PDA

View Full Version : Astronomy: what can we know?



Nereid
2010-Jul-14, 01:40 PM
We have not been outside the solar system yet, and even when we do, it is highly unlikely that we will have visited another star system, in the lifetimes of everyone reading this, today.

So, what can we know about the universe beyond our solar system? Indeed, what can we know about the Sun's core, or even the Earth's?

This thread is specifically for a discussion of this topic.

I was inspired to start it by a post, by Ken G, in a Q&A thread (http://www.bautforum.com/showthread.php?p=1761615):


Well, we do know some things about the darks!

For example, dark matter is cold, and non-baryonic.
"Know" is a strong word. That is certainly the prevailing view, but alternatives like modifications to gravity continue to be worked on. There are also some who feel that you'll never get everything you want from cold dark matter, and you'll need some more neutrino-like hot dark matter. I was just referring to the fact that we have no idea what the stuff is, even though some progress has been made on its attributes.

And dark energy is, today, indistinguishable from lambda, the cosmological constant.
True, but that may just be because the data is so limited. If you only have enough data to assert one statistic about the acceleration, you are likely to be able to fit that statistic with a theory that has one free parameter, and you're not going to be interested in entertaining theories with more than one. I'm not sure we are seeing anything beyond that in current dark energy studies, but it certainly isn't my area of expertise.
Do we know, for example, that other stars are composed, largely, of hydrogen and helium (most of them anyway)?

Do we know that there's a super-massive black hole at the centre of the Milky Way?

Do we know that Geminga is a ~sol-mass object, composed almost entirely of nuclear degenerate matter?

Do we know that dark matter is cold, and non-baryonic?

How should we go about deciding what we know (and what we don't)?

Does it even make sense to say we 'know' something, anything, about things beyond our solar system?

Ken G
2010-Jul-14, 03:21 PM
I would say the first thing we need to establish (and I'm sure you "knew" this would come up), is what meaning of "know" is in play. People not accustomed to delving particularly deeply into what they think they know (and consider the irony of that phrase for a moment) often imagine that "know" just means one thing-- you know it, or you don't. Taking that perspective will lead you almost immediately into logical inconsistencies. You can't even escape by saying you're "pretty sure"-- that's only testable if you have the opportunity to shift from "pretty sure" to "knowing" later on, so does not escape the requirement of knowing what you actually think "knowing" means in the first place.

So instead, we must recognize that "knowing something" can actually mean a lot of different things, and indeed, there is a whole branch of philosophy (yes, that dirty word), called epistemology, developed to treat that question. In a nutshell, epistemology asks the question of the OP, but in more general terms. This thread takes the general area of epistemology and focuses it specifically on what we mean by "astronomical knowledge." There's plenty of fruit for discussion, even in that more focused area!

To start that off, I might point out that we first have to say if by "astronomical knowledge", do we mean knowledge as established within the confines of the scientific method, like a machine for converting observations into what we will label as knowledge when theory gives us unification of and predictive power over those observations? Or are we not content with just that machine, and we want to assert that the output is "known" on a more fundamental level that transcends the label that the machine stamps on it like a trademark on an assembly line?

I'm going to assume that the "transcendence" problem is intractable, though others may feel free to try to address it, and just use "astronomical knowledge" in the "trademark" approach. We say it is knowledge if it pasts the scientific test of knowledge, nothing more-- with the caveat that when we use this meaning, it is possible to later "unknow" something we "thought we knew." Indeed, this happens all the time with scientific knowledge. What's more, we can even identify "levels of certainty", which are really nothing more than "grades" we put with the trademarks: like some eggs might be AAA eggs, and others only A eggs, or we can buy cereals with "national brands" or with "supermarket brands."

Taking that approach, I would go through the specific questions in the OP and try to determine which are the national brands, which are the supermarket brands, and which ones the FDA doesn't even yet "allow on the shelves," if you will. Probably no one would care about my personal opinions on those issues for those specific cases, many of which are not my area of expertise, so I think it matters more to simply identify that this is the process we are undertaking when we start to assess the meaning of scientific knowledge.

trinitree88
2010-Jul-14, 03:39 PM
Nereid. I'll invoke Lorentz Invariance, the Queen of all conservation laws.....the laws of physics must look the same in all reference frames ...in order for us to "know" anything in the universe. What works in a lab in Bangladesh, must match results from Tuva, London, Dusseldorf, Chicago, Brasilia, Kyoto, Seoul, Paris, Turin, Toronto, Adelaide, Beijing, etc... If it doesn't, we need a different physics text for every country, and extrapolating that, for every solar system, and every galaxy.
Most physical investigations involve reasonable extensions of our knowledge to date, using ever more sensitive apparatus, and better data processing. Some, and only a tiny "some", base extrapolations on house-of-cards assumptions which remain unproven, and indeed often untestable in our labs. I see lots of statements which claim more veracity than is justifiable by labwork. It's the nature of the beasts, investigative reporting, crime scenes, science labs. A good dose of common sense should be heard on a regular basis. To that end we can know a lot about the universe, but need to keep in mind that it's always possible to find a new way of looking at things. pete

Ken G
2010-Jul-14, 06:18 PM
Which brings up the interesting issue of objective knowledge-- does science give us objective knowledge because "real truth is objective", or does it give us that because it is inherently part of the "trademark" that it only gets stamped on what is objective? Your comments are consistent with the latter, and that's what I feel is the more supportable stance. Which provides a significant challenge to naive concepts of "knowing" that think in black-and-white terms.

Nereid
2010-Jul-14, 07:04 PM
I would say the first thing we need to establish (and I'm sure you "knew" this would come up), is what meaning of "know" is in play. People not accustomed to delving particularly deeply into what they think they know (and consider the irony of that phrase for a moment) often imagine that "know" just means one thing-- you know it, or you don't. Taking that perspective will lead you almost immediately into logical inconsistencies. You can't even escape by saying you're "pretty sure"-- that's only testable if you have the opportunity to shift from "pretty sure" to "knowing" later on, so does not escape the requirement of knowing what you actually think "knowing" means in the first place.

So instead, we must recognize that "knowing something" can actually mean a lot of different things, and indeed, there is a whole branch of philosophy (yes, that dirty word), called epistemology, developed to treat that question. In a nutshell, epistemology asks the question of the OP, but in more general terms. This thread takes the general area of epistemology and focuses it specifically on what we mean by "astronomical knowledge." There's plenty of fruit for discussion, even in that more focused area!

To start that off, I might point out that we first have to say if by "astronomical knowledge", do we mean knowledge as established within the confines of the scientific method, like a machine for converting observations into what we will label as knowledge when theory gives us unification of and predictive power over those observations? Or are we not content with just that machine, and we want to assert that the output is "known" on a more fundamental level that transcends the label that the machine stamps on it like a trademark on an assembly line?

I'm going to assume that the "transcendence" problem is intractable, though others may feel free to try to address it, and just use "astronomical knowledge" in the "trademark" approach. We say it is knowledge if it pasts the scientific test of knowledge, nothing more-- with the caveat that when we use this meaning, it is possible to later "unknow" something we "thought we knew." Indeed, this happens all the time with scientific knowledge. What's more, we can even identify "levels of certainty", which are really nothing more than "grades" we put with the trademarks: like some eggs might be AAA eggs, and others only A eggs, or we can buy cereals with "national brands" or with "supermarket brands."

Taking that approach, I would go through the specific questions in the OP and try to determine which are the national brands, which are the supermarket brands, and which ones the FDA doesn't even yet "allow on the shelves," if you will. Probably no one would care about my personal opinions on those issues for those specific cases, many of which are not my area of expertise, so I think it matters more to simply identify that this is the process we are undertaking when we start to assess the meaning of scientific knowledge.
My question is one of epistemology, but a quite restricted one.

Also, it's not astronomical knowledge in general, but a very specific subset, to do with 'beyond the solar system'.

More specifically, how does knowing that this critter is an individual of species X, within the phylum Y, differ from knowing that this star is an M dwarf?

Nereid
2010-Jul-14, 07:16 PM
Nereid. I'll invoke Lorentz Invariance, the Queen of all conservation laws.....the laws of physics must look the same in all reference frames ...in order for us to "know" anything in the universe. What works in a lab in Bangladesh, must match results from Tuva, London, Dusseldorf, Chicago, Brasilia, Kyoto, Seoul, Paris, Turin, Toronto, Adelaide, Beijing, etc... If it doesn't, we need a different physics text for every country, and extrapolating that, for every solar system, and every galaxy.
Most physical investigations involve reasonable extensions of our knowledge to date, using ever more sensitive apparatus, and better data processing. Some, and only a tiny "some", base extrapolations on house-of-cards assumptions which remain unproven, and indeed often untestable in our labs. I see lots of statements which claim more veracity than is justifiable by labwork. It's the nature of the beasts, investigative reporting, crime scenes, science labs. A good dose of common sense should be heard on a regular basis. To that end we can know a lot about the universe, but need to keep in mind that it's always possible to find a new way of looking at things. pete
In astronomy - the definition in the OP - there must be some kind of assumption concerning how the 'laws of physics' work, because we cannot (per the OP) 'go there' to conduct in situ experiments^.

So, with some such assumption in hand (maybe we need to work on it a bit, to nail it down carefully), can we say that we know dark matter is both cold and non-baryonic? If not, why not?

Can we say that we know SgrA* is a super-massive black hole, of mass ~4 million sols? If not, why not?

Can we say that we know that {star} has a surface gravity of X, radius Y, mass Z, and composition {details, including uncertainties}? If not, why not?

^ that's one reason why I set the scope as it is - we can, realistically, go to most parts of the solar system and conduct in situ experiments.

Ken G
2010-Jul-14, 08:08 PM
More specifically, how does knowing that this critter is an individual of species X, within the phylum Y, differ from knowing that this star is an M dwarf?I see, you are interested in the issue of how direct are our observations. With critters, we can watch them, inspect them, even dissect them. There are a wide range of avenues we can choose from to collect information, and many of them have essentially no level of interpretation (if it has hair, or doesn't have hair, is something that is more or less directly knowable!). But for distant objects, we get light at different frequencies, and if we are lucky, we get an image, and that's it. To connect with things we can manipulate in the lab, this requires some degree of inference, and introduces a full range of variation of how direct is our knowledge, and how much we need to make inferences from a self-consistent model. So there's an added dimension to just the quality of the brand (the "national brands" vs. "supermarket brands", which have more to do with the standard of inference that is required and the record of reliability that model has provided), which is how easily the answers could change if something fundamental changed in our self-consistent model. It is not a guarantee of anything once we determine what is a national brand and what is generic-- for millennia, the geocentric model was certainly a national brand, complete with advertising and loyal customers. But it was not direct information, it required an inference that turned out to be wrong (same for Enron stock). So the quality of the brand is not the same thing as the certainty of the knowledge.

Instead, all knowledge is provisional on the validity of certain other knowledge-- like an interconnected ecology. If a part is removed, the consequences percolate throughout, and a whole subsection of that ecology might go with it, while other areas are unaffected. So we need to track not only how much we'd be willing to bet something is true, which actually doesn't mean a whole lot, we also have to track what else depends on that being true, and how easily the whole house of cards comes down if a few key cards are taken away. Sounds like the European economy, doesn't it?

trinitree88
2010-Jul-14, 08:28 PM
In astronomy - the definition in the OP - there must be some kind of assumption concerning how the 'laws of physics' work, because we cannot (per the OP) 'go there' to conduct in situ experiments^.

So, with some such assumption in hand (maybe we need to work on it a bit, to nail it down carefully), can we say that we know dark matter is both cold and non-baryonic? If not, why not?

Can we say that we know SgrA* is a super-massive black hole, of mass ~4 million sols? If not, why not?

Can we say that we know that {star} has a surface gravity of X, radius Y, mass Z, and composition {details, including uncertainties}? If not, why not?

^ that's one reason why I set the scope as it is - we can, realistically, go to most parts of the solar system and conduct in situ experiments.

Nereid. 1. I'm not sure I'm giving up on baryonic matter being dark until the successor to the FUSE mission checks out hot gas in X-rays...SEE:http://science.nasa.gov/missions/fuse/ and I'm waiting for a more sensitive report on the ratio of molecular to atomic hydrogen with newer instrumentation.

2. I won't get into a brawl over Sgr*A until they put a couple of 10 meter telescopes on the moon, which won't happen soon. I'd like to see better resolution of the images as past history has shown that better resolution keeps turning up more unexpected structures.

3. I think the rationale for [star Q] with surface gravity X, radius Y, mass Z, and composition #*&*(%)^^#)*^#....has been one of the crowning achievemants of the scientific method. No argument and kudos to the community that did it.

pete

Nereid
2010-Jul-14, 10:17 PM
More specifically, how does knowing that this critter is an individual of species X, within the phylum Y, differ from knowing that this star is an M dwarf?I see, you are interested in the issue of how direct are our observations.
That's one aspect; another I smuggled in (and you seem to have missed it).


With critters, we can watch them, inspect them, even dissect them. There are a wide range of avenues we can choose from to collect information, and many of them have essentially no level of interpretation (if it has hair, or doesn't have hair, is something that is more or less directly knowable!).
But how do we know that the critter belongs to species X, in phylum Y?

How do we know that this star is an M dwarf?

Other than that we may be able to cut our poor critter up (or not; what if the only thing we have, of the critter, is a fossilised bone?), how does the knowing differ?


But for distant objects, we get light at different frequencies, and if we are lucky, we get an image, and that's it. To connect with things we can manipulate in the lab, this requires some degree of inference, and introduces a full range of variation of how direct is our knowledge, and how much we need to make inferences from a self-consistent model.
So, astronomical knowledge - of the restricted kind referenced in the OP - is (entirely?) dependent upon the building of self-consistent models?

And, as a corollary, astronomical knowledge can come in very different flavours, or strengths?


So there's an added dimension to just the quality of the brand (the "national brands" vs. "supermarket brands", which have more to do with the standard of inference that is required and the record of reliability that model has provided),
An ironic aside: many pairs of products differ only in the brand (and price!) - they have the same ingredients and are produced by the same factory (even on the same production line!).

(to be continued)

Nereid
2010-Jul-15, 12:15 AM
(continued)
[...]

Instead, all knowledge is provisional on the validity of certain other knowledge-- like an interconnected ecology. If a part is removed, the consequences percolate throughout, and a whole subsection of that ecology might go with it, while other areas are unaffected. So we need to track not only how much we'd be willing to bet something is true, which actually doesn't mean a whole lot, we also have to track what else depends on that being true, and how easily the whole house of cards comes down if a few key cards are taken away. Sounds like the European economy, doesn't it?
If I've understood you, astronomical knowledge is no different than any (science-based) knowledge; is that right?

Nereid
2010-Jul-15, 12:24 AM
Nereid. 1. I'm not sure I'm giving up on baryonic matter being dark until the successor to the FUSE mission checks out hot gas in X-rays...SEE:http://science.nasa.gov/missions/fuse/ and I'm waiting for a more sensitive report on the ratio of molecular to atomic hydrogen with newer instrumentation.
You seem to be saying that it's the strength of the consensus which counts; as long as someone, even fringe?, has a testable hypothesis based on a non-consensus model, we do not know stuff.

If so (and please chime in if not), how does this make astronomical knowledge different from any other (science-based) knowledge?


2. I won't get into a brawl over Sgr*A until they put a couple of 10 meter telescopes on the moon, which won't happen soon. I'd like to see better resolution of the images as past history has shown that better resolution keeps turning up more unexpected structures.

