Science vs. Pseudoscience

[Altherion]

The root of the word "science" means simply "knowledge" so pretty much anything can claim to be science, but our current conception of the word is knowledge that we can use to make accurate, verifiable predictions about the world. Unfortunately, not everything that claims to be scientific actually falls under that definition: there is a large variety of people ranging from quacks on the internet to professors in academia who do something that, on the surface, looks scientific, but in fact is nothing of the sort. We had a debate in the politics thread about the validity of certain results and I wanted to point out some general criteria for spotting pseudoscience (this can be difficult when it hides behind jargon and math).

First, any empirical result must either be based on a large, representative sample size or a prior probability distribution with a good explanation of why it makes sense in your case. This is a general requirement across fields because regardless of what it is you are measuring, you must eventually quantify your prediction in terms of probability and the two things I mentioned are the respective requirements for the frequentist and Bayesian varieties of the latter. Most people can come up with a large sample, but proving that it is representative is usually pretty hard. Physics and chemistry have it easy here because they mostly deal with things that are well and truly identical (this can be proven!), but this is a minefield for the social sciences.

Second, the subject matter of a result must either be proven to be independent of time or its evolution with time must be fully understood. A snapshot can be useful for understanding such evolution (as part of a series), but a single one should never be used to predict something in the future. Again, natural laws do not change with time (people have tested this pretty extensively), but some things about human beings do (and some do not).

Finally, a result must be verifiable and actually verified by means other than simply reproducing the original study (although for some disciplines, even the latter appears to be a challenge). The scientific method and even mathematics itself (which underlies all scientific results) have rather famous philosophical problems. The reason we trust them despite this is that they work: everything from cold medicine to washing machines to roads to smartphones is build on top of the natural and formal sciences. This is also true for some aspects of the social sciences... but not all.

Note that these rule out, for example, much of macroeconomics. It deals with large and diverse groups which makes getting a representative sample difficult, it is time dependent in an unpredictable way and it is almost impossible to test in practice. This doesn't mean that studying it is worthless -- as I mentioned above, with enough snapshots patterns start to emerge -- but it is very dangerous to make predictions based on its results. It's closest in nature to something like an almanac (which had a combination of common sense, genuine data such as the rising and setting times of the Sun... but also weather forecasts made months in advance). Despite the use of enough math to make it incomprehensible to most of the population and enough jargon to do so for most of the rest, it's not science so I see no reason for giving it any deference... and it's far from being the worst offender.

[Rippounet]

Thanks for taking the time.

There was such a debate before actually (can't remember where). It's true that using a strict definition of science "humanities" could be called "pseudo-sciences."

But you're using this for a very specific purpose. This was your argument:

Quote

More and more people believe that they're the ones who are disadvantaged and it is not obvious who is right. Academia has its position, but fewer and fewer people trust them anymore. I personally do not -- the people who do studies on these subjects sometimes can't even do math and always can't do science (i.e. this level of quality would never be accepted in any discipline where the results are expected to be used for something that must work).

You've been arguing for some time now that minorities are no longer disadvantaged in the US. But the studies on the matter don't use "samples," they try to use data about the entire population of the US, based for example on figures from the U.S. Census Bureau, the Bureau of Labor... etc.

So, semantics aside, what do you base your claim on? Or, to be more specific, why would the conclusions of the many studies on the topic be invalidated?

[OldGimletEye]

3 hours ago, Altherion said:

Note that these rule out, for example, much of macroeconomics. It deals with large and diverse groups which makes getting a representative sample difficult, it is time dependent in an unpredictable way and it is almost impossible to test in practice. This doesn't mean that studying it is worthless -- as I mentioned above, with enough snapshots patterns start to emerge -- but it is very dangerous to make predictions based on its results. 

Back around 2010 John Cochrane and the whole Robert Lucas posse at U of Chicago, along with many others, warned that the Federal Reserve expanding the monetary base by 5 times would produce rampant inflation. Their Keynesian opponents basically said, "meh don't worry about it." And the reason basically given by the Keynesians was that at low interest rates money and bonds become highly subsitutable and accordingly wouldn't cause inflation to go out of control. In other words the velocity of money was hardly stable and sensitive to the interest rate. It would appear the Keynesians made the right call on this matter, right in line with their models.

