the 'no miracles' argument for scientific realism- any thoughts?

[plant]

i know there are enormous free-wheeling discussions elsewhere regarding the nature of reality etc...
i'd like to open this post specifically to discuss Putnam's 'no miracles' argument in favour of scientific realism over anti-realism. Please confine discussion to this specific topic if possible.

In addition, lets assume that there is an objective, real world that exists independent of our consciousness.

Lets assume for the purposes of this discussion that there actually is a 'mind independent reality' and avoid radical skepticism.

to quote Paul Dicken "the scientific realism debate takes, as it's point of departure, the assumption that it is possible to have at least some knowledge about the external world" (A Critical Introduction to Scientific Realism)

the positive argument for realism is that it is the only philosophy that does not make the success of science a miracle....theories accepted in mature science are typically approximately true... Putnam 1975

to paraphrase...

our scientific theories are approximately true because they correlate (imperfectly) with an external real world.
our theories are not just made to fit the data (there are many possible theories which will give approximately the same results = underdetermination), but also make novel predictions which are falsifiable (Popper), and can be correlated with other data and theories from other experiments.

SOME THOUGHTS FOR:

it would be a miracle if a theory which posits electrons and atoms made accurate pedictions unless these entities and the relationships between them actually exist

this doesn't PROVE scientific realism.... but it surely makes it more plausible than the alternative?

'approximately' true... structural realism vs entity realism... i.e. protons don't really have 'positive' charge, they just have the opposite charge of an 'electron'.

SOME THOUGHTS AGAINST:

pessimistic induction: most of our old theories were false.. even the ones that had predictive value... phlogiston, ether

[swampyankee]

I'm not sure of your points.

Certainly, one of the fundamental and unprovable bases of science is that there is an objective reality that can be discovered by experiment and observation. I think there is also an associated acceptance that observation and experiment have limits. For example, within current understanding of the structure of the universe, neither experiment nor observation can extend beyond an event horizon. Similarly, we probably will never be able to demonstrate or invalidate the many worlds interpretation of quantum mechanics.

[plant]

i guess what i am getting at is:

lets assume for the sake of argument there is an objective real world out there that we can know something about. (whether there is or not is a separate thread of threadworms).

then,
does the 'no miracles' argument 'hold water'?
are there good arguments against it?

[plant]

... granted we may never know everything there is to know... e.g. what is happening in the 'unobservable' universe (if that term has any meaning?) 90 billion light years away... but does the 'no miracles' argument mean we can at least know SOMETHING about a real, external reality?

[Ken G]

In addition, lets assume that there is an objective, real world that exists independent of our consciousness.

That's like saying let's assume a supreme being exists, and then begin a debate on atheism.

to quote Paul Dicken "the scientific realism debate takes, as it's point of departure, the assumption that it is possible to have at least some knowledge about the external world" (A Critical Introduction to Scientific Realism)[/U][/I]

There are so many flaws in that statement I hardly know where to begin. But let me point out the critical one: what is meant by "the external world." In science, that phrase has to be a testable model, because so does everything have to be. So we need operational meanings, those are the meanings that allow tests. So what is the operational meaning of "external"? It means we are referring to a model that is based on an idealized dichotomy between "internal" and "external" worlds. Of course that is a false dichotomy, obviously, but it, like many artificial idealizations, spawns many classes of useful models. So the statement that we can have knowledge about the external world, to a scientific thinker, has to mean exactly one thing: it is demonstrably useful to utilize models that include that obviously artificial dichotomy. But here's the problem: stated in the scientific way, how is that an entry into realism?

the positive argument for realism is that it is the only philosophy that does not make the success of science a miracle....theories accepted in mature science are typically approximately true... Putnam 1975

This argument is completely specious for the simple reason that I can, with hardly any effort, suggest a list of potential reasons why mature sciences produces approximately true theories that neither involve miracles, nor naive assumptions that the term "external world" means anything but what I just said: a useful idealization we put into our models for all the reasons we always put idealizations into models.

our scientific theories are approximately true because they correlate (imperfectly) with an external real world.

Or, they are approximately true because we keep modifying them by trial and error until they allow successful predictions. That's science. The idea that there is some reason for this that is explained by some imagined "correlation" with an external real world requires a willingness to almost completely ignore the entire history of the enteprise of science.

What I don't get is why Putnam thinks he can attach the word "mature" to science, and all of a sudden the year 2017 AD is vastly different from the year 2017 BC. It's as though he thinks that philosophers in 2017 BC weren't also realists, and didn't also think they had a pretty good handle on "the external real world." News flash: they did. That should pretty much put his argument to bed right there.