3. I think the rationale for [star Q] with surface gravity X, radius Y, mass Z, and composition #*&*(%)^^#)*^#....has been one of the crowning achievemants of the scientific method. No argument and kudos to the community that did it.

pete
So why accept stuff about a star - which may be quite unresolved - and not about SgrA*?

How is the knowing different?

When did the knowing about a star become a crowning achievement? Why is there no expectation of unexpected surprises wrt an unresolved star, but there is wrt SgrA*?

baric
2010-Jul-15, 03:10 AM
"What can we know" from astronomy is, to me, one of the most fascinating aspects of the field. And it is also one of the fastest ways for me to glaze over the eyes of my non-enthusiast friends when I try to explain this particular fascination.

See that tiny point of light up there? This is what we can know from staring at it...

It's a star. Its mass. Is it moving towards us, or away? How fast it is moving. How old it is. How large it is. How hot it is. What is in its core and maybe when it might blow up.

Does it have any companion stars? If so, the same information about them.

We can tell if it is surrounded by a dust cloud, and if proto-planets are forming in that dust cloud.

Does it have any planets around it? If so, we can determine the approximate mass and length of orbit them, starting with the largest and innermost.

If some of these planets pass in front the star, we can estimate the planet's diameter and density and therefore infer its composition. We can analyze the composition of atmosphere. We can then estimate its temperature and the infer large-scale climate effects.

If there is an oxygen-rich atmosphere indicating Earth-like life, we can see that as well. With bigger telescopes, we'll see even more.

All from staring at a tiny point of light in the sky.


I LOVE ASTRONOMY.

Ken G
2010-Jul-15, 05:36 AM
(continued)
If I've understood you, astronomical knowledge is no different than any (science-based) knowledge; is that right?I'd say the difference is in degree, rather than in kind. There are still "national brands" in astronomy, ideas that seem reliable and are very widely accepted, and on which hang a lot of other things we believe to be true. But the "other dimension" of certainty that crops up, the directness or indirectness dimension, is often quite a bit more indirect for astronomy. That means astronomical knowledge tends to be more like a house of cards than in other branches of physics-- the indirectness means the knowledge relies on a self-consistent paradigm of some kind, and a shift in the paradigm can shift the interpretations of a wide array of observations without changing anything about the observations themselves (as happened, most famously, with the shift from a geocentric to a heliocentric model of our solar system).

That can certainly happen in physics too, so I'm not sure how much different this makes astronomy, but astronomy seems a bit more fragile. For example, distance scales in astronomy are based on a kind of "ladder", where all the farther distances are scaled to some closer distance, and if you change the closer distance (as happened to Hubble), you instantly change all the farther distances, without doing a thing to any of those farther observations. But physics also contains examples of paradigms we use to draw inferences and interpret observations, where a shift in the paradigm shifts the interpretation of a wide class of observations. For example, the current paradigm is that there is no aether, and coordinates that make the speed of light constant are the ones that connect with certain rationally expected axioms about the behavior of rigid objects when you rotate them and so forth (specifically, that rulers stay the same length when turned or moved slowly from place to place). However, if tomorrow some new observation forces us to conclude that there is an aether, and forces us to abandon the hope that rigid objects would possess those seemingly natural properties, then not a single observational outcome of all past experiments would need to be altered one iota-- they would just admit to a totally new interpretation.

Ken G
2010-Jul-15, 05:45 AM
"What can we know" from astronomy is, to me, one of the most fascinating aspects of the field. And it is also one of the fastest ways for me to glaze over the eyes of my non-enthusiast friends when I try to explain this particular fascination.And, I fear you'll find, that also holds for many on this forum! This thread should attract a small subset that share your fascination.


All from staring at a tiny point of light in the sky.
I start every introductory course on astronomy with the observation that of all the crazy crackpot theories you may have ever heard, that thunder is God bowling, or that stars are holes in a dome surrounded by a great fire, what must be viewed as by far the craziest of all is that unusually intelligent apes should be able to sit on the surface of a single randomly placed planet and figure out the internal operations of distant stars, and map out the entire history of the universe they inhabit. Yet the flip side of this amazement of how far we've come is a sober caveat that we have shown, throughout our history, a proclivity to overestimate how much we actually do have right, versus how much is just currently working for us. Still, those things in your list certainly do seem pretty solid...

astromark
2010-Jul-15, 06:04 AM
What are you saying... ? we know very little. No, We know lots of stuff... and very little of it is untested.

By imposing directly the constraints of science. Challenging. Testing. Building models. Pushing aside the unknown...

By laboratory tests we know of the spectrographic scale... We know the emission lines to look for., and have.

We know of red shift and, blue., and what that might tell us... How to look for what we do not see... x-ray and info-red...

How to detect a wobble in tracking which could signal the presence of other objects..

or the sudden dimming of a planetary body orbiting., or the gravitational lensing of a far more distant object...

This is a argument without substance... I call it science. What do you call it ? Please do not say faith... I will implode.

Ken G
2010-Jul-15, 12:17 PM
Of course it's science, everything we are talking about is science. And of course there are many things that we count ourselves as knowing that have significant validity to them. I'm just saying that scientific knowledge is not always what people imagine it is, and even a fairly cursory review of scientific history makes that point about as clearly as any truth of science. Yet of all the truths of science, that's the one that is most conveniently ignored. In science, knowledge is the "awareness of what works," which has an uneasy relationship with "truth." So when we ask "what do we know", the scientific version of that question is "what is currently working for us."

Science makes idealized models, always, no exceptions, and as far as science is concerned, a "true" model, like the idea that stars are balls of hot gas, is a model that works extremely well. Some models work only approximately well, others work spectacularly well-- is there a line you can draw and say that this is the moment the model became the truth? No, there is no such line. All you can do is keep track of the assumptions that the paradigm relies on, and how directly established are those assumptions, and what could change the model if what assumption changed, or if the accuracy target changed, or the domain of some approximation changed. It's hard work to carry around one's assumptions, so hard that scientists generally get a bit lazy about it and set them down-- and just imagine what they have is truth instead. It's lighter.

Nereid
2010-Jul-15, 12:45 PM
If I've understood you, astronomical knowledge is no different than any (science-based) knowledge; is that right?I'd say the difference is in degree, rather than in kind. There are still "national brands" in astronomy, ideas that seem reliable and are very widely accepted, and on which hang a lot of other things we believe to be true. But the "other dimension" of certainty that crops up, the directness or indirectness dimension, is often quite a bit more indirect for astronomy. That means astronomical knowledge tends to be more like a house of cards than in other branches of physics-- the indirectness means the knowledge relies on a self-consistent paradigm of some kind, and a shift in the paradigm can shift the interpretations of a wide array of observations without changing anything about the observations themselves (as happened, most famously, with the shift from a geocentric to a heliocentric model of our solar system).

That can certainly happen in physics too, so I'm not sure how much different this makes astronomy, but astronomy seems a bit more fragile. For example, distance scales in astronomy are based on a kind of "ladder", where all the farther distances are scaled to some closer distance, and if you change the closer distance (as happened to Hubble), you instantly change all the farther distances, without doing a thing to any of those farther observations. But physics also contains examples of paradigms we use to draw inferences and interpret observations, where a shift in the paradigm shifts the interpretation of a wide class of observations. For example, the current paradigm is that there is no aether, and coordinates that make the speed of light constant are the ones that connect with certain rationally expected axioms about the behavior of rigid objects when you rotate them and so forth (specifically, that rulers stay the same length when turned or moved slowly from place to place). However, if tomorrow some new observation forces us to conclude that there is an aether, and forces us to abandon the hope that rigid objects would possess those seemingly natural properties, then not a single observational outcome of all past experiments would need to be altered one iota-- they would just admit to a totally new interpretation.
So, with the exception of being, to some extent, somewhat more indirect, astronomical knowledge comes in the same range of greys as non-astronomical knowledge; some things we know with considerably certainty (whether astronomical or not), others with much less certainty. Or, perhaps, instead of certainty, there, what? consensus? reliability? acceptance? how-much-else-depends-on-it?

What are the key dimensions of scientific knowledge? And how, apart from directness, does astronomical knowledge differ?

I'd also like to examine "observations"; do these not depend, critically, on self-consistent models? Models built using self-consistent things from physics, or chemistry, say? Don't the observations themselves also come in varying shades of grey? How can "observations" - as astronomical knowledge - be cleanly separated from "interpretations"?

Ken G
2010-Jul-15, 01:35 PM
So, with the exception of being, to some extent, somewhat more indirect, astronomical knowledge comes in the same range of greys as non-astronomical knowledge; some things we know with considerably certainty (whether astronomical or not), others with much less certainty.That certainly seems reasonable to me.

Or, perhaps, instead of certainty, there, what? consensus? reliability? acceptance? how-much-else-depends-on-it? There's the rub! We are much better at "knowing" things than we are at knowing what it means to know!


What are the key dimensions of scientific knowledge? And how, apart from directness, does astronomical knowledge differ?
The key dimension must be self-consistency. We have a set of questions, and we find a unified scheme to answer them all, to some reasonable accuracy (or even to some spectacular accuracy, in some cases). It all "hangs together", so we call it "knowledge." Of course there are questions on the "fringe" that still come up "?", but we hope that we will answer them by adding to our knowledge, rather than by going back and replacing something. How often do we add, and how often replace? I've never tried to do a careful survey, and I'd like to think that most of the future of our knowledge will be more like adding and less like replacing, but then-- everybody always thinks that.

I'd also like to examine "observations"; do these not depend, critically, on self-consistent models? Models built using self-consistent things from physics, or chemistry, say?Yes, one cannot really separate the observations from the models they test, but we can (and must) form hierarchies of models, which are like the hierarchies of "directness"-- there are models we are so confident in (and some built right into the assumptions of science, like the model of "objectivity") that when we invoke them, we call it a "direct observation". So in effect, some models are used to test others. Is there a formula for doing that? No, it's just "common sense" or "logic" or some such agreed-on function of intelligence that we think we know what we are doing. We rarely get fundamentally shocked, but it still happens.


Don't the observations themselves also come in varying shades of grey? How can "observations" - as astronomical knowledge - be cleanly separated from "interpretations"?No formula for it-- it's more the function of peer review than anything else. So it's a matter of consensus on that one-- those "national brands."

George
2010-Jul-15, 05:00 PM
Excellent discourse, folks.

Extending (off a cliff, probably) Ken's use of "dimension", can we formulate an astroepistemological approach?

Here's my amateur round 1... :)

FQ = DQ * OD * TB * CD

FQ ~ Final Quality

Data Quality ~ What is the sigma value of the celesetial light data itself?

Objectivity Degree ~ What is the relative amount of objectivity that can be applied? E.g. GRB vs steady, white sunlight.

Theoretical Base – “Brand Respect” (respect for the theory itself). This may be two-fold: How does our information correlate with other information within the given applied theory? How strong are the arguments supporting the theory as it applies to our case (e.g. other lines of evidence)?

Categorical Degree ~ To what degree can we appropriately categorize the conclusion drawn from our information (i.e. Nereid's Species X, Phylum Y)? How accurate is our mapping for the information given?

[All values are 0 to 1]

Ken G
2010-Jul-15, 05:32 PM
I'd say you are bringing in a lot of the key themes, but it's hard to be either complete or quantitative about it, so I think that ultimately any attempt to apply something as formal as a formula will probably not be flexible enough. It suggests that we can, at any given time, accurately assess our own certainty about the various conclusions we draw, and I don't think that's actually possible. I wonder, if we could go back in time to Aristotle's day, or Newton's day, and give those geniuses a list of 10 important paradigms related to their work, and ask them to rank order them in terms of how valid we would view them today, would their ordering have any relation to the actual situation? I'm not sure it would, but it would be an interesting exercise. But I guess we can kind of tell when we're out on a limb, in many cases, it's just that the tendency when we get out there is to say "it may be a problem, but it'll probably work itself out," moreso than we probably have any right to actually think that.

It's a bit like weather forecasters-- when they are wrong, they never say "oops", nor give any apology, they just carry on and say "it was a forecast for you to use as you choose, it was never a promise."

George
2010-Jul-15, 06:57 PM
I'd say you are bringing in a lot of the key themes, but it's hard to be either complete or quantitative about it, so I think that ultimately any attempt to apply something as formal as a formula will probably not be flexible enough. Agreed. Yet its intent is not to produce a hard value but a relative one, yet one that clearly indicates that a filltration process has been applied so tht we all have a better appreciation for the credibility of what is presented. Can we produce something at least partially qualitative so we have a better idea on whether to bet the cat or the whole house?


It suggests that we can, at any given time, accurately assess our own certainty about the various conclusions we draw, and I don't think that's actually possible. Yes, that is always the case but it may be less the case if some formality is applied to the conclusive approach itself.


I wonder, if we could go back in time to Aristotle's day, or Newton's day, and give those geniuses a list of 10 important paradigms related to their work, and ask them to rank order them in terms of how valid we would view them today, would their ordering have any relation to the actual situation? I'm not sure it would, but it would be an interesting exercise. Indeed, it would be both interesting and valuable in determining which set of filters give us the best result, though we still have to be mindful to recognize the potential loss that can come from over filtration -- large gold nuggets can be accidentally discharged if the screen size is too small.

But let's take your good advice and jump into the past -- Galileo's day and the heliocentric conclusion he asserted.

Data Quality ~ He obtained relatively solid evidence based on the size and resolution of his better telescopes. Being an artist, his drawings were quite accurate, at least compared to what I have attempted.

As for his telescopes, he reportedly trashed over 90% of the hundreds he made due to poor quality.

It was obvious to him even with his smaller telescopes that Jupiter’s moons were probably not stars based on his ability to see their alignment (fit) to the ecliptic.

Objectivity Degree ~ Unfortunately for him, horrible, in the first decade at least. Only a few telescopes existed that had the ability to see what he saw. [Harriot seems to have had one of respectable quality, but was independent of Galileo and neither shared info.] His public demonstrations were few because of the negative results from poor weather, bad seeing, immature viewing skills of the general public, etc. He refused Kepler a telescope, IMO, because he was aware that Kepler had poor vision. [Kepler kept trying to get one for obvious reasons and finally received one from a friendly dignitary. Kepler confirmed Galileo’s claim.]

Theoretical Base ~ Big problem here. He was ATM up until his discovery and verification of the phases (gibbous and crescent) for Venus and Mercury. [Venus was uncooperative for most of 1610.]

Once Ptolemy's model was knocked-out, he still had problemse since the “theoretical base” was still in the hands of the theologians and, even more so, the philosophers in academia who all had been united against Galileo since they were still stuck on Aristotle. [Time and evidence changed all this, of course. There were, originally, many more clergy favorable to Galileo's Copernican view than the peripatetic philosophers.]

Categorical Degree ~ Fairly strong here, I think. The Medician Moons could be categorized as actually being moons of Jupiter and not some other stellar phenomena, though no absolutes were assured. The sunspots [true nature] were not easy to nail down and Apelles (Scheiner) was very argumentative against Galileo’s conclusion that they weren’t transits of orbiting objects. [Galileo stated he was unsure if the spots were on the Sun’s surface or if they were an atmospheric event similar to clouds, yet Solar spots they were.]

Thus, our low FQ value might have served him well when he chose the forks in the road that thwarted the opposing theological exegesis he assumed would change in light of his discoveries and eloquence.

Am I close?

trinitree88
2010-Jul-15, 07:17 PM
Excellent discourse, folks.

Extending (off a cliff, probably) Ken's use of "dimension", can we formulate an astroepistemological approach?