To some extent you are right that we don't have lots of macroeconomic data. It would be nice if we had aggregate data going back to 4000 BC, but we just don't have it. So to certain extent you are right about this matter and that certainly makes macro modelling and testing challenging. But the issues aren't are just about lack of macro data as there are issues with selecting an appropriate identification procedure. If you are wiling to assume a certain identification strategy then there is enough data to estimate the fiscal multiplier for instance..

You can do economics the way say Ludwig von Mises did it and begin with a few axioms and then claim from that you can deduce all the laws of economics from that or maybe you can be bit like Milton Friedman and basically say, so long as an economic model makes successful predictions the internals of the model doesn't matter.

I don't think either is the right approach. You need both formal models and data. Of course the models have make sense and if they don't make sense than they need to be disregarded. And of course, they need to be rejected if the data clearly rejects their predictions. Or they probably need to be disregarded if they can't seemingly explain major historical events.

One of the reasons I found the whole Robert Lucas RBC models utterly ridiculous is because of their claim that all markets clear instantaneously, allegedly because there is some guy that sums up all the demands and supplies and then sets prices, to clear them. And then the claim, that people always have correct price expectations and financial markets are complete. All these claims seem just bit too fantastic in the real world. So basically I eliminate that model because some of it's internals are pretty absurd. And then of course, it can't explain real world events like the massive unemployment during the Great Depression, nor does it account from the real world experience that money matters and is not neutral with regard to real variables.

So I reject the idea that macroeconomics is an exercise in anything goes. It's true of course that there isn't a lot of aggregate macroeconomic data out there, or as much as we would like, as it's only been fairly recent that governments have started to compile and collect it. However to some extent we can use microeconomics data, of which is in greater supply, to help build the internals of any macroeconomic model.

Also it is true that empirical economics (and not just macro) isn't like the hard sciences because largely the data is observational and a variety of issues can come up with the estimation of any model. In fact, in a US politics thread, I brought the issue of simultaneity and it's the reason why empirical economist look for natural or quasi natural experiments because those things help to give an exogenous event to identify an econometric model. And as I said in that post, it would be wise to not just look at one empirical paper but many in order to arrive at a conclusion. And of course asking whether the model makes sense.

I don't deny that getting at the truth in something like macro (and economics in general) is harder than say chemistry and physics where experiments can be controlled in the lab. Still, it's not exercise in "anything goes" if that is what you are trying to get at.

3 hours ago, Altherion said:

It deals with large and diverse groups which makes getting a representative sample difficult,

What if the population under study is GDP data or something like the monetary base. So long as the the data is stationary (or if the variables are co-integrated), I don't get your point here.

[Altherion]

1 hour ago, Rippounet said:

You've been arguing for some time now that minorities are no longer disadvantaged in the US. But the studies on the matter don't use "samples," they try to use data about the entire population of the US, based for example on figures from the U.S. Census Bureau, the Bureau of Labor... etc.

So, semantics aside, what do you base your claim on? Or, to be more specific, why would the conclusions of the many studies on the topic be invalidated?

The US Census Bureau and the rest do not actually survey the entire population of the US for economic statistics. The census has a form which is supposed to be filled out by everyone, but it is very short and while it does ask about race, there is little there to provide a picture of who is advantaged and who is disadvantaged. They used to have a longer form which would be sent to a subset of the population and they carry out a variety of other surveys, but all of this involves the same business with samples.

All of that said, I don't really have a quarrel with the Census Bureau or other government agencies (e.g. the Federal Reserve) which publish macroeconomic data. As I mentioned above, great care should be taken when using these snapshots for predictions, but they're better than nothing. However, there is nothing in this data that indicates who is disadvantaged and who is not -- they merely present a state of the system and not how the state was arrived at. The fact that at a given time, one group has more money than another says nothing about which one is disadvantaged -- it's entirely possible that, for example, the better off group works harder (on average) and/or has values better aligned with what it takes to make money right now (e.g. education). You need to tack things on to that data which do not follow from it in order to make the case about certain groups being disadvantaged and this is what people object to.