[Selfsim]

... the positive argument for realism is that it is the only philosophy that does not make the success of science a miracle....
...
to paraphrase...

our scientific theories are approximately true because they correlate (imperfectly) with an external real world.
...
it would be a miracle if a theory which posits electrons and atoms made accurate pedictions unless these entities and the relationships between them actually exist
this doesn't PROVE scientific realism.... but it surely makes it more plausible than the alternative?

Why is 'plausibility' being used to promote a philosophy over objective testability ... in a science forum?

[profloater]

in addition to those excellent points, how do the assumptions suggest no miracles? Or conversely why does any other assumption make science a miracle? Are we to also add an assumption that miracles ie unexplained phenomena, cannot exist? Unfortunately the miracle test is very unsatisfactory.

[Ken G]

To flesh out some of the arguments, let's start with a demonstration that the "external world" is obviously an attribute of an idealized model, not a universe. I'll ask, is my body part of my external world, or my internal world? How about my brain? If the realist holds that my body is made of particles obeying the same rules as in the "external world", and if all it does is feed signals to my brain, then it's clearly part of the "external world." But my brain is all that too, and it has parts, some that are not involved in my generation of this model attribute I call "external." So are the parts of my brain not so involved the external, or internal, world? And even the parts of my brain that are involved in this concept of an external world are made of those same atoms, following those same rules, are they not? And what about your brain, to me, would Putnam say that your internal world is my external world? So these are the telltale signs of model attributes-- when you dig into them, you find they involve simplifications and idealizations that are endemic to that particular model, but fail badly to be able to give a complete description of some "external reality." Or is Putnam entering into Cartesian dualism, in which the thoughts of the mind are something quite separate from the physical atoms and rules inhabiting the brain? Is that the "internal" vs. "external" world, Cartesian dualism that no physicist ever pays much attention to?

So here's the problem: if you are a realist, then there is no "internal" world, it's all external. But then if it's all external, then my thoughts are part of the external world too. So all my descriptions of the rules and attributes of the external world, included in those thoughts, are also part and parcel of the external world. So apparently, we have a tiny subset of the external world that approximates the truth of the external world! Interesting fantasy, that one. The point being, the effort to tear apart physical phenomena, such that we can say we have the reality and we have our approximately true thoughts about reality as two separate elements of that reality, is nothing but the attributes of an idealized model. We can simply observe this to be true, it's much easier to infer than some "external reality." So that's why I argue the very people who would separate the "maps from the territory" invariably confuse the two even in their own models. Oh the irony.

Also, a key argument that this thread is meant to address is the question of how can science work if it is not some fairly close approximation of the truth. This argument simply boggles my mind, for it seems to require a willingness to completely ignore scientific history. Over that history, we have many examples of models that are now viewed as nothing at all like the truth, but which worked quite well and were regarded as the truth in their day. And here I do not refer to the occasional missteps like the aether and phlogiston (assuming neither return at some point), I mean every single theory we have ever inferred and used. Just go right down the list of the biggies, the geocentric universe, the concept of indivisible atoms, Galilean relativity, the deterministic laws of Newton, the wave theory of light of Huygens, local realism, how long do you want me to go on for? All of these theories are approximately correct, but none of them are anything remotely approaching "the truth." In fact, they often seem closer to the opposite of the truth-- the geocentric model held that the Earth was stationary and completely different from the rest of the universe, indivisible atoms have been broken down into quarks and gluon fields involving some weird combination of real and virtual aspects, Galilean relativity was found to be completely wrong in any system of very fast-moving components (even though it was regarded as self-evidently true), quantum mechanics, without added bells and whistles, says god rolls dice, light is now regarded as comprised of particles not waves (though all particles are ruled by wavelike mechanics), and quantum entanglement has shown that systems can behave holistically over nonlocal distances (something Einstein's brand of realism regarded as so absurd it was self-evidently false).

So the one thing that is completely clear from history is that the habit of regarding physical truths to be knowable from limited experience is a recipe for being wrong. Yet the realists think they can do exactly that, and even imagine that the fact that they can do that is evidence for realism! It mystifies me. All anyone has to do is use realism to make even a single correct prediction about anything, and I'll be impressed. Let's see how that would have worked out for Aristotle-- realism would have told him orbits are circles, because that's the way reality has to work. Newton would have used realism to predict that all behaviors are deterministic, since that's not just his theory, it's how the real world works. And so on-- it's simply a category error, a scientific mistake, to confuse one's theories with what reality actually does. Yet the realists have seem to have missed this completely.