Here's my amateur round 1... :)

FQ = DQ * OD * TB * CD

FQ ~ Final Quality

Data Quality ~ What is the sigma value of the celesetial light data itself?

Objectivity Degree ~ What is the relative amount of objectivity that can be applied? E.g. GRB vs steady, white sunlight.

Theoretical Base – “Brand Respect” (respect for the theory itself). This may be two-fold: How does our information correlate with other information within the given applied theory? How strong are the arguments supporting the theory as it applies to our case (e.g. other lines of evidence)?

Categorical Degree ~ To what degree can we appropriately categorize the conclusion drawn from our information (i.e. Nereid's Species X, Phylum Y)? How accurate is our mapping for the information given?

[All values are 0 to 1]


Ah! The Drake equation is now being supplemented with the GEORGE Equation. Like it. pete

George
2010-Jul-15, 07:17 PM
It's a bit like weather forecasters-- when they are wrong, they never say "oops", nor give any apology, they just carry on and say "it was a forecast for you to use as you choose, it was never a promise."
Yet, this is a great analogy, I think. When meteorologists present a forecast do they not use a similar filtration system to give us a FQ (i.e. chance of something; percentage) that is always less than or equal to 1 (100%)? Recently they showed the "spaghetti trails" for the impending storm coming from Can Cun. These computer models take the atmospheric data and present a degree of quality in their projections. [In this case, all the trails headed for the middle of the Texas coast, though one trail ran straight to New Orleans -- a historical bias? ;)].

George
2010-Jul-15, 07:18 PM
Ah! The Drake equation is now being supplemented with the GEORGE Equation. Like it. pete
I hope it has efficacy, though the use of my name does me great injustice! ;)

[I had thought about the Drake equation, but I was too lazy to do subscript. :)]

Nereid
2010-Jul-15, 07:50 PM
Interesting ...
Excellent discourse, folks.

Extending (off a cliff, probably) Ken's use of "dimension", can we formulate an astroepistemological approach?

Here's my amateur round 1... :)

FQ = DQ * OD * TB * CD

FQ ~ Final Quality

Data Quality ~ What is the sigma value of the celesetial light data itself?
Curiously, this is not eternal; you can re-analyse historical data and get different sigmas!


Objectivity Degree ~ What is the relative amount of objectivity that can be applied? E.g. GRB vs steady, white sunlight.
I don't understand this; can you say more please?


Theoretical Base – “Brand Respect” (respect for the theory itself). This may be two-fold: How does our information correlate with other information within the given applied theory?
I.e. a self-consistent model?


How strong are the arguments supporting the theory as it applies to our case (e.g. other lines of evidence)?
This may not be helpful. For example, Newtonian gravity is dead; long live GR! Yet for a wide range of circumstances/situations (or regions of parameter space) it doesn't matter which you use, they're both equally appropriate.


Categorical Degree ~ To what degree can we appropriately categorize the conclusion drawn from our information (i.e. Nereid's Species X, Phylum Y)? How accurate is our mapping for the information given?
I don't understand this either; help!



[All values are 0 to 1]
Good.

George
2010-Jul-15, 08:48 PM
Curiously, this is not eternal; you can re-analyse historical data and get different sigmas! The use of sigma is applied to the quality of the data itself. This factor comes from the margin of error or deviation of the measurments taken?

Why is it that the Hi-Z team beat Saul Perlmutter's team in the announcement of the accelration of space? The Hi-Z team had perhaps half the number of supernova's to work with, yet they put more effort in the filter work in order to improve their sigma value, which produced a better fit with about half the data.

I suppose the corollary would be "garbage in, garbage out".


I don't understand this; can you say more please? Objectivity here is meant to address how much more the measurements can be done objectively. In the case of GRBs, the burst time is too quick to get many observers on-board. At the other extreme, the diurnal regularity and consistency of sunlight -- did anyone catch the prior heliochromological lick? (sorry, can't help it, ; I recognize the wise won't answer.) ) -- affords a vast amount of objective data.


I.e. a self-consistent model? I may be missing what you're asking, but every theory must be self-consistent, yet no two theories are equal in their prowess. For another extreme range in examples, compare GR to the Multiuniverse "theory", *cough*. If the astronomical claims we are making fit nicely within a well-established theory -- the more lines of evidence our established theory has the better -- then we are likely on a firmer base with a higher credibility factor than if our claim only fits some fringe hypothesis.


This may not be helpful. For example, Newtonian gravity is dead; long live GR! Yet for a wide range of circumstances/situations (or regions of parameter space) it doesn't matter which you use, they're both equally appropriate. Yes, both are theories that are highly respected. [It's a safe bet that Newton's laws are taught far more than Einstein's, or are my fellow engineers being suberverted by physicists these days. *wink*]

So if, for instance, we are obtaining exoplanetary data relating to low speed motions, and our data is consitent with Newtonian theory, then we have a TB factor that should be high. If, on the other hand, our data seems to work only with iron star ideas, we should have a low TB factor.

As far the Categorical Degree or factor, actually I was hoping I was tossing the ball to you on this one - I don't know what a phylum is? :) Nevertheless, my attempt was to address the more difficult and subjective aspect of a claim: how well does it fit into the little boxes we call categories? This seems redundant to the TB factor, but I think there is more to it than that, so I tossed it in.

Since you are using a biological example, consider the "finches" that Darwin sent to England. It became clear to him just how unclear the scientists were in categorizing these birds as finches. Darwin soon recognized that the line between varieties and species was very vague, which also contributed, no doubt, to his extensive views on natural selection. The greater our ability to categorize what is found, the greater the overal chance we have at supporting our claim.

I admit I need some help with all these, but this last is messier still. [I thought of calling it a Fudge Factor, but I'm dieting. ;)]

Nereid
2010-Jul-15, 09:24 PM
Curiously, this is not eternal; you can re-analyse historical data and get different sigmas!The use of sigma is applied to the quality of the data itself. This factor comes from the margin of error or deviation of the measurments taken?

Why is it that the Hi-Z team beat Saul Perlmutter's team in the announcement of the accelration of space? The Hi-Z team had perhaps half the number of supernova's to work with, yet they put more effort in the filter work in order to improve their sigma value, which produced a better fit with about half the data.

I suppose the corollary would be "garbage in, garbage out".
My point is that the contemporary estimate of the sigma may be different from a later, independent, analysis, using the same data.

The 1919 eclipse data for example: at the time, Eddington et al. declared an unambiguous signal; for many years afterwards (and even today) anti-relativists have claimed (with varying degrees of credibility) that the 1919 data were not clean enough to distinguish GR from Newton (IIRC a relatively recent re-analysis, of some of the data only, reached a Solomon-like conclusion: by today's standards, the data favour GR; by the data analyses of the time, they probably didn't).

Another one: Miller and the aether.



I don't understand this; can you say more please?

Objectivity here is meant to address how much more the measurements can be done objectively. In the case of GRBs, the burst time is too quick to get many observers on-board. At the other extreme, the diurnal regularity and consistency of sunlight -- did anyone catch the prior heliochromological lick? (sorry, can't help it, ; I recognize the wise won't answer.) ) -- affords a vast amount of objective data.
That seems more like repeatability, or verifiability.

A million sunrises may be a million instances of one phenomenon, or a million independent phenomena, or ...



I.e. a self-consistent model?
I may be missing what you're asking, but every theory must be self-consistent, yet no two theories are equal in their prowess. For another extreme range in examples, compare GR to the Multiuniverse "theory", *cough*. If the astronomical claims we are making fit nicely within a well-established theory -- the more lines of evidence our established theory has the better -- then we are likely on a firmer base with a higher credibility factor than if our claim only fits some fringe hypothesis.
What, then, does "How does our information correlate with other information within the given applied theory?" mean?

And doesn't this just import another, subjective, category ("well-established" vs fringe)?



This may not be helpful. For example, Newtonian gravity is dead; long live GR! Yet for a wide range of circumstances/situations (or regions of parameter space) it doesn't matter which you use, they're both equally appropriate.
Yes, both are theories that are highly respected. [It's a safe bet that Newton's laws are taught far more than Einstein's, or are my fellow engineers being suberverted by physicists these days. *wink*]

So, if we are obtaining exoplanetary data relating to low speed motions, and our data is consitent with Newtonian theory, then we have a TB factor that should be high. If, on the other hand, our data seems to work only with iron star ideas, we should have a low TB factor.

As far the Categorical Degree or factor, actually I was hoping I was tossing the ball to you on this one - I don't know what a phylum is? :) Nevertheless, my attempt was to address the more difficult and subjective aspect of a claim: how well does it fit into the little boxes we call categories? This seems redundant to the TB factor, but I think there is more to it than that, so I tossed it in.

Since you are using a biological example, consider the "finches" that Darwin sent to England. It became clear to him just how unclear the scientists were in categorizing these birds as finches. Darwin soon recognized that the line between varieties and species was very vague, which also contributed, no doubt, to his extensive views on natural selection. The greater our ability to categorize what is found, the greater the overal chance we have at supporting our claim.

I admit I need some help with all these, but this last is messier still. [I thought of calling it a Fudge Factor, but I'm dieting. ;)]
Ah, species and phyla! :)

The name has remained the same, over quite a long time (over a century), but its meaning has changed, rather drastically. So the critter remains the same, but the species (and even phylum!) classification may have changed.

George
2010-Jul-15, 10:08 PM
My point is that the contemporary estimate of the sigma may be different from a later, independent, analysis, using the same data.

The 1919 eclipse data for example: at the time, Eddington et al. declared an unambiguous signal; for many years afterwards (and even today) anti-relativists have claimed (with varying degrees of credibility) that the 1919 data were not clean enough to distinguish GR from Newton (IIRC a relatively recent re-analysis, of some of the data only, reached a Solomon-like conclusion: by today's standards, the data favour GR; by the data analyses of the time, they probably didn't). Yes, this is a nice example of the need to give the reliability of the data itself a value factor. IIRC, Eddington's data was marginal even in his eyes since he had so few photos to work with. The question at hand was what was the margin of error on those photos. He certainly had a margin of error in his measurements found on those few plates, and I suspect this would have presented a marginal DQ factor, though his expertise in understanding all the other variables may have given him the confidence to favor Einstein, whom he was no adversary to. :)


Another one: Miller and the aether. You mean Michelson-Morley? Another good example, I would guess. Their interferometry results would present a high quality factor for their data. The claim by others that it made the aether superfluous will still be valued by the other factors involved, which may be the ones presented. [The aether may still be out there but drug around by large masses like planets and star systems, though unlikely. :)]


A million sunrises may be a million instances of one phenomenon, or a million independent phenomena, or ... A million observers is always better than one. :)

[Nuts, the 5 o'clock madness has cut my post time. I'll respond the rest later. sorry]

astromark
2010-Jul-15, 10:18 PM
Is a zebra a black horse with white markings or a white horse with black markings.. ?

No you idiot mark... Its a Zebra. Oh...

You mean a quadruped Herbivore ... No, I mean a Zebra. sigh.... but what is a zebra ?

A variant quarter horse type animal of the African highlands...

"Astronomy, What can we know. ?" Lots. I would argue.

The correlation of many theory to affirm a hypothesis... confirmed.

Are we trying to argue something else ?

EDG
2010-Jul-15, 11:01 PM
I can't help but feel that discussions like this are just so much navel-gazing that misses the point. How much we can know about astronomy isn't really an issue - it's not going to change our quest to learn more, is it? Are we to enter into making observations about astronomy with preconceptions about what we can find? I'd rather just go out and look and see what's out there than fuss about the nature and extent of the knowledge we'd find. We'll learn more that way, and then we'll figure out ways to learn even more.

Ken G
2010-Jul-15, 11:58 PM
Agreed. Yet its intent is not to produce a hard value but a relative one, yet one that clearly indicates that a filltration process has been applied so tht we all have a better appreciation for the credibility of what is presented. Can we produce something at least partially qualitative so we have a better idea on whether to bet the cat or the whole house?
That's very hard-- what would it take for you to "bet the house" that some physical theory we now cherish will survive in a nearly unchanged form a thousand years from now? I'm really not sure I'd be willing to bet my house on any of our current theories. Sure an accurate theory won't suddenly stop being accurate, but it might be thought of as kind of a naive limit of a much more sophisticated, and totally different, theory-- like Newton's gravity and Einstein's (would Halley have bet his house that Newton's theory would continue to be held as our best model of gravity? He might have bet his house that his comet would return, but that's a prediction-- not a theory. Predictions can be made in the absence of a theory, the issue is why did he think that comet would return-- just because it always did, or because Newton created a convincing model that it would?)


Thus, our low FQ value might have served him well when he chose the forks in the road that thwarted the opposing theological exegesis he assumed would change in light of his discoveries and eloquence.

Am I close?Sounds pretty good to me. Certainly it's not quite the story we tell today!

Ken G
2010-Jul-16, 12:03 AM
I can't help but feel that discussions like this are just so much navel-gazing that misses the point. Ah, my favorite word, "navel-gazing". Does that word really mean anything? Sounds like another way to say "introspection", just spun to sound negative. Obviously, any effort to ascertain the meaning of what it is to know something is going to require introspection, so trying to make that sound negative is bemusing.

I'd rather just go out and look and see what's out there than fuss about the nature and extent of the knowledge we'd find.Well, notwithstanding more pointless spin words like "fuss", you are certainly welcome to make that personal choice. Others might feel that f you don't even understand what you know now, you certainly aren't going to understand what the new discoveries mean either.


We'll learn more that way, and then we'll figure out ways to learn even more.And that is a perfectly valid choice for anyone to make-- they can decide to try and learn, without understanding what learning is. Others can choose differently. This thread sounds like it is more attuned to the latter group, so it's not clear to me what contribution it makes to complain about their choices.

Nereid
2010-Jul-16, 12:34 AM
I can't help but feel that discussions like this are just so much navel-gazing that misses the point. How much we can know about astronomy isn't really an issue - it's not going to change our quest to learn more, is it?
I hope not!


Are we to enter into making observations about astronomy with preconceptions about what we can find?
We do that all the time; sometimes it's quite open, and conscious, other times it's not.

Most of the time - hopefully - it doesn't matter, because the data is taken by mindless machines (very carefully calibrated ones), and analysed with software routines that are also carefully calibrated (and published).

However, that is not always so ... the 1919 eclipse expedition is a very good example.


I'd rather just go out and look and see what's out there than fuss about the nature and extent of the knowledge we'd find. We'll learn more that way, and then we'll figure out ways to learn even more.
That may be so.

However, if people - other than you - enjoy discussing epistemology, wrt astronomical knowledge, who cares?

forrest noble
2010-Jul-16, 01:13 AM
Nereid,

I've "heard" you consider this line of questioning before and I think the general idea of it is of interest to many.


Do we know, for example, that other stars are composed, largely, of hydrogen and helium (most of them anyway)?

The existence of dark matter still must be considered hypothetical or theoretical since there is no consensus that dark matter has ever been observed or that it really exists other than baryonic matter of some kind. The same thing applies to dark energy.


Do we know that there's a super-massive black hole at the centre of the Milky Way?

As you know black holes have a number of different definitions and related theories concerning their true nature. However, there is a vast amount of evidence that supports the assertion that there is a strong gravitational influence at the center of most galaxies that seems to be concentrated in a relatively small volume which is seemingly strong support for a "black hole" type theory of some sort.


Do we know that Geminga is a ~sol-mass object, composed almost entirely of nuclear degenerate matter?
Neutron stars are solely a theoretical entity but it would be otherwise more difficult to explain the gravitational concentration of matter and its observed gamut of radiation and magnetic influences in another way -- so it seems like a good theory to me.