[OldGimletEye]

9 hours ago, Altherion said:

However, there is nothing in this data that indicates who is disadvantaged and who is not -- they merely present a state of the system and not how the state was arrived at.

Yeah, you might want to ask these guys about that:

https://voxeu.org/article/race-and-economic-opportunity-united-states

9 hours ago, Altherion said:

they merely present a state of the system and not how the state was arrived at. 

So we just have no flamin' clue how one group may have become disadvantaged? None at all? Nothing in US history that might just tip you off?

Really?

Seriously?

Face palm.

No make that a double face palm.

9 hours ago, Altherion said:

The fact that at a given time, one group has more money than another says nothing about which one is disadvantaged -- it's entirely possible that, for example, the better off group works harder (on average) 

So white people just work harder?

[Darth Richard II]

What the actual fuck?

[The Mance]

40 minutes ago, OldGimletEye said:

So white people just work harder

No no, its just that their values including having parents that can afford to pay for their education.

[Jo498]

The idealized model proposed is not fulfilled by a lot of "hard sciences" as for example geology and biology are far less general than most physics or chemistry and often deal with "single systems", say the ecology of one particular island or the anatomy of one species. (And as I seem to recall from the 1990s even hard physics often dealt with very "poor statistics", I think the top quark or one of the more exotic things back then orginally was discovered on a very narrow data basis of at first only a handful of "events".)

Anyway, the problem with economics as well as lots of sociology, psychology, nutrition science, educational science, even plenty of medicine (many fields that have done poorly in reproduction studies) is mainly that we have extremely complex systems that are very hard (or unethical) to screen from outside influences. (There are also new feedback loops. Chemical systems do not care about chemistry. But economical systems use economical theories to organize themselves.)

So often we can only do so much and the alternative would be not to do any inquiry into these fields at all which usually would be worse (maybe not always, though). It is true that often the mantle of science is used to present rather uncertain, restricted and ungeneralizable results as "scientific" but overall I'd say one could classify even the more problematic of these sciences more charitably as "protoscience" (like 18th century chemistry or so). They will maybe never leave that stage, it is simply naive to expect the universe to be such that all systems at all scales can be inquired into with methods that worked extremely well in physics and chemistry. So skepticism is certainly called for. Because of their immense societal influence and power I am more worried about this problem in medicine and economics than in sociology, though.

As for (most of) the humanities, they usually are not pseudoscience because their methods are entirely different, they deal with a different kind of "data" e.g. history and we (should) trust them to have developed appropriate methods that are simply different from natural sciences.

[Seli]

zeroth: All fields of science are interconnected. They all support each other and while steep simplifications will have to be made for any specific subfield to create their models and understanding in the end they cannot contradict each other.

The problem is apparent contradictions can either mean flawed understanding, bad models or bad faith arguments. And it is not always easy to distinguish between them.

9 hours ago, Altherion said:

...

All of that said, I don't really have a quarrel with the Census Bureau or other government agencies (e.g. the Federal Reserve) which publish macroeconomic data. As I mentioned above, great care should be taken when using these snapshots for predictions, but they're better than nothing. However, there is nothing in this data that indicates who is disadvantaged and who is not -- they merely present a state of the system and not how the state was arrived at. The fact that at a given time, one group has more money than another says nothing about which one is disadvantaged -- it's entirely possible that, for example, the better off group works harder (on average) and/or has values better aligned with what it takes to make money right now (e.g. education). You need to tack things on to that data which do not follow from it in order to make the case about certain groups being disadvantaged and this is what people object to.

And here that zeroth principle comes in quick and hard. Luckily we have dozens of other subsets that explain why differences exist. And even multiple political thought models of how (or even if it is necessary) to fix them.

[OldGimletEye]

14 hours ago, Altherion said:

Second, the subject matter of a result must either be proven to be independent of time or its evolution with time must be fully understood.

Well no, not exactly. People work with auto-correlated data all the time. The issue is weather those time series are stationary. That is why there are things like unit root test, the Durbin Watson statistic, the Johansen test and so forth.

[OldGimletEye]

14 hours ago, Rippounet said:

So, semantics aside, what do you base your claim on? Or, to be more specific, why would the conclusions of the many studies on the topic be invalidated?