[SkepticJ]

it would be a miracle if a theory which posits electrons and atoms made accurate pedictions unless these entities and the relationships between them actually exist

Are planets really on epicycles, orbiting the Earth? Yet that theory worked pretty well for a long, long time.

[swampyankee]

Miracles are, by definition, not understandable within current knowledge; since current knowledge is finite, there may be items of objective reality that are inexplicable.

[Hornblower]

I would say that the existence of conscious beings like ourselves who can think about and discuss this topic is a miracle.

[plant]

Ken, regarding the idea that our theories of 2017AD are as 'true' as our theories from 2017BC... i have some thoughts....

1. scientific methodology did not dvelop properly until the 12C at the earliest.... so democritus gets few points for being right for the wrong reasons for developing the atomic theory of matter. Ditto for Giordano Bruno for getting burned at the stake for believing in extrasolar planets- right for the wrong reasons.
2. lets compare newton vs einstein. sure the concepts are totally different bent spacetime vs invisible forces... but at low velocities the math is pretty similar isn't it? the einsteinian 'correction' is negligible at ordinary velocity- preumably that's why it took until modern times to measure this effect. Surely this is some form of structural/mathematical realism? Surely the fact that at low celocities the equations give the same results is an indication that there is a relationship in the real world between our models of mass, distance etc. Rest mass = mass at low velocities. Isn't all this cause for celebration that the structures of our scientific e indeed mapping on to an external reality?
3. re kepler.... isn't a circle just a special case on an ellipse?
4. re epicycles... yes one can retoactively fit any number of epicycles to fit the data... however such a theory will not make good predictions of other solar systems.
5. assuming that 'the future will resemble the past' in terms of our scientific theories is an examlpe of induction. The turkey thinks every day is going to be wondeful.... until the day before thanksgiving. I think most people would deny that induction is a good approach to science.
6. if our theories do not somehow correlate/map on to some external world... how is it posible, for example, that the kinetic theory of matter can ex-lain brownian motion seen in a cup of hot tea, and give us boyles 'law' et etc.... multiple experiemnts from different angles lead us to a common theory which can make predictions.
7. if ken can use his radical skepticism to deny even the possibility of an external real world.. then i can deny that anything exists outside of Ken's mind. Every one of ken's posts, everything is a figment of hhis own imagination. Ken chooses not to believe this is true. This is NOT scientific but a choice he makes. He chooses to believe there are thoughts and ideas outside his mind. Ideas he disagrees with. Well, i choose to believe (for the sake of this post only) that (as a starting asumption) there is an objective reality that we can know something about. I choose to escape from the solipsism!
8. Putnam, by the way, is past tense.. died last year RIP. Many philosphers of science have addresed the 'no miracles' argument and there are interesting arguments and discussions to be had as it seems to be the basis of what most 'scientific realists' believe.

[plant]

by the way i'm not pretending i am clever enough to understand the dens e logical thickets of an entire philosophical field...
if anyone understands and can translate into plain english these the ideas of...

Stathis Psillos
Arthur Fine
Bas van Fraassen
Lipton

... and how Bayesianism can help....


i'd be appeciative....!
trying to start up a converation at a dinner party regarding scientific realism has proven to be difficult.. even amongst my medical colleagues and scientists and rabid agnostics.... i think very few people in the sciences have ever thought about these issues. I think for most scientists...(maybe not physicists!) their day to day assumption is 'naive scientific realism'.... in the same way as 'a fatalist will still jump out of the way of an oncoming car" (Isaac Bashebis Singer)... so itis great to have a forum like this.... :-)
i digress....

[Hornblower]

4. re epicycles... yes one can retoactively fit any number of epicycles to fit the data... however such a theory will not make good predictions of other solar systems.

Why not? Planets in any other system will have orbits that fit Kepler's laws, so the same technique of approximating them to any desired degree of precision with a series of epicycles will work kinematically just as well there as here.

[George]

Plant, I expect Ken is arguing kind and you are addressing degree. It's not that today we can see more "turtles" upon which the Earth sits and have better techniques, but that the bottom one is still very indeterminate.

[plant]

..and if you are an idealist.. (or an anti-realist) there is no external world. it is all internal.....
..and if it's all internal to your own single mind.. how does one explain the predictive success of science without resorting to miracles...
...hence the post....

[George]

Why not? Planets in any other system will have orbits that fit Kepler's laws, so the same technique of approximating them to any desired degree of precision with a series of epicycles will work kinematically just as well there as here.

