parts of the brain, which including all sorts of feedback, feed-forward, self-referential strange loops and what-not.

Connectionism is one step along from admitting total ignorance as to how consciousness works - rather than saying 'I have no idea how consciousness works' one says 'Consciousness happens because the brain is so interconncted', but of course that leaves the question of what sort of connections produce consciousness and how much connectedness is needed to produce consciousness are left hanging!

Closely related is the idea that consciousness is 'emergent', a word that doesn't mean 'I haven't a clue how it works' but is used that way more often than not.
Diqiucun_Cunmin
Posts: 2,710
Add as Friend
Challenge to a Debate
Send a Message
10/2/2015 4:01:32 PM
Posted: 1 year ago
At 10/2/2015 2:55:16 PM, kp98 wrote:
Broadly, behaviourism concerns itself with only with what is observed. Behaviourists try to do without theorising about how a response (like pushing a lever) gets associated with a stimulus (such as a light flashing) - it merely notes that such an association is observed under certain stated conditions. Behaviourism is an 'anti-theory' of consciousness - it deliberately and assiduously avoids saying anything about consciousness either way, although it may provide data for the theoretically inclined to do so.
Do you believe that behaviourists' lack of interest in the mind and consciousness is purely methodological, then?
On the other hand connectionism is based on the idea that the appearent complexity of conscousness is due to the high degree of inter-conncectedness between the various parts of the brain, which including all sorts of feedback, feed-forward, self-referential strange loops and what-not.

Connectionism is one step along from admitting total ignorance as to how consciousness works - rather than saying 'I have no idea how consciousness works' one says 'Consciousness happens because the brain is so interconncted', but of course that leaves the question of what sort of connections produce consciousness and how much connectedness is needed to produce consciousness are left hanging!
LOL, though to be fair, plausible connectionist models have been proposed, at least in the field of linguistics (don't know about other fields).

(I do have doubts about connectionism since it seems to have little explanatory power for language phylogeny. I do believe the poverty of the stimulus argument has significant empirical support, and so I don't think language can be feasibly acquired by the learning model proposed by connnectionism.)
Closely related is the idea that consciousness is 'emergent', a word that doesn't mean 'I haven't a clue how it works' but is used that way more often than not.
LOL...
The thing is, I hate relativism. I hate relativism more than I hate everything else, excepting, maybe, fibreglass powerboats... What it overlooks, to put it briefly and crudely, is the fixed structure of human nature. - Jerry Fodor

Don't be a stat cynic:
http://www.debate.org...

Response to conservative views on deforestation:
http://www.debate.org...

Topics I'd like to debate (not debating ATM): http://tinyurl.com...
Diqiucun_Cunmin
Posts: 2,710
Add as Friend
Challenge to a Debate
Send a Message
10/2/2015 4:09:26 PM
Posted: 1 year ago
At 10/2/2015 4:01:32 PM, Diqiucun_Cunmin wrote:
At 10/2/2015 2:55:16 PM, kp98 wrote:
Broadly, behaviourism concerns itself with only with what is observed. Behaviourists try to do without theorising about how a response (like pushing a lever) gets associated with a stimulus (such as a light flashing) - it merely notes that such an association is observed under certain stated conditions. Behaviourism is an 'anti-theory' of consciousness - it deliberately and assiduously avoids saying anything about consciousness either way, although it may provide data for the theoretically inclined to do so.
Do you believe that behaviourists' lack of interest in the mind and consciousness is purely methodological, then?
On the other hand connectionism is based on the idea that the appearent complexity of conscousness is due to the high degree of inter-conncectedness between the various parts of the brain, which including all sorts of feedback, feed-forward, self-referential strange loops and what-not.

Connectionism is one step along from admitting total ignorance as to how consciousness works - rather than saying 'I have no idea how consciousness works' one says 'Consciousness happens because the brain is so interconncted', but of course that leaves the question of what sort of connections produce consciousness and how much connectedness is needed to produce consciousness are left hanging!
LOL, though to be fair, plausible connectionist models have been proposed, at least in the field of linguistics (don't know about other fields).