Personally, what I would like to know is what his precise issue or issues are with "identity politics".
Preferably, explained clearly and succinctly as possible without reams of bullshit thrown in.

[Chaircat Meow]

A very serious case occurs in biology and pertains to the theory of evolution through natural selection. This is accepted as fact by the overwhelmingly majority of the cognitive elite throughout the western world but is  incoherent. That is to say to state ‘I justifiably believe in evolution via natural selection,’ or ‘the evidence supports evolution via natural selection’ is like uttering the statement ‘no statement I utter is true.’ Evolution via natural selection might be true but no one could rationally believe it, so its plausibly a kind of pseudo-science (not methodologically but in terms of outcome).

The reason for this is that the probability we will have reliable cognitive faculties (i.e. truth finding mechanisms in our minds) is very low if we suppose evolution via natural selection. That is P(R I E&NS) is low, where R is having reliable faculties. This is so because it seems evolution via natural selection would not select for true beliefs but only the neural structures advantageous for survival.

If we take a bacterium or something like that, the actions that lead to its survival and flourishing will be purely physical and mechanical, the firing of certain neurons (or whatever bacteria have). The same goes for frogs, snakes, bats and so on. There will be a decision-making structure or brain structure (in the higher organisms) that responds to certain stimuli and generates decisions that lead to survival or death and extinction.

On evolution and natural selection that is all a decision-making process will be, a physical structure (brain structure in the higher organisms) that all things being equal leads to certain actions, and those actions will determine the likelihood of the organism’s survival. There is no reason to suppose that a true belief goes along with the neural structure, given it’s only the neural structure doing the determining which leads to the surviving.

This seems very plausible when we consider that things like ants, beetles and snakes probably don’t have anything like beliefs anyway. Once we reach organisms that do have beliefs about things we have no reason to imagine that true belief suddenly starts affecting survival chances, as it did not do that in other organisms (as they did not have beliefs). Therefore, if we suppose evolution and natural selection is true we have good grounds for thinking any and all beliefs generated by the belief-causing structure it provides are likely to be false, or as likely to be false as they are to be true, which includes the belief in evolution via natural selection.

For a fuller discussion see Alvin Plantinga - Where the Conflict Really Lies: Science, Religion, and Naturalism

As this theory is taught in schools as scientific and can’t be sensibly believed it is to be hoped President Trump will do something to support those suggesting alternative theories, such as religion.

It is a phoney excuse to say that this latter is not a scientific theory, because while true, evolution and natural selection as taught in schools contribute to forming a metaphysical picture of the universe that can’t be believed in addition to teaching about genuine scientific methods and discoveries. Until these functions of current teaching are separated people will continue to learn nonsense and it seems the best way to separate the actual scientific discoveries from the junk will be to propose an alternative metaphysical model that complements the past and current biological research.

[Altherion]

13 hours ago, OldGimletEye said:

Back around 2010 John Cochrane and the whole Robert Lucas posse at U of Chicago, along with many others, warned that the Federal Reserve expanding the monetary base by 5 times would produce rampant inflation. Their Keynesian opponents basically said, "meh don't worry about it." And the reason basically given by the Keynesians was that at low interest rates money and bonds become highly subsitutable and accordingly wouldn't cause inflation to go out of control. In other words the velocity of money was hardly stable and sensitive to the interest rate. It would appear the Keynesians made the right call on this matter, right in line with their models.

It's usually the case that some group is right. In fact, the same group can be right for quite a while... right up to the point when they're not.

13 hours ago, OldGimletEye said:

I don't deny that getting at the truth in something like macro (and economics in general) is harder than say chemistry and physics where experiments can be controlled in the lab. Still, it's not exercise in "anything goes" if that is what you are trying to get at.

It's not an exercise in "anything goes" -- you can certainly get things right and wrong and occasionally even have these things borne out with time. My point was that it cannot make reliable predictions in the same way as the natural sciences do. What will be the year of the next recession? What will be the unemployment rate (U3) in July 2019? Or the inflation rate? People spend a lot of effort answering questions like these (you could make a lot of money if you knew the answers), but the best we can do are those stochastic models that you refer to... and if that wasn't bad enough, sometimes (when there's an all-out war or major recession), they're not even close.