Yes. It's interesting how nice an ellipse forms from a single epicycle. The speed of a planet also slows and speeds-up along the deferent to better fit what actually is seen. But there is a difference. I wonder why Kepler was slow (years) to try the ellipse? Perhaps it may be more a story of putting the Sun at a foci?

Epicycles.pdf

[Hornblower]

Yes. It's interesting how nice an ellipse forms from a single epicycle. The speed of a planet also slows and speeds-up along the deferent to better fit what actually is seen. But there is a difference. I wonder why Kepler was slow (years) to try the ellipse? Perhaps it may be more a story of putting the Sun at a foci?

Epicycles.pdf

That sketch is not a first approximation of a Kepler ellipse. As generated, it gives maximum velocity at both ends of the major axis and minimum at both ends of the minor axis. On a Kepler ellipse the velocity is greatest at one end and least at the other end of the major axis. To do a rough approximation with one epicycle, it needs to be revolving retrograde at twice the deferent rate, with the epicycle's center on an eccentric deferent circle whose center is offset from the Sun. I did it myself in the early 1980s, when I still knew how to program in BASIC on a TRS-80 computer and use Newton's method from a college textbook to calculate the motion around a Kepler ellipse. By trial and error I could get the periodic error down to a roughly sinusoidal ripple that could be attacked with successive smaller epicycles at frequencies of multiples of 2 times that of the first epicycle. With 5 or 6 such terms I could fit a planet's angular position about the focus within an arcsecond for an orbit of eccentricity 0.05, about that of Jupiter. Kepler knew this could be done, but he considered it an outrageous kludge and decided to try other simple curves when it became clear that two constant rate circular terms would not fit Tycho's data.

I wonder why Kepler was slow (years) to try the ellipse? Perhaps it may be more a story of putting the Sun at a foci?

Kepler's thinking initially was two millenia behind the times, rooted in archaic Pythagorean mathematical ideals. He started with what seems like a crackpot idea to us, that is, thinking the orbits of the planets were somehow related to inscribed and circumscribed regular polyhedra. He imagined that by using each of the five polyhedra only once, he could account for the fact that there were exactly six known planets, along with the relative sizes of their orbits. This was hand in hand with the ancient dogma which insisted that only perfect constant rate circles, or vector combinations thereof, were permissible in the heavens. Fortunately he was an honest and persistent observer who was willing to change when it became clear that the ancient dogma could not keep up with increasingly precise measurements.

[SkepticJ]

. . . rabid agnostics

The Rabid Agnostics would make a great band name.

[gzhpcu]

Plant,
In a nutshell, this was my standpoint in the mammoth thread on reality:


For me a picture is worth a thousand words.

[Len Moran]

Well, i choose to believe (for the sake of this post only) that (as a starting asumption) there is an objective reality that we can know something about. I choose to escape from the solipsism!

No one anywhere can tell you what to believe or not believe - the debate between realism and idealism will rage on indefinitely, but hopefully it will be increasingly seen by the scientific community as being a philosophical debate that has no bearing on the remit of the scientific method. That scientific remit cannot venture into a conjectured mind independent realm - such a venture has to transparently admit to taking up a philosophical position and be happy doing so.

However one chooses to frame that admission, it should never be forgotten that it is an admission of a belief structure and that belief structures are only relevant to the individual who chooses to take them up, they have no empirical basis upon which we can use them to describe an objective reality and hence take to be a scientific truth. Some will argue otherwise, they will say that such and such philosophical argument (such as the no miracle argument) "pushes" the belief structure to a level that is universally "true" such that science can effectively dismiss the belief part and turn the discipline into a means of directly accessing a conjectured mind independent reality. But that just seems a quagmire to me, it leads to a free for all, after all, who can really say one set of beliefs is more likely than another? The no miracle argument can readily be philosophically rebuffed, but that doesn't make it wrong, it just establishes two philosophical camps with no possibility of applying empiricism to settle the matter.

It seems to me that you are looking for a personal view of nature, one that asserts there is a mind independent realty, one that escapes solipsism, and one that sees science as directly accessing a "real" mind independent realm. And that's all fine - go ahead and do it. But do it independently of science, don't try and marry up the two in an "objective manner" such that you "feel" there is a philosophical means by which your desires/beliefs for the existence of a mind independent realm that can be accessed via science can be established. It can't and never will be.

Thankfully, no matter what feathers are ruffled with regards to this issue of realism and science, no matter who gets hot and bothered over it all, science via the scientific method remains impervious to it all, it continues to give us utility in term of predictive models and interesting descriptive pictures. That it can never empirically find a mind independent "source" for that utility does not diminish that utility in the slightest. And it allows us as individuals to legitimately form our own personal belief structures concerning such a "source" (or even dismiss the idea of there being such a source) without imagining that we can find within that structure some universal "truth". The truth is true only to us as individuals.