Do we know that dark matter is cold, and non-baryonic?
again such ideas I think are speculative considering that dark matter as an unknown particle entity which may not exist at all. Such hypotheses are necessary and will be well considered provided they are not discussed as knowledge, which might be challenged by many.


How should we go about deciding what we know (and what we don't)?
The description of something as knowledge should follow the common definition of the word knowledge, which I think is good enough even though such "knowledge" in the future may turn out someday to be wrong. Such a definition of knowledge might be: "What is learned by education, perceptual experience, and reasoning." Knowledge of theory is still knowledge of sorts. The disagreement is generally only a matter of opinion as to whether a hypothesis or theory has graduated to the fact status or not and the degree of perceived certainty involved?


Does it even make sense to say we 'know' something, anything, about things beyond our solar system?

A cautious approach might be to refer to the theory involved but most educated people in the field, as you know, will realize the theory that you are assuming when you are discussing cosmology for instance, unless you think what you are saying could be misunderstood or challenged without added qualification.

forrest noble
2010-Jul-16, 01:16 AM
"navel-gazing?" -- you mean the phrase is not related to piercings?

George
2010-Jul-16, 03:03 AM
That's very hard-- what would it take for you to "bet the house" that some physical theory we now cherish will survive in a nearly unchanged form a thousand years from now? I'm really not sure I'd be willing to bet my house on any of our current theories. Yeah, the house is a bit much -- what does this say about your cat? :). Most bets, however, are conditional so that I would be willing to bet the house given a certain number of bowling balls falling downward at a vertical or near vertical angle... without wind... no magnets... enclosed environment, etc. This would be a relative fall, of course. Whether or not the Earth is rising or the ball is falling, I'd limit the bet to the cat, assuming our family hadn't already lost the one we had, though not from a bet. :)


Sure an accurate theory won't suddenly stop being accurate, but it might be thought of as kind of a naive limit of a much more sophisticated, and totally different, theory-- like Newton's gravity and Einstein's (would Halley have bet his house that Newton's theory would continue to be held as our best model of gravity? True, but he probably would have bet the cat, or at least the money he likely got back from Newton when he fronted him the money to publish a little book called Principia. :)

But my little confidence formula is more about a claim that is embodied or placed within a given theory, else why have a TB factor? I suppose we could expand it, or shrink it, in some fashion that helps reveal what is happening in the minds of scientists regarding how they discern credibility for or against a theory. Some of this helps me see what you and others see, though you and others have brought me further than I thought I'd ever go.


He might have bet his house that his comet would return, but that's a prediction-- not a theory. Predictions can be made in the absence of a theory, the issue is why did he think that comet would return-- just because it always did, or because Newton created a convincing model that it would?) Yes, a predicition, or rather the likelihood of a prediction coming true, is more to what I am attempting to quantify, at least to some degree.

How good was Halley's data regarding the comet?
Was there a high degree of objective information available and would be available at the time of return?
How confident was he in the path comets take? [Speaking of betting, I know Halley, after betting someone or two (Hooke?) was thrilled when he learned that Newton had a mathematical solution to elliptical paths (calculus), thus won the bet, I think. Newton couldn't find his work, which prompted Halley to encourage Newton to publish, which he did, though Halley had to front the money, IIRC.]
Could he comfortably categorize his comet within the shelfwork of astronomy during his time?

My interest is frameworking science in ways that reveal its true identity and methodology, including its limitations (both self-imposed and otherwise). Your comments have helped me understand a great deal more about how scientists see things, or should see things. Some people just know when it's going to rain, I like the meteorological odds. :)

George
2010-Jul-16, 03:39 AM
What, then, does "How does our information correlate with other information within the given applied theory?" mean? This takes our supposition, conjecture or hypothesis derived from the data and compares it to related hypothesis that are embodied in the theory at hand. For instance, a survey showing the lack of metals at z = 10 or greater distance would be a logical fit for BBT. There is a comfort level here that is of some value (pun, I think). But, admitedly, it is often when we are most comfortable when the snake finally appears.


And doesn't this just import another, subjective, category ("well-established" vs fringe)?Well, there will always be a subjective element to any hypothesis or viewpoint, but I don't see why any claim can't be compared favorably or otherwise to the existing claims already whithin the theory.


Ah, species and phyla! :) I looked it up and found it relates to verbena. ;)
[A sendero on my father's small ranch.]

astromark
2010-Jul-16, 05:34 AM
Only looking at what I have taken from something George has just said...

about some data recorded from 1919...
and we have found it necessary to modify as newer information gets us closer to the facts...

and Edg is right... lets build the robotics craft bristling with tecknowledgies and get it out there...

But then I can still see a point where just by looking at information as we see it... halves the distance...
sort of...

Warp eight Mr Sulu... engage.

Navel gazing... yes., but its my navel and its not light years away...:eh:

Disinfo Agent
2010-Jul-16, 10:27 AM
Are we to enter into making observations about astronomy with preconceptions about what we can find? I'd rather just go out and look and see what's out there than fuss about the nature and extent of the knowledge we'd find. We'll learn more that way, and then we'll figure out ways to learn even more.That sounds like a great plan, until you try it. As Nereid noted earlier, all observation requires some amount of interpretation. We always enter everything with preconceptions. Preconceptions are just assumptions, and there is no science without assumptions. I'd go even further: there is no meaningful discussion without assumptions.


But how do we know that the critter belongs to species X, in phylum Y?

How do we know that this star is an M dwarf?I recently found out (http://www.bautforum.com/showthread.php/105660-Things-Named-Poorly?p=1758680#post1758680) that taxonomic borders have been redrawn since I learned taxonomy in school.

But, although I try not to quibble over names, I suggest that classification, while an important and necessary part of science, is not what we usually understand by 'knowledge'. A reclassification doesn't tell us anything new; it only reorganises what we already knew. A hypothetical science that consisted only of classifying previous observations would be no science at all.

George
2010-Jul-16, 02:00 PM
That sounds like a great plan, until you try it. As Nereid noted earlier, all observation requires some amount of interpretation. We always enter everything with preconceptions. Preconceptions are just assumptions, and there is no science without assumptions. I'd go even further: there is no meaningful discussion without assumptions. I am pleased to assume you are correct. :) No doubt, there is a great deal of range of degree since hard facts will minimize the amount of assumption that is needed in limited circumstances.


But, although I try not to quibble over names, I suggest that classification, while an important and necessary part of science, is not what we usually understand by 'knowledge'. A reclassification doesn't tell us anything new; it only reorganises what we already knew. I am unclear on this. Why would something be reclassified if something new wasn't learned?

Ken G
2010-Jul-16, 02:29 PM
Navel gazing... yes., but its my navel and its not light years away...
That's actually rather well put!

Ken G
2010-Jul-16, 02:32 PM
A hypothetical science that consisted only of classifying previous observations would be no science at all.Right, because science isn't a Google lookup table. We need to feel that we are gaining some kind of understanding, but to understand what that even means, we need to look to the process, not just its results. If all we have any understanding of are our results, then all we have is a dogmatic clinging to the current canon of wisdom. Are there any historical analogs of that happening, and did it slow future discovery?

It seems to me epistemology is a bit like structural engineering. The structural engineer does not just look at a building and say "yup, that building is standing all right, let's go build some other ones now", which is pretty much how I interpret EDG's advice on the matter. It's true that if you want a city, you're going to have to build lots of buildings, and if some of them fall, you can always just replace them with others. But all the same, we do have structural engineers, and they are capable of looking deeply at how a structure is built, where the loads are, and what supports are dependent on what other supports, such that other kinds of questions can be answered about that building: questions like, will it sustain a hurricane, could another level be added to it, and if it needed to be taken down, where would be the most effective place to put the dynamite?

Those are the questions of epistemology-- gazing at the navel of the structures we build, rather than just going on to the next one in a kind of Darwinian "survival of the fittest" approach to architecture (or science). And despite how poorly many people understand the process of "looking under the hood" of our current science, and how frankly unnecessary it is for the daily functioning of a scientist, it would seem to be an important component of the science done by those who generate the truly startling new ways to understand our reality. Those people are always considered to be philosophers, even if they don't self-identify as such (since many people reserve the term "philosophy" for other people's philosophical views).

Disinfo Agent
2010-Jul-16, 02:47 PM
I am unclear on this. Why would something be reclassified if something new wasn't learned?You've got a point there. To tell the truth, reclassifications in science often come as the consequence of learning something new:

blue algae are now classified as bacteria because of advances in molecular biology
Pluto is no longer classified as a planet because of the discovery of the Kuiper Belt
A reclassification may also become valuable for illustrating a new insight into a field of study. The periodic table is a classic example:


The success of Mendeleev's table came from two decisions he made: The first was to leave gaps in the table when it seemed that the corresponding element had not yet been discovered. Mendeleev was not the first chemist to do so, but he went a step further by using the trends in his periodic table to predict the properties of those missing elements, such as gallium and germanium. The second decision was to occasionally ignore the order suggested by the atomic weights and switch adjacent elements, such as cobalt and nickel, to better classify them into chemical families. With the development of theories of atomic structure, it became apparent that Mendeleev had inadvertently listed the elements in order of increasing atomic number.

Wikipedia (http://en.wikipedia.org/wiki/Periodic_table)Mendeleev's periodic table didn't just reorganise existing knowledge about the chemical elements. It also pointed in the direction of new discoveries.

Pure reclassification does not bring new insights - but I must admit that instances of pure classification in science are not easy to come by. Usually there is something else going on, some broader rationale... And we're back to the importance of interpretation! :)

George
2010-Jul-16, 03:08 PM
It seems to me epistemology is a bit like structural engineering. Nice analogy.

I would assume it should also be analogous to any epistemological quantification attempt. Maybe it is:

Handbooks have not only vast amounts of data available to the structural engineer, but the data itself is very, very reliable (especially if one understands how those material property values were determined). High quality structures that the building must support will likely require quality data.

The best structure will accommodate lots of people that can crawl all over it, and it will have lots of windows to improve the objective view of the world around it.

The structure will not likely accommodate things hanging off it that are in violation of mainstream science, though not trying new things will cause it to only accommodate the inevitable mundane stuff.

The residents are well-advised to organize all their stuff, and can put their office locations, at least, on the lobby index.

:)

Ken G
2010-Jul-16, 03:11 PM
Mendeleev's periodic table didn't just reorganise existing knowledge about the chemical elements. It also pointed in the direction of new discoveries.
Excellent example-- Mendeleev approached the periodic table like a kind of structural engineer, making decisions based on attributes he wanted that structure to possess, he didn't blindly follow a recipe like ordering with atomic weight. If he hadn't thought a great deal about what the knowledge of elements ought to look like, he couldn't have discovered those improvements.

Ken G
2010-Jul-16, 03:19 PM
I would assume it should also be analogous to any epistemological quantification attempt. And like the architecture of buildings, the architecture of knowing may be a bit different when our goals are different, or for different types of "buildings." Your themes are the important ones, but there may be no single best way to combine them that suits every situation. Scientific knowledge is really a much more complicated animal then we'd like to believe, I think stars themselves are easier to understand!

George
2010-Jul-16, 03:20 PM
Pure reclassification does not bring new insights - but I must admit that instances of pure classification in science are not easy to come by. Usually there is something else going on, some broader rationale... And we're back to the importance of interpretation! :) Yes, I think I see where you are in this. It can be seen as two separate things involved since the reclassification is an effect and not a cause. Understanding the cause is the key to appreciating the reclassification, which is in agreement with what Ken is saying.

George
2010-Jul-16, 03:40 PM
And like the architecture of buildings, the architecture of knowing may be a bit different when our goals are different, or for different types of "buildings." Your themes are the important ones, but there may be no single best way to combine them that suits every situation.
My attempt to use your structural analogy and merge it into an epistemological one was outside my fun zone, but I was curious if I could get close. I would much rather have used your analogy applied to the quality of a specific claim since no two buildings are ever exactly the same and will not have the same quallity value.


Scientific knowledge is really a much more complicated animal then we'd like to believe,... Yet they are much easier to ride if you can manage to get a saddle on them. Y'all know how to handle these better than I can, hence my need for some cinching of that which is not a cinch. :) If some simple sort of structure could serve as a set of general filters to help others (eg Joe Public) consider the credibilty of any scientific claim, then so much the better. The Moan Hoax ideas, Face on Mars and the host of other wild unsubstantiated stuff might have much more trouble "smuggling in" (borrowing Nereid's cool earlier phrase) their self-serving ideas when they surreptitiously crawl up upon the shore of...(guess).... the Isle of Science.


I think stars themselves are easier to understand! I thank the stars they are easier to understand, [in my limited way]! :)

Nereid
2010-Jul-16, 06:12 PM
A bit late on classifications, and systems therefor ...

Reclassifications are agents of chaos, and of deeper knowing.

An awful lot of scientific knowledge is built on the bedrock of classifications - birds are built for flying, elliptical galaxies lack gas, all Cepheids have the same P-L relationship, and so on. And the building often keeps growing, with more and more knowledge built on top of the conclusions based on the conclusions based on ... the classifications.

Come the revolution, and a lot falls down.

What's built from the rubble is, (almost always? sometimes?) better in so many ways, but it means that older scientific material is not so easy to read and understand.

So, back to astronomy.

WOBAFGKM (+RNS) - a classification scheme for stars; any revolutions past? hints of revolutions to come?

How do we *know* {this star} is a K dwarf? a white dwarf? a neutron star? a binary containing a black hole?

Will it still be such a star next year? in a decade's time? a century's? (assuming it doesn't go supernova)?

How intricately does astronomical knowledge depend on classifications?

And what are classification schemes anyway?

George
2010-Jul-16, 08:10 PM
An awful lot of scientific knowledge is built on the bedrock of classifications - birds are built for flying, elliptical galaxies lack gas, all Cepheids have the same P-L relationship, and so on. And the building often keeps growing, with more and more knowledge built on top of the conclusions based on the conclusions based on ... the classifications. Yes. I suppose it helps, for me at least, to see classifications as an organized effect that comes from some sort of causal relationship, even if it remains unstated. Ellipitcal galaxies are more blobby in shape, older, and about out of gas. [Good grief, no wonder my behavior gets too elliptical around here!] First came the shapes (e.g. Wolf), then the lack of gas (Curtis?), which, with bigger scopes, allowed a modern classification scheme (Hubble), then their age, which, in turn, helped in the explanation for their shape and lack of gas.

Once what is observed is orgainized into classifications, then the hunt for the reason becomes more exciting since "why" gets louder and louder as the classification bins begin to fill. It also gets easier, though some causal reasons -- never any absolute ones -- may never be found.

The trick is to not overclassify and to be ready to caveat when something is suspect.

[I'm speculating on all this, admittedly.]


Come the revolution, and a lot falls down.

What's built from the rubble is, (almost always? sometimes?) better in so many ways, but it means that older scientific material is not so easy to read and understand.

So, back to astronomy.

WOBAFGKM (+RNS) - a classification scheme for stars; any revolutions past? hints of revolutions to come? No "T" with that jam? ;)
Don't forget their colors. Oops, here's another small place where your rubble metaphor applies.


How do we *know* {this star} is a K dwarf? a white dwarf? a neutron star? a binary containing a black hole? Is not the merit of each category, and each object within each category, dependent upon Data, Objectivity, Theory and Category (subjective fit).


Will it still be such a star next year? in a decade's time? a century's? (assuming it doesn't go supernova)? Some probability might be assignable.