14 hours ago, OldGimletEye said:

What if the population under study is GDP data or something like the monetary base. So long as the the data is stationary (or if the variables are co-integrated), I don't get your point here.

Sure, you can study the GDP data or many other parameters -- but what do you intend to do with them?

13 hours ago, OldGimletEye said:

Yeah, you might want to ask these guys about that:

https://voxeu.org/article/race-and-economic-opportunity-united-states

This is a perfect example of pseudoscience. They have a large set of data -- so far, so good. But then what? The rest is a set of conjectures quite similar to what you might find on these forums. There is no rigorous chain of arguments; nothing that would lead one to say "it must be this way and no other" -- it is trivial to come up with an alternative interpretation of their data that is just as plausible. And then they follow it up with a vague set of recommendations.

4 hours ago, OldGimletEye said:

Well no, not exactly. People work with auto-correlated data all the time. The issue is weather those time series are stationary. That is why there are things like unit root test, the Durbin Watson statistic, the Johansen test and so forth.

Determining that a process is stationary is part of the way toward figuring out its time dependence. If the stochastic fluctuations are sufficiently small, it might be enough to qualify as fully understanding it... assuming that it actually remains stationary over long time scales and not just until certain people start hanging certain other people from lampposts or something of the sort.

[OldGimletEye]

55 minutes ago, Altherion said:

It's usually the case that some group is right. In fact, the same group can be right for quite a while... right up to the point when they're not.

Question is why are they seemingly right and then what is the reason why they end up being wrong?

What is your explanation for this? Are they right just because of luck?

And if they end up being very wrong should they revise the model they are using or should they say "oh no worries. things might change back making us right again".

Should the rampant inflation around the corner crowd revise their beliefs? Or how about the skills gap crowd? Should they just continue believing what they believe because things might in the future change making their beliefs correct?

55 minutes ago, Altherion said:

It's not an exercise in "anything goes" -- you can certainly get things right and wrong and occasionally even have these things borne out with time. My point was that it cannot make reliable predictions in the same way as the natural sciences do. What will be the year of the next recession? What will be the unemployment rate (U3) in July 2019? Or the inflation rate? People spend a lot of effort answering questions like these (you could make a lot of money if you knew the answers), but the best we can do are those stochastic models that you refer to... and if that wasn't bad enough, sometimes (when there's an all-out war or major recession), they're not even close.

Certainly it is very difficult to predict when the next recession will happen or what unemployment will be from 2 years. However, that is not the same thing as saying no predictions can't be made at all. See the example I gave in my first post.

Certainly I don't know exactly what the U3 rate will be in July 2019. But that isn't being unscientific I don't think. That is merely being upfront that our current state of knowledge doesn't allow us to make those kind of predictions precisely.

55 minutes ago, Altherion said:

Sure, you can study the GDP data or many other parameters -- but what do you intend to do with them?.

1. To see if Republican Party claims of "superior business friendly supply growth" has any empirical bases at all.

2. To get an idea of what the fiscal multiplier might be (assuming of course a particular identification strategy).

3. To see how monetary policy affects growth.

4. To test whether recessions can cause Hysteresis effects.

Just a few examples.

55 minutes ago, Altherion said:

This is a perfect example of pseudoscience. They have a large set of data -- so far, so good. But then what? The rest is a set of conjectures quite similar to what you might find on these forums. There is no rigorous chain of arguments; nothing that would lead one to say "it must be this way and no other" -- it is trivial to come up with an alternative interpretation of their data that is just as plausible. And then they follow it up with a vague set of recommendations.

No this is pretty much horseshit by you. The authors explain their empirical strategy in their paper.

Here is a copy of it.

http://www.equality-of-opportunity.org/assets/documents/race_paper.pdf

And if it so trivial to come up with an alternative explanation, then do it.

And now that you have taken such a keen interest in "rigorous arguments" then puuuleeze, for the love god, give us your "rigorous argument" why "identity politics" is so bad. Despite constantly complaining about it, I've yet to see you put your objections to it in a forthright and comprehensible manner.