Your beliefs concerning realism may be "common sense" to you, but there are a range of realist positions in which to believe. I am a realist also, but I don't take up naive realism or any kind of representative realism - I just believe there is some kind of mind independent realm and within that realm there is a "something" that "feeds" into phenomena such that our reality doesn't entirely depend on us. But what that "something" may be is anyone's guess, I don't believe that this realm exists in space and time and has any "causal mechanisms" - but it simply isn't open to empirical investigation. And that's my belief - I happen to personally believe that it is much more plausible than your belief based upon my thinking on this issue. I can't say that you are wrong, I just personally believe you are wrong, but who knows, perhaps you are right! But what's important is that neither of us pretends that we can place our beliefs into science and then forget that they are belief structures - we need to properly separate our belief structures pertaining to an independent realm from the remit of the scientific method.

[SkepticJ]

. . . I mean every single theory we have ever inferred and used.

I get your point, but you are, strictly speaking, wrong here: you're not noting the old theories that haven't been falsified. Archimedes understood that objects float because they displace more weight/mass than they have, and he understood why parabolas can focus light (though he probably didn't build the legendary "death ray"). Ancient peoples understood that there is a direct relationship between the mass and speed of an object and how much of a "punch" it has: water wheels weren't just built anywhere, they were built where there was a lot of slow flowing water, or less of fast flowing water, and there were tailored designs for each flow-type. Why does water put a fire out? I'm sure the ancients thought because it cooled the fire, like it cools us off when we're hot. Why does sand? Because it keeps the air from getting to the fire; they knew air was essential for some reason because blowing on fire makes it more intense, such as in a metal forge, and you can snuff a lamp out by covering it.

The Theory of Evolution will turn 200 in a few more decades, and the basic mechanism of natural selection is extremely unlikely to be wrong.

[profloater]

The Rabid Agnostics would make a great band name.

Yes indeed but there still seems to a lot of confusion about any agnostic stance in this debate, including in Plant's posts. Agnostics do not deny the existence of external reality they just know that they cannot know how it works. Denial would be a very extreme solipsism view and not in line with all the evidence. There is a huge consensus about the scientific model of how everything works and that's fine, no miracles included, unless he means the big bang, and no deities either, but it's rather unkind to call an agnostic rabid, it's rather a well thought out standpoint and not at all extreme. Unless you don't choose to understand what "can't know" means.

[Ken G]

1. scientific methodology did not dvelop properly until the 12C at the earliest.... so democritus gets few points for being right for the wrong reasons for developing the atomic theory of matter. Ditto for Giordano Bruno for getting burned at the stake for believing in extrasolar planets- right for the wrong reasons.

True, but the main example I used, Ptolemy's geocentric model, was prepared in a way that is pretty close to modern scientific standards.

2. lets compare newton vs einstein. sure the concepts are totally different bent spacetime vs invisible forces... but at low velocities the math is pretty similar isn't it?

That's the point-- the realist doesn't say the maths are similar to reality, they say the description is close to the reality! We know the maths are all going to be similar because we already have the quantitative measurements, so any maths that aren't similar are not going to give those results.

Surely this is some form of structural/mathematical realism?

I would simply call it "objectivity"-- that a set of quantitative measurements can be agreed on by any who conduct the measurements. As soon as you regard that set as your constraints for any theory, the theory has to approximate those results. But there are many theories that will reduce to those results in the necessary limit. This is the main logical error of realism: they think that if you get the answer right, they must have the uniquely correct process for arriving at that answer. But actually, many processes lead to the same answer, so you can't reason backward from the answer to the process, and still regard the process as "the truth."

Isn't all this cause for celebration that the structures of our scientific e indeed mapping on to an external reality?

No, because it is not a one-to-one mapping. Will realists be happy to say that their thoughts about reality correspond to a reality that might be like ours, or might be like some other that looks quite similar to ours superficially but is actually totally different if you dig deeper? Would that satisfy them? In other words, if there is some other reality that actually does obey Newton's laws, where stars never undergo core collapse supernovae so have very little iron in the planets, would the realists of Newton's day be happy to know they got that other reality right, but not ours?

3. re kepler.... isn't a circle just a special case on an ellipse?

Absolutely, that's the point. The realists pre-Kepler thought the reality was of circular orbits. The actual reality allows more elliptical orbits, so we have phenomena that are due to elliptical orbits in our reality that would not be in the one they got right. Would that satisfy them, to get the wrong reality right?