We consciously or subconsciously look at probability with almost everything we buy. Warranties are provided to offset the concern for the probability of failure, even if only slight. The huge insurance business has their own terminology for this very thing.


How intricately does astronomical knowledge depend on classifications? Good one. I assume that it varies on the class itself, but history has examples where the object's class is critical to the path astronomy takes. The Type Ia sn class is an obvious example of amazing things wrought from the "eye" of the astronomer.


And what are classification schemes anyway? Taxonomy, speaking of species. :)

Robert Tulip
2010-Jul-17, 01:34 AM
What can we know? We know the earth is a round planet. This knowledge is not from the evidence before our eyes, because to naive perception the earth looks flat. We rely on mathematics for this simple knowledge.

If we accept that mathematics is reliable to guarantee the truth of the simple knowledge that earth is a planet, we are justified to rely on the consistency of mathematical logic to guarantee the truth of abundantly confirmed knowledge about stars and the solar system.

Knowledge of stellar physics can present appropriate caveats, such as the question of whether the cosmological constant in fact exists. Perhaps the universe is so weird that apparent constants are not to be relied on? This is over my head.

Ken G presents a consistent argument, that past error means all knowledge is provisional. This implies that at some level, we need obvious axioms to support claims that knowledge is objective truth. For example, astrophysics assumes the consistency of the cosmological constant, effectively taking this as an axiom, a matter of faith. Knowledge rests on faith in the coherence and consistency of scientific observation. Many scientists, following Hume and Popper, would like to eradicate faith entirely from epistemology, but then they are left with David Hume in the aridity of a solipsist scepticism, where we see no necessary connection between a cause and effect. Knowledge requires that we accept what Kant called synthetic a priori judgements, effectively taking on faith that the framework of space-time is real.

The examples of nineteenth century scientists who wrongly claimed that physics was nearly complete are a cautionary tale. However, the sheer consistency of modern cosmology justifies faith that the core findings of mainstream astrophysics provide correct knowledge.

This has a social implication. When scientists are too timid about the provisional status of their findings, even when those findings are central to cosmology, the general community can gain the false impression that modern science is no more reliable than non-scientific claims. With various anti-scientific proponents advocating their ideas without blushing or worrying about their accuracy, scientists have a responsibility to ensure that the doubt about core scientific knowledge is limited to the absurd semantic hypothesis, a la Descartes, that logically everything we perceive could be delusory.

Ken G
2010-Jul-17, 02:42 AM
WMany scientists, following Hume and Popper, would like to eradicate faith entirely from epistemology, but then they are left with David Hume in the aridity of a solipsist scepticism, where we see no necessary connection between a cause and effect. Knowledge requires that we accept what Kant called synthetic a priori judgements, effectively taking on faith that the framework of space-time is real.Perhaps we can identify levels of knowledge-- the knowledge we can obtain without such a priori judgements, bolstered by the knowledge we obtain with them. Other levels of knowledge might also be identified. We don't want to restrict ourselves from embracing various levels of knowledge, but we can track their different nature.


The examples of nineteenth century scientists who wrongly claimed that physics was nearly complete are a cautionary tale.Yes, I don't think the perspective of scientists in 1900 can be overstressed. They had a few indications something was wrong (like, how was the Sun in equilibrium for billions of years), but it didn't dampen their enthusiasm-- they figured they'd be minor tweaks. Today, we have dark energy and dark matter, and view them as minor tweaks to our current theories. The parallel is somewhat disturbing.
However, the sheer consistency of modern cosmology justifies faith that the core findings of mainstream astrophysics provide correct knowledge.Consistency is a fickle master.


This has a social implication. When scientists are too timid about the provisional status of their findings, even when those findings are central to cosmology, the general community can gain the false impression that modern science is no more reliable than non-scientific claims. With various anti-scientific proponents advocating their ideas without blushing or worrying about their accuracy, scientists have a responsibility to ensure that the doubt about core scientific knowledge is limited to the absurd semantic hypothesis, a la Descartes, that logically everything we perceive could be delusory.But there is a middle ground-- and scientists often get egg in their faces when they pretend too much certainty. We know LIGO will find gravitational waves, we know ITER will teach us what we need to know to get fusion, we know the LHC will find the Higgs, we know there is some extra kind of matter out there that we only see gravitationally, etc. Do we really know any of these things at all? We might be right, we might be wrong-- identifying our best current model is not the same thing as asserting truth. When the price tags are so high, science may face a bit of a crisis in the near future-- if it doesn't deliver on everything it "knows" to be true.

Nereid
2010-Jul-17, 08:00 AM
Old data, new analysis: An Explanation of Dayton Miller's Anomalous "Ether Drift" Result (http://arxiv.org/abs/physics/0608238) (link is to arXiv abstract).

Now we know why his result was what it was; at the time, no one knew.

Nereid
2010-Jul-17, 08:04 AM
A hypothetical science that consisted only of classifying previous observations would be no science at all.Right, because science isn't a Google lookup table.
But to what extent is it a set of nested giant lookup tables?

Classification at one level gives way to classification at another level, plus some interesting equations?


We need to feel that we are gaining some kind of understanding, but to understand what that even means, we need to look to the process, not just its results. If all we have any understanding of are our results, then all we have is a dogmatic clinging to the current canon of wisdom. Are there any historical analogs of that happening, and did it slow future discovery?

It seems to me epistemology is a bit like structural engineering. The structural engineer does not just look at a building and say "yup, that building is standing all right, let's go build some other ones now", which is pretty much how I interpret EDG's advice on the matter. It's true that if you want a city, you're going to have to build lots of buildings, and if some of them fall, you can always just replace them with others. But all the same, we do have structural engineers, and they are capable of looking deeply at how a structure is built, where the loads are, and what supports are dependent on what other supports, such that other kinds of questions can be answered about that building: questions like, will it sustain a hurricane, could another level be added to it, and if it needed to be taken down, where would be the most effective place to put the dynamite?

Those are the questions of epistemology-- gazing at the navel of the structures we build, rather than just going on to the next one in a kind of Darwinian "survival of the fittest" approach to architecture (or science). And despite how poorly many people understand the process of "looking under the hood" of our current science, and how frankly unnecessary it is for the daily functioning of a scientist, it would seem to be an important component of the science done by those who generate the truly startling new ways to understand our reality. Those people are always considered to be philosophers, even if they don't self-identify as such (since many people reserve the term "philosophy" for other people's philosophical views).
So, there's nothing special, or different, about astronomical knowledge?

Nereid
2010-Jul-17, 08:23 AM
An awful lot of scientific knowledge is built on the bedrock of classifications - birds are built for flying, elliptical galaxies lack gas, all Cepheids have the same P-L relationship, and so on. And the building often keeps growing, with more and more knowledge built on top of the conclusions based on the conclusions based on ... the classifications.Yes. I suppose it helps, for me at least, to see classifications as an organized effect that comes from some sort of causal relationship, even if it remains unstated. Ellipitcal galaxies are more blobby in shape, older, and about out of gas. [Good grief, no wonder my behavior gets too elliptical around here!] First came the shapes (e.g. Wolf), then the lack of gas (Curtis?), which, with bigger scopes, allowed a modern classification scheme (Hubble), then their age, which, in turn, helped in the explanation for their shape and lack of gas.

Once what is observed is orgainized into classifications, then the hunt for the reason becomes more exciting since "why" gets louder and louder as the classification bins begin to fill. It also gets easier, though some causal reasons -- never any absolute ones -- may never be found.

The trick is to not overclassify and to be ready to caveat when something is suspect.

[I'm speculating on all this, admittedly.]
There's a devil in the details!

What is a pulsar?

Originally a pulsar was something fixed on the sky which emitted EMR (electromagnetic radiation) in the radio, in short, well-timed pulses.

Later some pulsars were found to also emit EMR in other wave bands, also in short, well-timed pulses.

Today, a pulsar is a neutron star which ... same word, different meaning.

Classification schemes change, based on new knowledge.

How, and why, does the new knowledge become so accepted that it gets incorporated into classification schemes? Is astronomy (beyond the solar system) different from any other branch of science?



[...]

How do we *know* {this star} is a K dwarf? a white dwarf? a neutron star? a binary containing a black hole?
Is not the merit of each category, and each object within each category, dependent upon Data, Objectivity, Theory and Category (subjective fit).

Yes, it is ... but does constructing a classification scheme imply something qualitatively different? Something about the acceptance of some underlying self-consistent models (or theories)? Or a desire for there to be such a model (not yet developed)?

And are classification schemes in astronomy (beyond the solar system) different, in some way, from any other?


[...]

How intricately does astronomical knowledge depend on classifications?
Good one. I assume that it varies on the class itself, but history has examples where the object's class is critical to the path astronomy takes. The Type Ia sn class is an obvious example of amazing things wrought from the "eye" of the astronomer.

[...]
The history of the role of classification (in astronomy) is clear.

Classifications are essential, and come with a great deal of baggage.

But how is astronomy (beyond the solar system) different? Are classification in astronomy just like classifications in any branch of science?

Ken G
2010-Jul-17, 02:18 PM
But to what extent is it a set of nested giant lookup tables?

Classification at one level gives way to classification at another level, plus some interesting equations?Nested classification misses a key element of science, in my view-- the development of explanatory principles. A classification scheme notices similarities and differences among the key attributes, and has some degree of predictive power as a result (as in the periodic table), but by itself it cannot identify unifying principles that help us understand why those classifications work (models, symmetries, rules). We can note that mammals share several characteristics, and so when we find some of those characteristics in a new species we can expect the others too (and be right in most, but not all, cases), but the only reason we have that predictive power is that there exists some underlying reason why those attributes appear together. Science must seek that underlying reason to complete the unification of the "mammal" category, and the categories themselves just provide guidance in the seeking of those underlying reasons. It all leads to one of the trickiest questions of all in science: what "counts" as an explanation, and what does it mean when some scheme is unifying, quantitative, predictive, and self consistent?



So, there's nothing special, or different, about astronomical knowledge?I don't think there's anything fundamentally special about it. It is highly divorced from our daily experience, but so is the study of atoms, or the study of low-temperature superfluids, or fusion reactors. Probably the most important difference is that we cannot adjust the experimental parameters, we can only observe the experiments already set up, so that's an important difference but it is largely mitigated by the sheer magnitude of all the different "experiments" already set up and running out there-- almost every law of physics one can name is playing out somewhere out there. We have to idealize the systems we are observing to get at those laws, because we don't get to "control" the experiments to focus them, and as a result there are quite a few results that come up "?" for a long time, but eventually they get brought into the consensus. It doesn't seem fundamentally different to me, and the education that an astronomer receives closely parallels that of a physicist.

Disinfo Agent
2010-Jul-17, 02:51 PM
What can we know? We know the earth is a round planet. This knowledge is not from the evidence before our eyes, because to naive perception the earth looks flat. We rely on mathematics for this simple knowledge.

If we accept that mathematics is reliable to guarantee the truth of the simple knowledge that earth is a planet, we are justified to rely on the consistency of mathematical logic to guarantee the truth of abundantly confirmed knowledge about stars and the solar system.Math plays a role in all of that, but the crucial reason why we believe those things is empirical.


Ken G presents a consistent argument, that past error means all knowledge is provisional. This implies that at some level, we need obvious axioms to support claims that knowledge is objective truth. For example, astrophysics assumes the consistency of the cosmological constant, effectively taking this as an axiom, a matter of faith. Knowledge rests on faith in the coherence and consistency of scientific observation. Many scientists, following Hume and Popper, would like to eradicate faith entirely from epistemology, but then they are left with David Hume in the aridity of a solipsist scepticism, where we see no necessary connection between a cause and effect. Knowledge requires that we accept what Kant called synthetic a priori judgements, effectively taking on faith that the framework of space-time is real.

The examples of nineteenth century scientists who wrongly claimed that physics was nearly complete are a cautionary tale. However, the sheer consistency of modern cosmology justifies faith that the core findings of mainstream astrophysics provide correct knowledge.

This has a social implication. When scientists are too timid about the provisional status of their findings, even when those findings are central to cosmology, the general community can gain the false impression that modern science is no more reliable than non-scientific claims. With various anti-scientific proponents advocating their ideas without blushing or worrying about their accuracy, scientists have a responsibility to ensure that the doubt about core scientific knowledge is limited to the absurd semantic hypothesis, a la Descartes, that logically everything we perceive could be delusory.I think the greatest danger comes not from acknowledging the provisional status of knowledge, but from neglecting it. Though I admit that sometimes it's right to be assertive (e.g. global warming, evolution, vaccines, etc.)

Kant noted that synthetic a priori judgements existed in math, and then proceeded to argue that they were also possible in philosophy. I accept the former, but I think he was mistaken about the latter. It's in mathematics - and only within a purely mathematical framework - that theories can be absolutely proven.


So, there's nothing special, or different, about astronomical knowledge?I'd like to echo Ken, above. Whatever there is of special in astronomical knowledge lies in its content and subject matter, not in its methods. The scientific method is the same for everyone.

George
2010-Jul-17, 08:34 PM
What can we know? We know the earth is a round planet. This knowledge is not from the evidence before our eyes, because to naive perception the earth looks flat. We rely on mathematics for this simple knowledge.

If we accept that mathematics is reliable to guarantee the truth of the simple knowledge that earth is a planet, we are justified to rely on the consistency of mathematical logic to guarantee the truth of abundantly confirmed knowledge about stars and the solar system.
Math plays a role in all of that, but the crucial reason why we believe those things is empirical.

Indeed, math (geometry) was the tool used to take what was observed – a spherical Earth due to movement of the celestial axis proportional to observations as one travels north or south; differences in the length of the Sun’s shadow for different latitudes – and quantify it rather accurately, especially for the BC boys.

So even the claim of the Earth’s shape was, at some point, well-founded on empirical evidence.


Old data, new analysis: An Explanation of Dayton Miller's Anomalous "Ether Drift" Result (link is to arXiv abstract).

Now we know why his result was what it was; at the time, no one knew. Thanks for the clarification. [Reading only the abstract, I would have assumed signal drift was common enough to have been taken more seriously back then. Ham and commercial radios required constant tuning until solid state came along.]


Consistency is a fickle master. Nice one! I hadn’t thought of it as a master, but I see your point, at least for science. [Is this a Ken original, as far as you can recollect?]


But there is a middle ground-- and scientists often get egg in their faces when they pretend too much certainty. We know LIGO will find gravitational waves, we know ITER will teach us what we need to know to get fusion, we know the LHC will find the Higgs, we know there is some extra kind of matter out there that we only see gravitationally, etc. “The Party of Know!” Are they politically affiliated? ;)


What is a pulsar?

Originally a pulsar was something fixed on the sky which emitted EMR (electromagnetic radiation) in the radio, in short, well-timed pulses.

Later some pulsars were found to also emit EMR in other wave bands, also in short, well-timed pulses.

Today, a pulsar is a neutron star which ... same word, different meaning.
[I recall their original classification - LGM (Little Green Men) :)]

The data didn’t change but did become more comprehensive and much more accurate.

Playing with the suppositional formula – the first two probability factors are likely high, but the latter two, theory and classification, are poor. The QF is even worse for the case of the LGM claim (Little Blue Men, perhaps, given a pulsar’s temperature ;) ).


But how is astronomy (beyond the solar system) different? Are classification in astronomy just like classifications in any branch of science? I too agree with Ken’s fundamental view of this. Why would it be different since the methodology itself seems rather simple – arranged bins on shelves that get rearranged when deemed appropriate?