Quote

And then they follow it up with a vague set of recommendations.

Oh the irony in this line. I've asked you several times what you'd do about neo-liberalism and have never gotten a clear response, other than some vague rant about "identity politics".

55 minutes ago, Altherion said:

Determining that a process is stationary is part of the way toward figuring out its time dependence. If the stochastic fluctuations are sufficiently small, it might be enough to qualify as fully understanding it... assuming that it actually remains stationary over long time scales and not just until certain people start hanging certain other people from lampposts or something of the sort.

Well actually stochastic fluctuations can be quite large and the time series can be stationary so long as it's means and variances don't change.

I don't disagree that such time series can have structural breaks and so forth. But even that can give us useful information about institutional changes and so forth.

[TrueMetis]

1 hour ago, Chaircat Meow said:

This seems very plausible when we consider that things like ants, beetles and snakes probably don’t have anything like beliefs anyway. Once we reach organisms that do have beliefs about things we have no reason to imagine that true belief suddenly starts affecting survival chances, as it did not do that in other organisms (as they did not have beliefs). Therefore, if we suppose evolution and natural selection is true we have good grounds for thinking any and all beliefs generated by the belief-causing structure it provides are likely to be false, or as likely to be false as they are to be true, which includes the belief in evolution via natural selection.

A person who believes they can fly by flapping their arms is going to have their survival chances affected, a person who believes they can survive snakebites because of their faith is going to have their survival chances affected. There is every reason to believe that evolution via natural selection does bring about reliable* cognitive because our ability to make successful predictions is second only to, and deeply intertwined with, our ability to work together in abilities that allow us to live.

Untrue beliefs that we hold anyway are the sickle cell anemia's of belief, they typically don't have very negative effects, usually because they're beliefs about things that are irrelevant to daily life anyway, and have a mass benefit in keeping the tribe united. At least that was true back when god was the local mountain and the most that happened was the ground shook when he was pissed off. Now when there's a bunch of different gods that can talk to you and everyone and their mother is willing to claim their interpretation of an old book is correct the negatives are much larger and serve to divide more than unite. But that's both a recent development and is being superseded by other forces anyway. So natural selection never had much of a chance to act upon those kinds of beliefs anyway.

 

Evolution is pseudoscience only to the breathtakingly ignorant. If you cannot imagine a way for true beliefs to affect survival chances your imagination is stunted.

 

*At least generally and at the scale we live in.

[Seli]

1 hour ago, Chaircat Meow said:

...

The reason for this is that the probability we will have reliable cognitive faculties (i.e. truth finding mechanisms in our minds) is very low if we suppose evolution via natural selection. That is P(R I E&NS) is low, where R is having reliable faculties. This is so because it seems evolution via natural selection would not select for true beliefs but only the neural structures advantageous for survival.

Luckily we know that our brains and our mind do not offer reliable cognitive faculties, so this is not a problem at all. Of course the whole point of selection is that very low probability situations can still be reached by accident due to that pressure.

 

1 hour ago, Chaircat Meow said:

...

If we take a bacterium or something like that, the actions that lead to its survival and flourishing will be purely physical and mechanical, the firing of certain neurons (or whatever bacteria have). ..

Proteins and other chemicals moving inside them, bacteria are single celled organisms so don't have specialized cells such as neurons. Doesn't stop them from reacting to and changing their environment though.

[Chaircat Meow]

1 hour ago, TrueMetis said:

A person who believes they can fly by flapping their arms is going to have their survival chances affected, a person who believes they can survive snakebites because of their faith is going to have their survival chances affected. There is every reason to believe that evolution via natural selection does bring about reliable* cognitive because our ability to make successful predictions is second only to, and deeply intertwined with, our ability to work together in abilities that allow us to live.

Untrue beliefs that we hold anyway are the sickle cell anemia's of belief, they typically don't have very negative effects, usually because they're beliefs about things that are irrelevant to daily life anyway, and have a mass benefit in keeping the tribe united. At least that was true back when god was the local mountain and the most that happened was the ground shook when he was pissed off. Now when there's a bunch of different gods that can talk to you and everyone and their mother is willing to claim their interpretation of an old book is correct the negatives are much larger and serve to divide more than unite. But that's both a recent development and is being superseded by other forces anyway. So natural selection never had much of a chance to act upon those kinds of beliefs anyway.