4. re epicycles... yes one can retoactively fit any number of epicycles to fit the data... however such a theory will not make good predictions of other solar systems.

In a reality where epicycles were "true", then they would. In ours, they won't. That's the point-- our reality is not the reality that the Ptolemaic realists believed in. I think it would have been small consolation to them to find out they would have been right in some other reality!

5. assuming that 'the future will resemble the past' in terms of our scientific theories is an examlpe of induction. The turkey thinks every day is going to be wondeful.... until the day before thanksgiving. I think most people would deny that induction is a good approach to science.

Science doesn't assume its theories, it tests them. I think it's very ironic that you don't count the future resembling the past as a good basis for a scientific model-- given that every single scientific model we have exhibits that property!

6. if our theories do not somehow correlate/map on to some external world... how is it posible, for example, that the kinetic theory of matter can ex-lain brownian motion seen in a cup of hot tea, and give us boyles 'law' et etc.... multiple experiemnts from different angles lead us to a common theory which can make predictions.

"How is it possible that..." finish with any one of the vastly surprising things we've had to put into our models. So what? This is commonplace.

7. if ken can use his radical skepticism to deny even the possibility of an external real world..

If you plan to change my words into things I never said, you will not come very close to understanding what I am saying. What I showed, and it really is plain as day, is that the concept of an "external real world" is clearly a model, like Ptolemy's model of the solar system, or Bohr's model of the atom. This should be perfectly clear, it really boggles my mind why it isn't. Note this isn't "denying" the external world, any more than I "deny" Bohr's model of the atom-- it is noticing what we are doing when we talk about these things.

then i can deny that anything exists outside of Ken's mind.

Do whatever you want, but if you cannot present the kinds of logical arguments I'm presenting here, I shouldn't imagine it will have much point.

Many philosphers of science have addresed the 'no miracles' argument and there are interesting arguments and discussions to be had as it seems to be the basis of what most 'scientific realists' believe.

That's true, but I generally find the level of the discussion is pretty uninformed, since they are rarely able to draw examples from science. There's a kind of gulf where philosophers don't know much physics and physicists think philosophers are "navel gazing," so the dialog is missing a lot.

[Ken G]

I get your point, but you are, strictly speaking, wrong here: you're not noting the old theories that haven't been falsified.

Actually, what I said applies even to theories that have not been falsified. I wasn't saying they will later prove wrong because I have a crystal ball (we might go extinct next year and never falsify any of them), I was saying they should all be regarded as approximate whether they are falsified or not. What is "unrealistic" is thinking the ones that haven't been falsified are true, until they get falsified. Given the history, that strikes me as downright silly.

Archimedes understood that objects float because they displace more weight/mass than they have, and he understood why parabolas can focus light (though he probably didn't build the legendary "death ray").

That is how he came to understand those phenomena, yes. But did he know why objects receive an upward force equal to the weight of the water they displace? And even today, we say it is due to a pressure gradient in the water, but do we know why that pressure gradient is there? (It requires a model of gravity.) And notice that even "pressure" has many models that describe how it works-- and they are different!

This is the whole point. Every person who thinks the phases of the Moon are caused by the shadow of the Earth think they live in a reality where that is true. If it never gets falsified, they live and die and never know. So how do realists distinguish themselves from those people? Because their realism is the "true" realism?

Why does water put a fire out? I'm sure the ancients thought because it cooled the fire, like it cools us off when we're hot. Why does sand? Because it keeps the air from getting to the fire; they knew air was essential for some reason because blowing on fire makes it more intense, such as in a metal forge, and you can snuff a lamp out by covering it.

The problem with realism is not that it fails to get working models, it is that it doesn't understand that it is using models in the first place.

Thus there are two basic flaws in realist thinking, both of which make it quite naive:
1) It is constantly mistaking models that have not yet been falsified for models that are "close to the truth." When one looks at the history of science, this is a dubious perspective, examples abound (not phlogiston).
2) It doesn't recognize the difference in type between a model and a reality. Depending on how we use the latter word, either they are of the same type, which means "reality" is just another model, or they are not of the same type, in which case there is no sense to which one can be "close to" the other. Is the number 2 "close to" a married couple, because there are 2 people? The number is an attibute of a way of thinking about a married couple, it's not a married couple and it's not "close to" a married couple. (Of course, I'm of the view that even what we call a "married couple" is a model we have in mind, but I'm talking about internal inconsistencies in the thinking of the realists themselves.)