Of course, astronomy has a unique “flavor” to it since it is all about light but in such a big way; it’s heavy-duty light taxonomy. (sorry, I’ll be less moronic till next Friday) It is also about the really, really big stuff, which doesn’t even have words since "trillion" -- our biggest common numeric word, though a million times more than Bible’s largest math term, “thousand” -- is usually a trivial astronomical term to describe what is out beyond our Horizon (hmmm, is this a decent pun?).

Nereid
2010-Jul-18, 11:53 AM
Nereid,

I've "heard" you consider this line of questioning before and I think the general idea of it is of interest to many.


Do we know, for example, that other stars are composed, largely, of hydrogen and helium (most of them anyway)?
The existence of dark matter still must be considered hypothetical or theoretical since there is no consensus that dark matter has ever been observed or that it really exists other than baryonic matter of some kind. The same thing applies to dark energy.


Do we know that there's a super-massive black hole at the centre of the Milky Way?

As you know black holes have a number of different definitions and related theories concerning their true nature. However, there is a vast amount of evidence that supports the assertion that there is a strong gravitational influence at the center of most galaxies that seems to be concentrated in a relatively small volume which is seemingly strong support for a "black hole" type theory of some sort.

Do we know that Geminga is a ~sol-mass object, composed almost entirely of nuclear degenerate matter?
Neutron stars are solely a theoretical entity but it would be otherwise more difficult to explain the gravitational concentration of matter and its observed gamut of radiation and magnetic influences in another way -- so it seems like a good theory to me.


Do we know that dark matter is cold, and non-baryonic?
again such ideas I think are speculative considering that dark matter as an unknown particle entity which may not exist at all. Such hypotheses are necessary and will be well considered provided they are not discussed as knowledge, which might be challenged by many.


How should we go about deciding what we know (and what we don't)?
The description of something as knowledge should follow the common definition of the word knowledge, which I think is good enough even though such "knowledge" in the future may turn out someday to be wrong. Such a definition of knowledge might be: "What is learned by education, perceptual experience, and reasoning." Knowledge of theory is still knowledge of sorts. The disagreement is generally only a matter of opinion as to whether a hypothesis or theory has graduated to the fact status or not and the degree of perceived certainty involved?


Does it even make sense to say we 'know' something, anything, about things beyond our solar system?
A cautious approach might be to refer to the theory involved but most educated people in the field, as you know, will realize the theory that you are assuming when you are discussing cosmology for instance, unless you think what you are saying could be misunderstood or challenged without added qualification.(bold added)

Lots of terms in this post by forrest noble!

Let's start with 'theoretical entity'.

You go out at night, and look up at the sky (which is, for our purposes, cloudless). You see a bright star, and you know (how?) that it is called Sirius. Is Sirius a 'theoretical entity'?

You observe Sirius through your telescope, and see that it is a double star; there's a faint 'companion' nearby. You know that the bright star is called Sirius A, and the faint companion Sirius B. Is Sirius B a 'theoretical entity'?

You read some astrophysics textbooks, and from them you come to know (how?) that Sirius A is a main sequence star, which shines, indirectly, by nuclear fusion in its core, that it is supported against gravitational collapse by gas pressure, and so on. You come to know that Sirius B is a dwarf star, which shines because it is hot, and that it is supported against gravitational collapse by degenerate electron pressure, and so on.

Are these descriptions of Sirius A and Sirius B 'theoretical entities'?

Must the 'existence of' 'stars which shine by nuclear fusion' 'be considered hypothetical'? Do such stars 'really exist'?

Must 'the existence of' 'stars supported by electron degeneracy pressure' 'be considered hypothetical'? Do such stars 'really exist'?

(to be continued)

Robert Tulip
2010-Jul-18, 12:58 PM
We know LIGO will find gravitational waves, we know ITER will teach us what we need to know to get fusion, we know the LHC will find the Higgs, we know there is some extra kind of matter out there that we only see gravitationally, etc. Do we really know any of these things at all? We might be right, we might be wrong-- identifying our best current model is not the same thing as asserting truth. When the price tags are so high, science may face a bit of a crisis in the near future-- if it doesn't deliver on everything it "knows" to be true.

You make the mainstream hostage to the frontier.
http://en.wikipedia.org/wiki/LIGO
http://en.wikipedia.org/wiki/ITER
http://en.wikipedia.org/wiki/Large_Hadron_Collider

Looking at these examples you have raised, they assume the certainty of core knowledge within mainstream astrophysics, and seek to extend it at the unknown frontier. Science does not know where the frontier will go. We do know that stellar knowledge well within the mainstream, of the type queried in this thread, is true.

Disinfo Agent
2010-Jul-18, 04:19 PM
There are degrees in knowing. I think this is Ken's main point. It's definitely my main point in these discussions. As scientists we should try not to mislead others into a false sense of certainty. It's hard, because our profession makes us passionate about science, and especially prone to falling into dogmatism. Nevertheless, Robert, you are quite right in insisting that there are lots of things in science that we can state with a great deal of certainty. It's OK to say that we 'know' these things. It's a useful shorthand, and it's consistent with everyone's everyday use of the word.

I don't like it when experts use complicated jargon that obscures issues instead of illuminating them (although I acknowledge that it's hard to escape this vice - a professional hazard...) I believe that clarity is a virtue, not a sign of lack of sophistication. Sometimes we need to simplify in order to be clear, and that can end up misleading, but it's a risk worth taking. It's better to mislead and then be forced to correct a misconception, than to stay in our comfort zone and refuse to communicate with the 'lay'.

But scientists should demand more understanding from themselves than they do from non-scientists. It's OK to tell the public 'We know this - it's well-established science'. But we should also acquaint ourselves with the limitations of that knowledge, and of the science that produced it. It's a balancing act.

When a conversation turns this abstract, one sometimes gets the feeling that one is saying trivialities. Like everything else, philosophy is more useful when it focuses on concrete problems, and often becomes sterile when it wanders off into the abstract. I wonder if we could come up with a few concrete problems to discuss...

George
2010-Jul-18, 05:53 PM
There are degrees in knowing. Yes, and this is another example where we have language inadequacies in certain areas. To "know" something is a statement of finality even if we understand it comes in degrees. It can often be a "cut to the chase" term in order to skip over all the facts that establish what is claimed as "known".

A more accurate use of "know", however, at least for science, should be applied only to the objective facts themselves. The theories that are based on what we know are not final things, but works in progress. Many understand this and use this against science (e.g. "A theory is just a theory"). When scientists argue that the theory is what we know is true, inevitably the scientist, if contested long enough, will be forced to reveal what are the real "known" things (i.e. objective evidence) that justifies his or her support for the theory itself.

Thus, there are more tactful ways to handle discussions to the general public when the topic is more the theory or broad issue than the supportive facts themselves. BBT, for example, can only be made credible by introducing the important facts and "lines of evidence" (another weak phrase, btw) that will allow the listener to see the "why" behind scientists confidence for the theory.

To put it bluntly, it's selling. We are all involved in selling either as "sellers" or "buyers", but we often have little interest in perfecting the art of either role. I'm no different and my use of puny puns, oxymorons, and poor grammar should be solid proof that I prefer having fun, entertaining others, learning new things, admiring the adventures of others, etc. over having to overcome a listener or group of listeners puerile incognizance of a very respectable scientific principle or theory. [Example: Just last night, my son and I had a freindly argument just last night after watching the new move Inception. When gravity was removed in one scene, I pointed out an inconsistency in the movie since the viewer was to assume this was not falling. He, at first, disagreed. ]



I don't like it when experts use complicated jargon that obscures issues instead of illuminating them (although I acknowledge that it's hard to escape this vice - a professional hazard...) I believe that clarity is a virtue, not a sign of lack of sophistication. Sometimes we need to simplify in order to be clear, and that can end up misleading, but it's a risk worth taking. I think it is so much harder for scientists today to be lucid enough to convey the much more complicated and deeper physics involved. QM and GR are not theories that can be presented in any sort of detail in a short public presentation or interview. The listener will be lost the first time they hear "quantum" or "tensor", which is why I think quantum mechanics should have been called Fred or Sue. :)


It's better to mislead and then be forced to correct a misconception, than to stay in our comfort zone and refuse to communicate with the 'lay'. This is unfortunate if true, and you may well be correct, at least in the cases where the misleading is trivial. Any corrections that must be made will be proof that they don't "know" what they said they knew. Once trust is lost, it is hard to regain it.


But scientists should demand more understanding from themselves than they do from non-scientists. It's OK to tell the public 'We know this - it's well-established science'. But we should also acquaint ourselves with the limitations of that knowledge, and of the science that produced it. It's a balancing act. Yes and I would assume such gymnastics are not part of the formal curriculum, but, of course, we can help them here. :)

Nereid
2010-Jul-18, 09:06 PM
Back to astronomy, beyond the solar system.

One interesting thing: no matter how much the theories of physics change, or the nature of science, astronomical observations are pretty robust.

For example, Sirius was observed to have a brightness of {some number}, when I observed it {date, time, place}; or the separation of Sirius A and B was observed to be {", PA}, when I observed it {date, time, place, equipment}.

Later, possibly even centuries later, someone may take those observations and make some (usually minor) adjustments to them, just as the SDSS team tweaked their published data, as the WMAP team refined their analyses, and as the HIPPARCOS data was re-reduced.

And as long as the 'meaning chains' associated with the observations remains clear - or easily re-constructed - we can know that all astronomical observations are robust. 'Meaning chains'? Think of 'main sequence star' - it has a pure observational meaning, and a theory-based one ... determining which is intended, wrt a particular set of observations, is key, to compiling a great database of timeless observations.

Astronomical observations which are less than robust? Written descriptions, in everyday language, of what something looked like, through a telescope say. Think of historical catalogues such as the Messier list, for example. Note that robust does not, necessarily, mean useful; think of the multiple sets of observations of GRBs, in the early days, produced by a dozen different observatories, using different telescopes, filters, detectors, reference stars, through different airmasses, etc.

Ken G
2010-Jul-19, 02:25 AM
There are degrees in knowing. I think this is Ken's main point. It's definitely my main point in these discussions.Right, it isn't black-and-white, I "know it or I don't", like the capital of Kenya. What's more, and this is the really dicey part, it is extremely hard to know how well you know something. In the history of science, some of the most tightly held "truths" of all turned out to be nothing but a mirage of sorts, useful as far as it went, but hardly a fundamental truth of how things really work. Nevertheless, we are not likely ever going to decide that most stars are not powered by fusion of hydrogen their cores, that seems "rock solid" as a starting point. Rather, it is what else they are doing that may surprise us, or what "hydrogen fusion" is really about might surprise us.


Nevertheless, Robert, you are quite right in insisting that there are lots of things in science that we can state with a great deal of certainty. It's OK to say that we 'know' these things. It's a useful shorthand, and it's consistent with everyone's everyday use of the word.Yes, I agree, all we are talking about in this thread is what we mean when we use that word, and how we should start by understanding our own meaning before explaining it to nonscientists. And above all, remember Feynman's definition of science: belief in the fallibility of experts.
I wonder if we could come up with a few concrete problems to discuss...We could talk about how certain we are that general relativity describes the motion of the universe as a whole? Whether or not quantum mechanics provides a fundamental description of what is happening on the micro scale?

Ken G
2010-Jul-19, 07:08 AM
And as long as the 'meaning chains' associated with the observations remains clear - or easily re-constructed - we can know that all astronomical observations are robust. 'Meaning chains'? Think of 'main sequence star' - it has a pure observational meaning, and a theory-based one ... determining which is intended, wrt a particular set of observations, is key, to compiling a great database of timeless observations.Observations are generally robust-- that is true in laboratory physics as well. It connects with the degrees of knowing-- we can know that an observation came out a certain way (unless dishonesty is involved), and we can infer that another similar one will come out similarly (unless error was involved). So already we encounter several "degree of separation" from pure knowing. Then we tack on the next degrees of separation, which have to do with the "meaning chains" applied to the observation. If the meaning chain only requires that we have a self-consistent model of the internal physics of a star, say, then it's probably pretty solid, but if the meaning chain requires that general relativity applies to the universe as a whole (with, say, a cosmological constant and a bunch of dark matter), then we have an inferential milieu that could change overnight if a better approach to the unknowns is developed.

These meaning chains, and their interdependences, are so complex that we almost never even recognize what their impact is until something fundamental gets altered in some particular way. In a way science does evolve a bit like speciation-- we know what lifeforms can be viable in a given niche, but we have no idea what other lifeforms could also have lived in that niche, nor what a change in the niche will do to the most viable lifeforms until we see the outcome of the evolutionary process that simply tries out and discards. Since in science, the old observations generally continue to be valid, we have the unusual aspect that in science, the new "viable" theory has to be viable not only in the new observational niche, but in the old one as well. That means it has to have been a theory that could have been accepted given the previous observations, it simply lacked a good enough reason, or enough imagination on our part. In the previous environment, the new theory was outcompeted by the old one, but in the new environment, it is selected by its new survival advantages (which oftentimes have to do with our simple ability to even conceive of the new theory-- necessity is the mother of invention).

George
2010-Jul-19, 12:36 PM
These meaning chains, and their interdependences, are so complex that we almost never even recognize what their impact is until something fundamental gets altered in some particular way. In a way science does evolve a bit like speciation-- we know what lifeforms can be viable in a given niche, but we have no idea what other lifeforms could also have lived in that niche, nor what a change in the niche will do to the most viable lifeforms until we see the outcome of the evolutionary process that simply tries out and discards. Since in science, the old observations generally continue to be valid, we have the unusual aspect that in science, the new "viable" theory has to be viable not only in the new observational niche, but in the old one as well. That means it has to have been a theory that could have been accepted given the previous observations, it simply lacked a good enough reason, or enough imagination on our part. In the previous environment, the new theory was outcompeted by the old one, but in the new environment, it is selected by its new survival advantages (which oftentimes have to do with our simple ability to even conceive of the new theory-- necessity is the mother of invention). Yes, "survival of the fittest" where the critter here is a mathematical construct that must adapt to the changing environment of Nature as it shrinks itself to get into the smaller and smaller crevaces, or ascends higher and higher into the vast expanse.

Ken G
2010-Jul-19, 02:28 PM
Nice metaphor-- shall we take the second one?

Nereid
2010-Jul-19, 02:36 PM
Suppose we have lots and lots and lots of good observations of the position, on the sky, of a star, over a long time period (say several years), with relatively evenly spread data (say, once every two days).

From this data, and a good, sound, analysis of it, we conclude that the star is 2.12 +/- .11 pc from us.

Where, on the continuum of knowing, do we place this: "we know {this star} is 2.1 +/- .1 pc from us, in {time period}"?

We also have lots of very good data in the form of spectra, covering the range from the Lyman limit to the FIR.

From this data, and a good, sound, analysis of it, we conclude that the star has an effective temperature of 10,050 +/- 130 K, and a surface gravity of 14.3 +/- 0.6 g.

Where, on the continuum of knowing, do we place this: "we know {this star} is has an effective temperature of 10,050 +/- 130 K, and a surface gravity of 14.3 +/- 0.6 g, in {time period}"?

We also have some very good data in the form of long-baseline optical interferometric observations.

From this data, and a good, sound, analysis of it, we conclude that the star has a radius, in the V band, of 2.45 +/- 0.32 sols (a 'sol' in this case is the radius of the Sun).