 

Evolution is pseudoscience only to the breathtakingly ignorant. If you cannot imagine a way for true beliefs to affect survival chances your imagination is stunted.

 

*At least generally and at the scale we live in.

The thing that causes the response to the given stimuli (say a crocodile in a river) is the neuronal state. The brain gets the stimuli and neurons fire and this determines how we act (so say evolutionists by NS anyway). So it is not correct to say the belief causes the responses, and thereby gets selected if it conduces to survival given E&NS.

Belief is a mental property, it isn't something necessary to respond to stimuli in such a way that conduces to survival (as my bacterium example showed). Indeed, you could have a false belief and respond to the stimuli in the survival enhancing way, say if you thought the crocodile filled river would get hurt feelings if you waded in it, and that hurting rivers' feelings is something to be avoided. 

So the question is why, given E&NS we would have true beliefs when true beliefs don't cause reactions which lead to survival and thereby get selected for. E&NS would only suggest we have the right neuronal responses to the stimuli not that we would have true beliefs. 

[TrueMetis]

47 minutes ago, Chaircat Meow said:

The thing that causes the response to the given stimuli (say a crocodile in a river) is the neuronal state. The brain gets the stimuli and neurons fire and this determines how we act (so say evolutionists by NS anyway). So it is not correct to say the belief causes the responses, and thereby gets selected if it conduces to survival given E&NS.

You missed some steps, we receive a stimuli and our neurons fire, we interpret the stimuli, then we determine how we act. To use your excellent crocodile example, do you interpret the object in the water as a crocodile, or a log? Belief it's a log leads to one response, belief it's a crocodile leads to a very different one.

Quote

Belief is a mental property, it isn't something necessary to respond to stimuli in such a way that conduces to survival (as my bacterium example showed). Indeed, you could have a false belief and respond to the stimuli in the survival enhancing way, say if you thought the crocodile filled river would get hurt feelings if you waded in it, and that hurting rivers' feelings is something to be avoided.

Belief is absolutely something necessary to respond to stimuli and can be very advantageous, or very disadvantageous depending on the belief. A belief all log shaped objects in water are crocodile is a superior belief to the belief that all log shaped objects in water are logs. And will certainly be selected for false or not. However a belief that it may be one or that other and that can be checked is superior to either and will be selected for even more.

Quote

So the question is why, given E&NS we would have true beliefs when true beliefs don't cause reactions which lead to survival and thereby get selected for. E&NS would only suggest we have the right neuronal responses to the stimuli not that we would have true beliefs. 

Because they do cause reactions that lead to survival, me the guy who believes that a log shaped object could be either a crocodile or a log and because of that belief responds to log shaped objects by checking has a survival advantage because I can cross river more easily than the person who believes all log shaped objects to be crocodiles and therefore won't be able to cross the river as often, and an extreme advantage over the person who believes all log shaped objects are logs who gets eaten more often.

The issue here is that you're for some reason under the impression that our beliefs are not part of what our response to neural stimuli will be. Our response to stimuli is directly affected by our beliefs, because our beliefs can change what we think that stimuli is.

[Chaircat Meow]

36 minutes ago, TrueMetis said:

You missed some steps, we receive a stimuli, we interpret the stimuli, then we determine how we act. To use your excellent crocodile example, do you interpret the object in the water as a crocodile, or a log? Belief it's a log leads to one response, belief it's a crocodile leads to a very different one.

Belief is absolutely something necessary to respond to stimuli and can be very advantageous, or very disadvantageous depending on the belief. A belief all log shaped objects in water are crocodile is a superior belief to the belief that all log shaped objects in water are logs. And will certainly be selected for false or not. However a belief that it may be one or that other and that can be checked is superior to either and will be selected for even more.

Because they do cause reactions that lead to survival, me the guy who believes that a log shaped object could be either a crocodile or a log and because of that belief responds to log shaped objects by checking has a survival advantage because I can cross river more easily than the person who believes all log shaped objects to be crocodiles and therefore won't be able to cross the river as often, and an extreme advantage over the person who believes all log shaped objects are logs who gets eaten more often.