The Theory of Evolution will turn 200 in a few more decades, and the basic mechanism of natural selection is extremely unlikely to be wrong.

Ah, so if a theory lasts 200 years, then it's the truth? In a universe we successfully model as being 13.8 billion of what we mean by years old, 200 years is the standard a realist should hang their hat on? Ptolemy's geocentric model lasted 7 times longer than that. Besides, to claim that the theory of evolution has "lasted" requires using the device of intentionally vague thinking. In actual fact, the theory has undergone significant modifications-- there wasn't even an understanding of genetics when it was first suggested. So we are free to reinterpret our basic guidelines as we go along, and the more vague those guidelines are framed the more they will tend to persist, but here's the problem: the realist does not have the freedom to be vague, and they do not have the freedom to "cherry pick" what aspects of their prior realism they have chosen to be the true ones, throwing out the rest at will! They need their whole story to be the actual reality, no "reinterpretation" allowed (if you need to reinterpret it to be the actual reality, then it ain't currently the actual reality). A classic example is the motion of the Earth: almost all realists from more than 500 years ago thought the Earth was really stationary, then they all thought it was really moving, now we think there's no such thing as "really moving" because all motion is relative! For crying out loud, will the realists who are talking about the actual reality please stand up? When realism becomes a "slippery slope" of dropping and altering its elements, then it simply isn't realism at all, it's an internally inconsistent world view.

[plant]

if you set up a 'straw man' of naive scientific realism...this is pretty easy to burn down.
i don't think there are many naive realists out there in the 'philosophy of science' community...
when you say 'the realist does not have the freedom to be vague"... i disagree. I think most structural realists would say something like 'our best scientific theories are approximately true in the sensse that the mathematical telationships between entities are approximately correct BECAUSE they map on to an external reality'. if they did not, it would be a miracle that science can progress and can make pedictions. surely the reverse isn't logical.. is our reality because' / defined by what we find in our scientific theories? why then would our theories be so consistent, reproduciable and predictive?

why does the radical skeptic/ anti-realist have the freedom to believe that anything at all exists outside his own mind? odd that the internal thoughts and models in ones mind seem to be so miraculously internally-consistent? why could they not be a surreal, dreamlike imagining?

[Ken G]

if you set up a 'straw man' of naive scientific realism...this is pretty easy to burn down.

Actually, naive realism is pretty much what Putnam was selling.

when you say 'the realist does not have the freedom to be vague"... i disagree. I think most structural realists would say something like 'our best scientific theories are approximately true in the sensse that the mathematical telationships between entities are approximately correct BECAUSE they map on to an external reality'.

Let's look at that sentence. Now let's remove everything from the BECAUSE onward. How has it changed? Well, we got rid of all the stuff we have zero evidence for, and lost nothing of what was important. So if including it is "realism", I call it not only naive, but also useless.

if they did not, it would be a miracle that science can progress and can make pedictions.

And that's the part that is pretty naive. All we have to do is look at theories that were mathematically approximate, but not even remotely close to what was "really going on." I could give a very long list, but let's focus instead on one very telling example that realist philosophers seem to be completely unaware of (given the thinness of their arguments): Galilean relativity.

Galileo thought speeds added like v_tot = v₁ + v₂, it seemed almost obvious and certainly gibed with experience. Enter the realist philosopher who thinks this velocity addition law "approximates reality." Of course it does not, what approximates reality is f(v_tot) =f( v₁ )+ f(v₂) for some function f that was unknown to Galileo, and to all the Galilean realists (and apparently also the modern realists too). Here's the point: almost any function f that is zero when v is zero (which is just the meaning of being stationary, we associate 0 with being stationary regardless of what the function f is) will have the property that when v is small, the equation can be approximated by a first-order Taylor expansion:
v_tot*f'(0) = v₁*f'(0) + v₂*f'(0)
where f' is df/dv. When we cancel the f'(0), this expression implies Galileo's "law!" So what can we conclude from the idea that Galileo's law is "approximately the true reality" on the grounds that it approximately works at small v? Answer: essentially nothing. If we regard the "reality" as the f(v) function, then Galileo's law tells us almost nothing at all about the reality, other than that v is small, which is what the realist apologists call "approximately the true reality" simply because we live in a "reality" of small v! It's not our understanding of the reality that is approximately right, it is only that we only understand small v. Again, if the realists would be happy to say that we know nothing about the "real" function f(v), but can make mathematical approximations that work for almost any reality since we only experience small v, then I'd be happy. Does that sound like what they say? Nope. Until they can make arguments consistent with what I just said, I see little reason to pay their naivete much attention.