(I do have doubts about connectionism since it seems to have little explanatory power for language ontogeny. I do believe the poverty of the stimulus argument has significant empirical support, and so I don't think language can be feasibly acquired by the learning model proposed by connnectionism.)
Closely related is the idea that consciousness is 'emergent', a word that doesn't mean 'I haven't a clue how it works' but is used that way more often than not.
LOL...

Fixed
The thing is, I hate relativism. I hate relativism more than I hate everything else, excepting, maybe, fibreglass powerboats... What it overlooks, to put it briefly and crudely, is the fixed structure of human nature. - Jerry Fodor

Don't be a stat cynic:
http://www.debate.org...

Response to conservative views on deforestation:
http://www.debate.org...

Topics I'd like to debate (not debating ATM): http://tinyurl.com...
kp98
Posts: 729
Add as Friend
Challenge to a Debate
Send a Message
10/2/2015 5:51:32 PM
Posted: 1 year ago
I would agree that behaviourism is principally a methodological guideline. Given the difficulty of forming a convincing theory of 'mental phenomena' gathering data without premature over-interpretion behaviourism is perhaps not a bad idea.

I understand a bit better where you are coming from now. If we want to know how children learn stuff (such as language) we can't really dissect their brains, or stick probes in their skulls (if we did we wouldn't really know what we were looking for). So we have to try to infer how they learn language purely from observation - ie from watching their external behaviour, in particular what they say.

Getting the best data is surely done best if one does not assume a model, ie taking a 'behaviourist' stance. It is then a question of whether a connectionist theory or, say, a symbol-processing model best explains what was observed.

It's not my field so I don't know what models are fashionable amongst linguistics experts, but my hunch is that a model works by combining - or connecting - a number of otherwise independent functional units is more likely to be near the truth how brains actually work.

That said, it is one thing to invent a model that seems to produce the same results as the human brain and a different thing to show the brain actually works that way. A behaviourist shouldn't care as long as the behaviour is the same, but I don't think anybody is that much of a behaviourist in practice.
Diqiucun_Cunmin
Posts: 2,710
Add as Friend
Challenge to a Debate
Send a Message
10/3/2015 1:59:09 AM
Posted: 1 year ago
At 10/2/2015 5:51:32 PM, kp98 wrote:
I would agree that behaviourism is principally a methodological guideline. Given the difficulty of forming a convincing theory of 'mental phenomena' gathering data without premature over-interpretion behaviourism is perhaps not a bad idea.

I understand a bit better where you are coming from now. If we want to know how children learn stuff (such as language) we can't really dissect their brains, or stick probes in their skulls (if we did we wouldn't really know what we were looking for). So we have to try to infer how they learn language purely from observation - ie from watching their external behaviour, in particular what they say.

Getting the best data is surely done best if one does not assume a model, ie taking a 'behaviourist' stance. It is then a question of whether a connectionist theory or, say, a symbol-processing model best explains what was observed.
To clarify, I don't think behaviourism is merely a methodological precept - that's just what I understood from your post. I think it embraces a strongly empiricist doctrine, assuming that the mind is tabula rasa and is shaped by external stimuli. The contribution of a priori knowledge in the mind/brain is negligible. At least that's the impression I've got - I haven't read any first-hand sources from behaviourists, so it could well be that I was deceived by rationalists, lol.
It's not my field so I don't know what models are fashionable amongst linguistics experts, but my hunch is that a model works by combining - or connecting - a number of otherwise independent functional units is more likely to be near the truth how brains actually work.
The current Chomskyan orthodoxy in linguistics stresses modularity over connectionism as well. The notion of the language faculty is particularly apparent in the works of Chomsky, Pinker and others (although they disagree on finer details). In fact, much of the symbolicism in cognition (such as classical AI) were initially inspired by Chomsky's work.

Judging by linguistic and aphasiological evidence, I think I'm inclined to agree with the orthodox stance in linguistics as well. Connectionism, as I've said above, seems improbable in light of stimulus poverty phenomena.
That said, it is one thing to invent a model that seems to produce the same results as the human brain and a different thing to show the brain actually works that way. A behaviourist shouldn't care as long as the behaviour is the same, but I don't think anybody is that much of a behaviourist in practice.
In Chomskyan linguistics, this is the difference between descriptive adequacy and explanatory adequacy. The ultimate goal is of course to reach explanatory adequacy :) and that requires the synergy of linguistics with other fields, like psychology, neuroscience and so on.
The thing is, I hate relativism. I hate relativism more than I hate everything else, excepting, maybe, fibreglass powerboats... What it overlooks, to put it briefly and crudely, is the fixed structure of human nature. - Jerry Fodor

Don't be a stat cynic:
http://www.debate.org...