The issue here is that you're for some reason under the impression that our beliefs are not part of what our response to neural stimuli will be. Are response to stimuli is directly effected by our beliefs, because our beliefs can change what we think that stimuli is.

Put it this way, if some kind of Cartesian dualism, or monism, or whatever is true, E&NS as it currently stands is wrong. The only theories of mind that seems compatible with E&NS are various kinds of materialism. You could (as a materialist) say that beliefs and all other mental events just are neurons firing (Daniel Dennett, for example), or you could say they are something over and above neurons firing but are nevertheless totally causally determined by neurons firing (an example would be John Searle’s ‘biological naturalism’).

Either way everything that causes our response to a stimuli is a neuron firing, not a belief. You might call a very large set of neurons firing together in a certain pattern a belief in the way you call a particular concatenation of molecules a table but in causal terms that’s all it is, a bunch of neurons firing; it has no causal properties the constituents do not possess. If this is not the case E&NS is goosed anyway.

Now, all these neurons have to do to conduce to my survival and ensure my genes are propagated is to respond to the stimuli (whatever they may be) in ways that maximise my chances of survival. They have to fire in the right patterns or whatever. Belief, as a mental property features nowhere in this causal account, so beliefs, such as in my case that I’m an especially handsome feline are just by products of the neurons firing; they don’t cause anything and they are themselves causally determined by neurons firing. So they can’t be selected for and so we have no reason on E&NS to think we have been selected to have true beliefs.

So that’s one thing, beliefs don’t cause anything on E&NS unless you're using the word belief to just mean pattern of neurons firing.

Furthermore as my example of the river showed you could have all the survival maximising responses and yet have all the wrong beliefs. Let us say a tribe eat a very nasty tasting plant that actually has antibiotic properties and only endure the taste because they think (totally wrongly) it makes them less likely to be bitten by ants (which it doesn't).

Or maybe I run away from lions and tigers because if I don’t the lion or tiger will catch me and make friends with me, and if I make friends with a lion or a tiger a giant penguin called Gwagwa will come and kill me and I don't want to get on the wrong side of Gwagwa (because who would).

My beliefs would conduce to survival because they would correlated with the neuronal firings that cause me to eat the nasty antibiotic and flee the lions and tigers but the accompanying beliefs would not be true and this would not matter in evolutionary terms.

[A wilding]

4 hours ago, Chaircat Meow said:

A very serious case occurs in biology and pertains to the theory of evolution through natural selection. This is accepted as fact by the overwhelmingly majority of the cognitive elite throughout the western world but is  incoherent. That is to say to state ‘I justifiably believe in evolution via natural selection,’ or ‘the evidence supports evolution via natural selection’ is like uttering the statement ‘no statement I utter is true.’ Evolution via natural selection might be true but no one could rationally believe it, so its plausibly a kind of pseudo-science (not methodologically but in terms of outcome).

The reason for this is that the probability we will have reliable cognitive faculties (i.e. truth finding mechanisms in our minds) is very low if we suppose evolution via natural selection. That is P(R I E&NS) is low, where R is having reliable faculties. This is so because it seems evolution via natural selection would not select for true beliefs but only the neural structures advantageous for survival.

As has already pointed out, the whole point of theory of evolution is that it results in outcomes that are very low priority, only provided that there is sufficient evolutionary pressure to get there. Your reference to "reliable cognitive faculties" appears to me to be essentially talking about free will, as you then start talking about religion. Free will is beyond the scope of theory of evolution (at least in its current form) and more generally is a problem that science has not (yet) solved (except perhaps by showing the we have somewhat less free will than we sometimes think).

Just about all scientific theories have limits and points where they break down, possibly ones that have not been discovered yet. To imply, as you do that evolution can't be "sensibly believed" because it has such limits is, I feel, somewhat disingenuous (and I note the conflation of scientific evidence with belief there).

 

As this theory is taught in schools as scientific and can’t be sensibly believed it is to be hoped President Trump will do something to support those suggesting alternative theories, such as religion.

But now I am really not sure that I have not missed a sarcasm flag somewhere ...