Now, the modern realists who imagine they live in some kind of exalted age where the veil of truth has finally been lifted might say we now know the f(v) function (it's an inverse hyperbolic tangent in special relativity, by the way). But they're just not paying attention. It's the same thing all over again-- their "reality" requires there to be no gravity, so again they only understand the reality of their experience, that of low gravity. A host of "actual realities" branch outward from this ignorance as soon as you turn on a stronger, i.e. more "real," gravity. And so it goes-- will the true "reality" that the realists think they approximately understand please stand up? We can easily see that all they understand is the restriction of that reality onto their limited experience. Which is my whole point-- that's the definition of naivete.

surely the reverse isn't logical.. is our reality because' / defined by what we find in our scientific theories? why then would our theories be so consistent, reproduciable and predictive?

This is called the fallacy of the excluded middle, the claim that a proposition must be true if its opposite seems false. You can do better than fallacies, surely. The correct argument is we have no idea how to add a "because" to that statement that has any meaning or value, or is testable in any way. That's the part you have excluded in your fallacy.

why does the radical skeptic/ anti-realist have the freedom to believe that anything at all exists outside his own mind?

First of all, everyone has the freedom to believe whatever they like. We aren't talking about beliefs, we are talking about logical arguments that connect with the data available to us. Like the one I just gave about Galileo's "law," that all the realists before 1900 would have said was the "approximately true reality" (if they didn't claim it was clearly the actual reality), but which we now see was nothing more than the statement that all the velocities we experience are small, in the grand scheme of "how reality actually works." Is that the same thing as a realist philosophy, that we don't know anything about the reality because all versions look pretty much the same when all you experience is small velocities? No, that doesn't sound like realist philosophy to me at all, but it's what an understanding of science's history calls for. Relativity was not the discovery that we were approximately right about how velocities really add, it was the discovery that all the velocities we normally experience should be regarded as too small to tell how velocity addition "actually works." And so it goes with lesson after lesson, another realist lynchpin falls to the wayside, but still they keep coming back for more.

[plant]

But surely Ken, you have decided in a non-scientific way to have faith in the idea of an external real world. Either that, or you are typing this yourself.
You have decided. This is not a 'default' position that requires no faith/belief?
If you want to be logically consistent, you must deny the existence of anything outside your own individual mind?

By the way, i sense people's exasperation!I posted this thread as i find it intersting, and i greatly appreciate everyone's input. I don't necessarily hold these beliefs but i'm interested to hear what others feel about them. I don't want to join the cult-of-Ken without understanding exactly where you're coming from... although perhaps i will end up agreeing with you! I'm going to take a little time to properly digest what has been written so far.... thanks again.

[plant]

what is my rest mass? lets say 70kg (i wish)
what is my mass when travelling in a car at 100miles an hour relative to the ground? 70.00000000000000000000000000000000000000000000000 001 (let's say)

so.. isn't newton is an approximation of einstein which will be an approximation of ??? who will be an approximation of XXXXX theory developed by artifical intelligence which we will not have the mental capacity to understand.
the 'structural realist' says - "so what if newton didn't 'know' space-time was curved by mass.. in the future perhaps we will also reject this model.. however the math/ relationships remain approximately true because they map on to a real external world."

[Ken G]

But surely Ken, you have decided in a non-scientific way to have faith in the idea of an external real world.

Huh?

Either that, or you are typing this yourself.

I have lots of evidence that the idealized model we call "the external real world" is a good one, we use it every day. I also use Newton's laws almost every day. But that doesn't require me to thrust my head into the sand and imagine Newton's laws are "the approximately real world," they are merely good approximations to the data I need to connect with via some kind of useful idealized model. We know the laws work in various situations-- from whence comes the need to claim that makes it the approximate reality? What does that ever add to Newton's laws?

You have decided. This is not a 'default' position that requires no faith/belief?

Again, I have no idea why you think any of this involves faith or belief without evidence.

If you want to be logically consistent, you must deny the existence of anything outside your own individual mind?

Nope. Do I need to point out more fallacies in your thinking? You should correct the last one I mentioned first!

I don't want to join the cult-of-Ken without understanding exactly where you're coming from...

Yup, another logical fallacy. Someone can present a logical argument with actual examples from real science, and you can counter that argument by branding it a "cult"? That's the fallacy of "argument by labeling."

although perhaps i will end up agreeing with you! I'm going to take a little time to properly digest what has been written so far.... thanks again.

Then at least you have an open mind! Forgive me if I argue too strenuously, I really would like to see a real logical counterargument, but since I haven't (including Putnam's), it makes me more strident.