Where, on the continuum of knowing, do we place this: "we know {this star} is has a radius, in the V band, of 2.5 +/- 0.3 sols, in {time period}"?

forrest noble
2010-Jul-19, 07:11 PM
.............Where, on the continuum of knowing, do we place this: "we know {this star} is has a radius, in the V band, of 2.5 +/- 0.3 sols, in {time period}"?

I think stellar theory is pretty solid. Maybe a few of the details you mentioned above might change over time but I believe the general concepts are pretty well founded and "known."

Examples of "Knowledge" that few of us might not be surprised if there were charges to in its most accepted versions might be "black hole theory," dark matter theory, dark energy theory, supernovae theory.

On the other hand Big theories such as the BB, would come as a great surprise to many if it were completely replaced. Quantum Theory has its basis in Quantum Mechanics which is based upon countless observations. If the theory changed much of the foundation would still remain, just a different perspective might be realized which might change the way the theory is applied for instance. Particle theory has many parts to it so if a part of it were to change that would not be surprising to many. If the foundation of particle theory such as quark theory were replaced by a string configuration of some kind for instance, some would be surprised and others less surprised.

Bottom line is it's all a matter of degree. When you are in the company of people who think alike or have a similar education you could use the words "we know this or that" and it would seldom be challenged. If you were talking at a conference where you thought scientists of differing opinions could be there you might qualify your statements to some extent or another. One might say that for every idea of knowledge out there, there is a tolerance of probability implicitly attached to it which in many cases would change over time, which of course is just a matter of opinion as to what that tolerance might be at any given point in time.

As far as the major theories out there, I don't think it will take a lifetime to see big theoretical changes to what some presently might call "knowledge."

Ken G
2010-Jul-19, 07:17 PM
Suppose we have lots and lots and lots of good observations of the position, on the sky, of a star, over a long time period (say several years), with relatively evenly spread data (say, once every two days).

From this data, and a good, sound, analysis of it, we conclude that the star is 2.12 +/- .11 pc from us.I'd put that pretty high up on the continuum, because it relies only on trigonometry in Euclidean space, and the motion of the Earth around the Sun, both of which have a host of connected contingencies that rely on them being accurate. Still, we can't say we really know that distance to be "true", because we don't know what other theories of geometry or motion might allow the same parallax observations to lead to a different distance determination. For example, imagine we do some new experiment that requires an absolute aether frame whose effects were so far covered up by the Lorentz transformation to other frames where most of the laws of physics maintained the same form (but not the hypothetical new discovery). If that were to happen, we might be tempted to evaluate all distances in that new absolute aether frame, and that could alter every distance we use-- without altering any parallax observations (which would then be viewed as distorted by our motion in that absolute aether frame).

We also have lots of very good data in the form of spectra, covering the range from the Lyman limit to the FIR.

From this data, and a good, sound, analysis of it, we conclude that the star has an effective temperature of 10,050 +/- 130 K, and a surface gravity of 14.3 +/- 0.6 g.That would be perhaps not as high up in the continuum, because it could be altered without fundamentally shaking any of our current views of the nature of space and time, it might just require a recalibration of some currently trusted approach. Perhaps a new form of turbulence is found that puts different energy into different species, and requires new spectral fits. The models of stellar atmospheres are easily shifted by some change in the fundamental assumptions-- maybe the ionization is found to be oscillating rather than in a steady state, or we might need new spatial and temporal inhomogeneities, who knows. We don't assume any of these modifications are needed until we have evidence they are needed, but such evidence can appear at any time, and the approaches to determining stellar atmospheric conditions can change just as quickly.

We also have some very good data in the form of long-baseline optical interferometric observations.

From this data, and a good, sound, analysis of it, we conclude that the star has a radius, in the V band, of 2.45 +/- 0.32 sols (a 'sol' in this case is the radius of the Sun).The angular size would be high up on the continuum, because you are coming pretty close to actually seeing it with the interferometry, but of course the actual radius also requires distance, so is no better than the distance determination. Again, distortions from the spacetime itself might alter what we see from what we perceive as the actual reality-- distortions we have no current reason to expect or model, but all that could change with some new window that is opened up.

Nereid
2010-Jul-19, 07:26 PM
How many dimensions does 'knowing', in beyond-the-solar-system astronomy, have?

No one seems to disagree that 'knowing' comes in at least one dimension, a cline, from super-rock-solid knowing (the observations, we all seem to agree, modulo some minor caveats) to wild speculation (essentially no logic, no math, appeals to intuition, not even any reference to any observations).

If astronomical knowledge does have but one dimension, to what extent can we place examples of such knowledge on the cline? Can we always make pair-wise comparisons, for example (a star's distance is 'stronger knowledge' than it's surface gravity, say, in my previous example)? And what shorthand should we adopt for 'strength of knowledge' (i.e. how close to observations it comes)?

Or does astronomical knowledge have more than one dimension?

What is the taxonomy of astronomical knowledge? How does it differ from the taxonomy of epistemology in general?

Ken G
2010-Jul-19, 10:15 PM
Or does astronomical knowledge have more than one dimension?I would say that all knowledge, not just astronomical, has more than one dimension. Indeed, it has so many dimensions that it is a much harder problem then the knowledge itself, which is why science has never attempted to tackle it! It's more akin to the problems one encounters in artificial intelligence than it is like the problems of coming up with physical laws.


What is the taxonomy of astronomical knowledge? How does it differ from the taxonomy of epistemology in general?

I'd say it's too interconnected with scientific epistemology in general to be fundamentally different. It is less personal, so perhaps we can set it off from the more personal forms of epistemology (like ethics or values), but that's true of all of science.

Jerry
2010-Jul-20, 12:48 AM
I love this topic!

Science is all about extrapolations; into space-and-time into both the past and into the future. Here is a helpful analogy:

In the 19th century there were very good extrapotations of the minimal age of the earth that were based upon simple evaluations of the strata, fossils and measurable erosion rates. There was an equally good extrapolation of the mass and composition of the sun. But the the known laws of energy conservation created a paradox: Nothing could burn for as long as the sun had surely endured with known chemical reactions and gravitational heating. Something big was missing. The paradox made it impossible to explain the age of the earth without a magic energy factor.

The discovery of natural radiation processes both explains the necessary age of the sun and provides a way to test a minimal age for the earth, given us confidence in our evaluation of the physical processes governing the earth.

But how much confidence should we have in this most-simple solition? I wouid feel a lot better if the lithium problem did not exist, and if the most distance galaxies observe were not so old. Since our best extrapolations are at odds with our best observations, there should be just as much concern today about 'big missing pieces' as there was in the 19th century.

Dark energy and Dark matter are 'big missing pieces' as well - observations that are at odds with the simplest extrapolations of our best-guess theories. What we know best is that we don't know.

George
2010-Jul-20, 02:21 AM
Science is all about extrapolations; into space-and-time into both the past and into the future. Here is a helpful analogy:

In the 19th century there were very good extrapotations of the minimal age of the earth that were based upon simple evaluations of the strata, fossils and measurable erosion rates. There was an equally good extrapolation of the mass and composition of the sun. But the the known laws of energy conservation created a paradox: Nothing could burn for as long as the sun had surely endured with known chemical reactions and gravitational heating. Something big was missing. The paradox made it impossible to explain the age of the earth without a magic energy factor.

The discovery of natural radiation processes both explains the necessary age of the sun and provides a way to test a minimal age for the earth, given us confidence in our evaluation of the physical processes governing the earth. As Tommy Smothers would say to Dickie...."take it" Ken. :) [It was Ken that introduced me to yet another surprise in the science behind the Earth's age that ironically reinforces your correct view.]

Ken G
2010-Jul-20, 07:58 AM
George is referring to the fact that Eddington did a calculation of the heat transport from the Earth's interior, and got the answer that given the heat transfer we measure from the interior now, the Earth could not be billions of years old. Eddington's mistake was he used a very inefficient heat transfer process (it really wasn't the neglect of radioactive heating, that's a common but erroneous answer), and with such an inefficient process, the Earth's core would need to still be extremely hot to support the heat flux we now measure. In actuality, convection in the molten mantle greatly assists the heat transport, so the Earth can support the current flux we measure with a much cooler (and older) core. The fact that the Sun had a similar problem (it could not be billions of years old, given its current heat transport, for the opposite reason-- it's core would have cooled off too much by now without a heat source) was a coincidence that was quite unfortunate for Eddington. But you can see where he was coming from-- he did one calculation that said the Sun was millions of years old (based on an egregious underestimate of how much energy it had available to supply the current heat flux) and another calculation that said the Earth was millions of years old (based on an egregious overestimate of the core temperature the Earth would need to supply the current heat flux). Getting answers that seemed consistent was encouraging to Eddington, and steeled him to go against the entire geologic community of the day. But as I said, consistency is a fickle master.

Robert Tulip
2010-Jul-20, 09:30 AM
Getting answers that seemed consistent was encouraging to Eddington, and steeled him to go against the entire geologic community of the day. But as I said, consistency is a fickle master.

The inconsistency in Eddington's hypothesis was fairly rapidly exposed. The advance of science since then means the frontiers of consistency have massively expanded. We see this in the ability of BAUT to present a standing offer via the ATM forum for anyone to argue some inconsistency in mainstream knowledge. As far as I have heard, no ATM proponent has ever demonstrated an inconsistency that was not already well known to astrophysics. The frontiers of knowledge, looking outside the mainstream rather than attempting to go against the mainstream, is where the question of what is known and what is unknown becomes relevant.

George
2010-Jul-20, 01:14 PM
George is referring to the fact that Eddington did a calculation of the heat transport from the Earth's interior, and got the answer that given the heat transfer we measure from the interior now, the Earth could not be billions of years old. Eddington's mistake was he used a very inefficient heat transfer process (it really wasn't the neglect of radioactive heating, that's a common but erroneous answer), and with such an inefficient process, the Earth's core would need to still be extremely hot to support the heat flux we now measure. In actuality, convection in the molten mantle greatly assists the heat transport, so the Earth can support the current flux we measure with a much cooler (and older) core. The fact that the Sun had a similar problem (it could not be billions of years old, given its current heat transport, for the opposite reason-- it's core would have cooled off too much by now without a heat source) was a coincidence that was quite unfortunate for Eddington. But you can see where he was coming from-- he did one calculation that said the Sun was millions of years old (based on an egregious underestimate of how much energy it had available to supply the current heat flux) and another calculation that said the Earth was millions of years old (based on an egregious overestimate of the core temperature the Earth would need to supply the current heat flux). Getting answers that seemed consistent was encouraging to Eddington, and steeled him to go against the entire geologic community of the day. Nice and lucid. I think you meant Lord Kelvin, however. It was Eddington that introduced radioactivity as an explanation for an older Earth, correcting Kelvin's conductive transfer age calculation. [Added: My mistake, Ken is correct, of course. It is Eddington here.]


But as I said, consistency is a fickle master. :) Yes, nice example.

Nereid
2010-Jul-20, 01:47 PM
Or does astronomical knowledge have more than one dimension?I would say that all knowledge, not just astronomical, has more than one dimension. Indeed, it has so many dimensions that it is a much harder problem then the knowledge itself, which is why science has never attempted to tackle it! It's more akin to the problems one encounters in artificial intelligence than it is like the problems of coming up with physical laws.
Would you care to say more?

Limiting ourselves to only astronomical knowledge, per the OP, what might some of the dimensions of such knowledge be?



What is the taxonomy of astronomical knowledge? How does it differ from the taxonomy of epistemology in general?
I'd say it's too interconnected with scientific epistemology in general to be fundamentally different. It is less personal, so perhaps we can set it off from the more personal forms of epistemology (like ethics or values), but that's true of all of science.
So if astronomical knowledge (per the OP) is, to a first order, no different than scientific epistemology, how should we comment on forest noble's and Jerry's posts?

Can we say, for example, that there is no difference - in terms of knowing - between CDM (cold dark matter) in astrophysics and, say, some as yet undiscovered, long extinct, Chordate species? Or between a ~4 millions sol-mass SMBH at SgrA* and, say, an as yet undiscovered isotope of element 118?

Ken G
2010-Jul-20, 02:01 PM
The inconsistency in Eddington's hypothesis was fairly rapidly exposed.True enough-- but such is not always the case.

The advance of science since then means the frontiers of consistency have massively expanded. We see this in the ability of BAUT to present a standing offer via the ATM forum for anyone to argue some inconsistency in mainstream knowledge. As far as I have heard, no ATM proponent has ever demonstrated an inconsistency that was not already well known to astrophysics. The frontiers of knowledge, looking outside the mainstream rather than attempting to go against the mainstream, is where the question of what is known and what is unknown becomes relevant.Also true-- and yet a balance must be struck. Although a successful ATM challenge is rare, and usually comes from well-trained scientists, nevertheless the next great discovery will be ATM-- for a spell. Einstein was the originator of special relativity, not because he was the only one who had access to the observations or mathematical skill, but rather because he was the first who was willing to get suitably ATM (and that was because of his largely ATM philosophy about the pre-eminence of the observer over space and time).

Nereid
2010-Jul-24, 03:31 PM
Returning to this ...

Suppose we have lots and lots and lots of good observations of the position, on the sky, of a star, over a long time period (say several years), with relatively evenly spread data (say, once every two days).

From this data, and a good, sound, analysis of it, we conclude that the star is 2.12 +/- .11 pc from us.I'd put that pretty high up on the continuum, because it relies only on trigonometry in Euclidean space, and the motion of the Earth around the Sun, both of which have a host of connected contingencies that rely on them being accurate. Still, we can't say we really know that distance to be "true", because we don't know what other theories of geometry or motion might allow the same parallax observations to lead to a different distance determination. For example, imagine we do some new experiment that requires an absolute aether frame whose effects were so far covered up by the Lorentz transformation to other frames where most of the laws of physics maintained the same form (but not the hypothetical new discovery). If that were to happen, we might be tempted to evaluate all distances in that new absolute aether frame, and that could alter every distance we use-- without altering any parallax observations (which would then be viewed as distorted by our motion in that absolute aether frame).

We also have lots of very good data in the form of spectra, covering the range from the Lyman limit to the FIR.

From this data, and a good, sound, analysis of it, we conclude that the star has an effective temperature of 10,050 +/- 130 K, and a surface gravity of 14.3 +/- 0.6 g.That would be perhaps not as high up in the continuum, because it could be altered without fundamentally shaking any of our current views of the nature of space and time, it might just require a recalibration of some currently trusted approach. Perhaps a new form of turbulence is found that puts different energy into different species, and requires new spectral fits. The models of stellar atmospheres are easily shifted by some change in the fundamental assumptions-- maybe the ionization is found to be oscillating rather than in a steady state, or we might need new spatial and temporal inhomogeneities, who knows. We don't assume any of these modifications are needed until we have evidence they are needed, but such evidence can appear at any time, and the approaches to determining stellar atmospheric conditions can change just as quickly.

We also have some very good data in the form of long-baseline optical interferometric observations.

From this data, and a good, sound, analysis of it, we conclude that the star has a radius, in the V band, of 2.45 +/- 0.32 sols (a 'sol' in this case is the radius of the Sun).The angular size would be high up on the continuum, because you are coming pretty close to actually seeing it with the interferometry, but of course the actual radius also requires distance, so is no better than the distance determination. Again, distortions from the spacetime itself might alter what we see from what we perceive as the actual reality-- distortions we have no current reason to expect or model, but all that could change with some new window that is opened up.
All of Ken G's comments have been couched in terms of some set of, or kind of, theories of physics (and underlying maths). And they have included examples of how the conclusions drawn, from observations, might be skewif.

But doesn't this just raise an even more fundamental question (or set of same)?