Response to conservative views on deforestation:
http://www.debate.org...

Topics I'd like to debate (not debating ATM): http://tinyurl.com...
Diqiucun_Cunmin
Posts: 2,710
Add as Friend
Challenge to a Debate
Send a Message
10/3/2015 2:20:01 AM
Posted: 1 year ago
At 10/3/2015 1:59:09 AM, Diqiucun_Cunmin wrote:
At 10/2/2015 5:51:32 PM, kp98 wrote:
I would agree that behaviourism is principally a methodological guideline. Given the difficulty of forming a convincing theory of 'mental phenomena' gathering data without premature over-interpretion behaviourism is perhaps not a bad idea.

I understand a bit better where you are coming from now. If we want to know how children learn stuff (such as language) we can't really dissect their brains, or stick probes in their skulls (if we did we wouldn't really know what we were looking for). So we have to try to infer how they learn language purely from observation - ie from watching their external behaviour, in particular what they say.

Getting the best data is surely done best if one does not assume a model, ie taking a 'behaviourist' stance. It is then a question of whether a connectionist theory or, say, a symbol-processing model best explains what was observed.
To clarify, I don't think behaviourism is merely a methodological precept - that's just what I understood from your post. I think it embraces a strongly empiricist doctrine, assuming that the mind is tabula rasa and is shaped by external stimuli. The contribution of a priori knowledge in the mind/brain is negligible. At least that's the impression I've got - I haven't read any first-hand sources from behaviourists, so it could well be that I was deceived by rationalists, lol.
Just to add that this is the reason I drew the link between connectionism and behaviourism. Connectionism favours an empiricist model of learning.
It's not my field so I don't know what models are fashionable amongst linguistics experts, but my hunch is that a model works by combining - or connecting - a number of otherwise independent functional units is more likely to be near the truth how brains actually work.
The current Chomskyan orthodoxy in linguistics stresses modularity over connectionism as well. The notion of the language faculty is particularly apparent in the works of Chomsky, Pinker and others (although they disagree on finer details). In fact, much of the symbolicism in cognition (such as classical AI) were initially inspired by Chomsky's work.

Judging by linguistic and aphasiological evidence, I think I'm inclined to agree with the orthodox stance in linguistics as well. Connectionism, as I've said above, seems improbable in light of stimulus poverty phenomena.
That said, it is one thing to invent a model that seems to produce the same results as the human brain and a different thing to show the brain actually works that way. A behaviourist shouldn't care as long as the behaviour is the same, but I don't think anybody is that much of a behaviourist in practice.
In Chomskyan linguistics, this is the difference between descriptive adequacy and explanatory adequacy. The ultimate goal is of course to reach explanatory adequacy :) and that requires the synergy of linguistics with other fields, like psychology, neuroscience and so on.
The thing is, I hate relativism. I hate relativism more than I hate everything else, excepting, maybe, fibreglass powerboats... What it overlooks, to put it briefly and crudely, is the fixed structure of human nature. - Jerry Fodor

Don't be a stat cynic:
http://www.debate.org...

Response to conservative views on deforestation:
http://www.debate.org...

Topics I'd like to debate (not debating ATM): http://tinyurl.com...
n7
Posts: 1,360
Add as Friend
Challenge to a Debate
Send a Message
10/3/2015 5:08:01 AM
Posted: 1 year ago
Connectionism doesn't seem to be behaviorist. Behaviorism states a mental state is simply a behavioral state. Anxiety IS having an increased heart rate, trembles, and breathing faster. A connectionist might state anxiety is your neural network being processing in a certain way. You might be able to control your breathing, heart rate, and trembles, but you can still feel anxiety.
404 coherent debate topic not found. Please restart the debate with clear resolution.


Uphold Marxist-Leninist-Maoist-Sargonist-n7ism.