For example, what are the unknown unknowns concerning our ideas of distance (wrt stars we cannot visit, in person, in our lifetimes)? And are these unknowns any more, or less, "high up on the continuum" than unknown unknowns in chains leading to effective temperature or surface gravity?

George
2010-Jul-24, 08:22 PM
For example, what are the unknown unknowns concerning our ideas of distance (wrt stars we cannot visit, in person, in our lifetimes)? And are these unknowns any more, or less, "high up on the continuum" than unknown unknowns in chains leading to effective temperature or surface gravity? Nice question. [Are these second order unknowns?] This seems theory specific. Is the physics associated with stellar temperature better, or more robust, than another?

Robert Tulip
2010-Jul-26, 12:00 AM
Returning to this ...
All of Ken G's comments have been couched in terms of some set of, or kind of, theories of physics (and underlying maths). And they have included examples of how the conclusions drawn, from observations, might be skewif.

But doesn't this just raise an even more fundamental question (or set of same)?

For example, what are the unknown unknowns concerning our ideas of distance (wrt stars we cannot visit, in person, in our lifetimes)? And are these unknowns any more, or less, "high up on the continuum" than unknown unknowns in chains leading to effective temperature or surface gravity?

Here we see the 'Rumsfeld epistemology', ranging from the 'known knowns', through the 'known unknowns' to the 'unknown unknowns'. There are no 'unknown knowns' because this category is logically inconsistent.

Astronomy is well placed to demarcate between these categories of knowledge.

My impression is that the theory of distance that supports general relativity is accurate, a 'known known'. This theory of distance rests on an assumption that the laws of physics are consistent throughout the universe, and that observation is consistent with a cosmological constant.

Theories of stellar physics rest on knowledge of a cosmological constant. I do not have the grasp of physics to say what likelihood there may be that an 'unknown unknown' would invalidate the foundation of knowledge provided by modern observation of the universe. It seems this frontier of knowledge would be seen through analysis of inconsistencies in the standard models. Postulation of explanations such as mass that does not engage with light (dark matter), and also possibly string theory, pushes towards both the 'known unknown' and the 'unknown unknown'.

Ken G
2010-Jul-26, 03:30 AM
I do not have the grasp of physics to say what likelihood there may be that an 'unknown unknown' would invalidate the foundation of knowledge provided by modern observation of the universe.Nor does anyone else, who is faithfully using science as their epistemological too, because it is not possible for science to convey knowledge of unknown unknowns.

It seems this frontier of knowledge would be seen through analysis of inconsistencies in the standard models.Those are just the known unknowns, but I'll grant you that when you find known unknowns, you probably shouldn't be too surprised if a few unknown unknowns come along for the ride.


Postulation of explanations such as mass that does not engage with light (dark matter), and also possibly string theory, pushes towards both the 'known unknown' and the 'unknown unknown'.Right, and scientists have always had something on the frontier that was just as troubling. Their choice has always been to be unimaginative and assume that the problems are not harbingers of completely new scientific discoveries, or be imaginative and assume the opposite. But it seems to me that the general tendency throughout history has always been to overestimate the reliability of the current state of scientific certainty.

Nereid
2010-Jul-26, 05:43 AM
Returning to this ...
All of Ken G's comments have been couched in terms of some set of, or kind of, theories of physics (and underlying maths). And they have included examples of how the conclusions drawn, from observations, might be skewif.

But doesn't this just raise an even more fundamental question (or set of same)?

For example, what are the unknown unknowns concerning our ideas of distance (wrt stars we cannot visit, in person, in our lifetimes)? And are these unknowns any more, or less, "high up on the continuum" than unknown unknowns in chains leading to effective temperature or surface gravity?Here we see the 'Rumsfeld epistemology',
Totally OT, but why does he get credit for this? I mean, the idea is far, far older than him, surely?


ranging from the 'known knowns', through the 'known unknowns' to the 'unknown unknowns'. There are no 'unknown knowns' because this category is logically inconsistent.

Astronomy is well placed to demarcate between these categories of knowledge.

My impression is that the theory of distance that supports general relativity is accurate, a 'known known'. This theory of distance rests on an assumption that the laws of physics are consistent throughout the universe, and that observation is consistent with a cosmological constant.
The principle that the laws of physics are the same (a stronger condition than merely being consistent) throughout the observable universe is a basic one; take that away then surely there is no astronomical knowledge worth anything at all.

But beyond that principle, to invoke GR, or any theory of distance, is just as arbitrary as any other, isn't it?

I mean, how do you establish a hierarchy, or scale, of theories (etc) in terms of what is more fundamental? Or, in Ken G's words, how do you work out what's higher up on the continuum?

BTW, what do you mean by "and that observation is consistent with a cosmological constant."?



Theories of stellar physics rest on knowledge of a cosmological constant. I do not have the grasp of physics to say what likelihood there may be that an 'unknown unknown' would invalidate the foundation of knowledge provided by modern observation of the universe. It seems this frontier of knowledge would be seen through analysis of inconsistencies in the standard models. Postulation of explanations such as mass that does not engage with light (dark matter), and also possibly string theory, pushes towards both the 'known unknown' and the 'unknown unknown'.
It seems, to me, that you're reducing everything (well, all astronomical knowledge beyond pure observations) to just one test; namely, model consistency. Is that so?

Nereid
2010-Jul-26, 05:53 AM
I do not have the grasp of physics to say what likelihood there may be that an 'unknown unknown' would invalidate the foundation of knowledge provided by modern observation of the universe.Nor does anyone else, who is faithfully using science as their epistemological too, because it is not possible for science to convey knowledge of unknown unknowns.
It certainly seems that way.

However, if there are levels within astronomical knowledge (beyond a division into pure observation and everything else), or a continuum (per a previous post of yours), then aren't we smuggling in judgements about the unknown unknowns?

Back to my previous example: estimates of distance, radius, effective temperature, and surface gravity of a star (not the Sun).

By putting distance and radius higher up on the continuum than effective temperature and surface gravity, aren't we making judgements about the unknown unknowns? For example, aren't we saying that whatever the unknown unknowns about distance are, they will prove to be less disruptive (I'm sure there are better words!) than the unknown unknowns about effective temperature?



[...]
Postulation of explanations such as mass that does not engage with light (dark matter), and also possibly string theory, pushes towards both the 'known unknown' and the 'unknown unknown'.
Right, and scientists have always had something on the frontier that was just as troubling. Their choice has always been to be unimaginative and assume that the problems are not harbingers of completely new scientific discoveries, or be imaginative and assume the opposite. But it seems to me that the general tendency throughout history has always been to overestimate the reliability of the current state of scientific certainty.
So, once again back to astronomical knowledge ...

CDM is pure astronomical knowledge; what puts it lower on the continuum than, say, super-massive black holes (which are also pure astronomical knowledge)?

And, to make a bridge to the Earth, were neutrinos just like CDM is today, in the period between 1931 and 1957 (or 2002)?

Ken G
2010-Jul-26, 11:21 AM
However, if there are levels within astronomical knowledge (beyond a division into pure observation and everything else), or a continuum (per a previous post of yours), then aren't we smuggling in judgements about the unknown unknowns?To me, this "continuum" we are talking about is not a continuum of likelihood of being right or wrong, because we never really get to assess such a likelihood. It's a continuum of violence to our current ideas, that's all-- perhaps even a continuum of violence to the entire process of doing science. If someday we discover that the laws of physics themselves are undergoing gradual aribitrary changes, then the discipline of physics itself would be rocked. If that never happens, we will continue to assume that the laws of physics stay constant, and place that at the "top" of the "continuum." The next tier might include models that demonstrably helped us achieve amazing feats, like getting to the Moon or accelerating particles to TeV energy. The models might be wrong, and might be later shown to have only a kind of coincidental connection with these feats of accomplishment (such as using Newton's laws to fly to the Moon-- I'm not sure if GR was a requirement for getting there), but we can be assured that they would continue to be useful in the same contexts for which they were once useful. Then lower in the "continuum" might be models we could alter overnight without any significant impact on anything we've demonstrably accomplished, like if dark energy is reintepreted as being unnecessary, or if it stems from something other than a cosmological constant. The discovery of dark energy led to no dramatic changes in any other area of physics, and neither would its un-discovery! It is a part of the "house of cards" on which no other cards have yet been placed, so that's the sense to which it is "low" in the continuum.



By putting distance and radius higher up on the continuum than effective temperature and surface gravity, aren't we making judgements about the unknown unknowns?Only if we interpret our placement as an assessment of likely correctness. If we instead look at how many other things we'd have to have wrong, and what kind of fundamental impact it would have on the self-consistent stories we tell, then we are not necessarily assessing unknown unknowns, we are looking at knowns-- we know what our models help us understand, we just don't know if that makes them true. We never get to know that, it's a known unknown.


For example, aren't we saying that whatever the unknown unknowns about distance are, they will prove to be less disruptive (I'm sure there are better words!) than the unknown unknowns about effective temperature?
Yes, but how disruptive something would be can be assessed by how many cards we rely on to be placed on top of it. In principle, anyway.

Jerry
2010-Jul-26, 03:48 PM
CDM is pure astronomical knowledge; what puts it lower on the continuum than, say, super-massive black holes (which are also pure astronomical knowledge)?
CDM is like a fart in a crowd.


And, to make a bridge to the Earth, were neutrinos just like CDM is today, in the period between 1931 and 1957 (or 2002)?
Or 2010. The lack of agreement between the solar neutrino count and that expected from the solar engine may be telling us that we are missing something even more fundamental than a winking neutrino. More fundamental would be a new radius for the proton; based upon the orbital of a muon. On large and small scales, our measurements are only as good as the weakest of our theoretical understandings multiplied by the inaccuracy of our measuring sticks: unknowns multiplied by unknowns.

Which brings us back to black holes, the sump where particle physics and relativistic physics collide and bleed profusely. We can see the blood, be we are kidding ourselves if we think we have the blow by blow description.

Ken G
2010-Jul-26, 04:46 PM
Perhaps one could characterize that stance as the "extremely pessimistic" one, a kind of foil against the "overly optimistic" one that generally prevails. I might not agree with each of the opinions expressed, but I can support the overall point-- it's impossible to use science to know what you don't know, all you can do is seek consistency, and use the cracks in that consistency as a guide to where the next big discoveries may come.

Jerry
2010-Jul-26, 04:47 PM
... Then lower in the "continuum" might be models we could alter overnight without any significant impact on anything we've demonstrably accomplished, like if dark energy is reintepreted as being unnecessary, or if it stems from something other than a cosmological constant. The discovery of dark energy led to no dramatic changes in any other area of physics, and neither would its un-discovery! It is a part of the "house of cards" on which no other cards have yet been placed, so that's the sense to which it is "low" in the continuum.

The problem with a big pile of maybes is that the fundamental tiers can be buried by an undecypherable maze.


Only if we interpret our placement as an assessment of likely correctness. If we instead look at how many other things we'd have to have wrong, and what kind of fundamental impact it would have on the self-consistent stories we tell, then we are not necessarily assessing unknown unknowns, we are looking at knowns-- we know what our models help us understand, we just don't know if that makes them true. ..
Help us understand, or help us engineer? Newton's explaination helped us get to the moon because the mathematical model is a close enough depiction of how a large rocket will behave in near space. The Eygptions had a good enough handle on gravity to build a level pyrimid. In both cases, the theory was not as important as knowing how to use a plumb bob and a square. (Not to mention feeding a very large workforce with visions of granduer.)

Our comfort zone for theories shouldn't necessarily extend as far as our comfort zone for engineering.

On the other hand, I am reasonably comfortable when we look at the spectra of a star and say 'that is a red dwarf, just like our sun, and we can tell that star is moving away from us or towards us at a quantifiable rate because of the small shifts in the spectral lines.' In this case our theory is supported by both geometry and chemistry.

I see three distinct cases:

1) Our observations agree with locally measurable features, such a spectral lines; and we have at least two reasonably independant avenues of investigation. The identification of galaxies as being similar to our own fits in this group because we can study both the spectra and the basic kinetical relationships.

2) We can use at least one well-understood local tool; but we are extrapolating beyond others. Black holes fit here, because we can observe doppler behaviour consistant with an accretion ring. The iffy part is that the theory behind black holes is not fleshed out as well as we would like it to be.

3) Our observations are at odds with our predictions. This is what we find in the invention of Dark Energy and Dark Matter (not to mention the surface chemistry of Titan). In these cases, we are making observations that are at odds with the simplist predictions of how we expect matter to behave. If we had found baryonic dark matter; DM would move into class 1. But we didn't. The confidence in the current model is based upon the elimination of the usual suspects; not a detection whimps either locally or in any spectral signature.

I am very uncomfortable with how DM and DE are presented to the public; and even more so the funding institutions. I wish we heard more of the 'this doesn't make sense' than the 'we aim to prove'.

Ken G
2010-Jul-26, 04:52 PM
I'm more confortable with dark matter and dark energy as things that do "make sense", but I agree with your other points about how to describe "the continuum" here, and also how to avoid sounding more sure than we really have any right to be. We can certainly say that DM and DE are "surprises" relative to our previous experience, but surprises relative to previous experience can still make sense, and indeed that is a pretty fair description of much of the process of making sense that has always been the history of science. The error tends to be in over-extrapolating the "trueness" of the result, rather than, as you put it, the effectiveness of the engineering.

Jerry
2010-Jul-26, 05:06 PM
Perhaps one could characterize that stance as the "extremely pessimistic" one, a kind of foil against the "overly optimistic" one that generally prevails. I might not agree with each of the opinions expressed, but I can support the overall point-- it's impossible to use science to know what you don't know, all you can do is seek consistency, and use the cracks in that consistency as a guide to where the next big discoveries may come.

Skeptical. Pessimism implies a solution set cannot be found; an extremely skeptical view is that knowledgable, bright and well trained persons don't have a clue how dull their tools are.

Robert Tulip
2010-Jul-27, 02:12 PM
Totally OT, but why does he get credit for this? I mean, the idea is far, far older than him, surely?
The wiki (http://en.wikipedia.org/wiki/Unknown_unknown)on 'unknown unknowns' gives prominent place to the former secretary for bringing the term to prominence, but notes that it appeared in military thought from 1984. It also notes that 'unknown knowns' do exist in the form of what I would call error and false belief. As a method of ranking the confidence we have in knowledge and perception it appears the 'unknown unknown' has been discussed more in military strategy than as a formal category in philosophy and epistemology. In astronomy, the unknown unknown may include speculation and imagination about the frontiers of mathematical possibility.

It bears comparison to Plato's allegory of the cave, and his diagram of the divided line, ranging from false belief and appearance through true belief to true knowledge as increasingly reliable categories in epistemology, with mathematical consistency serving to prove knowledge of the truths of reason as more reliable than fleeting sense impressions.


The principle that the laws of physics are the same (a stronger condition than merely being consistent) throughout the observable universe is a basic one; take that away then surely there is no astronomical knowledge worth anything at all.
This axiom of identity is a foundation of knowledge, starting from the observation that there is one universe in which the same laws of physics are universally consistent. I would have thought that consistency is a stronger condition than sameness, as the same laws might conceivably apply inconsistently, in a way that we cannot yet detect.

But beyond that principle, to invoke GR, or any theory of distance, is just as arbitrary as any other, isn't it?
GR provides a consistent story about the distance of galaxies. Other theories do not. Hence it is not arbitrary to prefer a theory that is consistent with observation over possible others. The scales of astronomical distance appear quite robust.

I mean, how do you establish a hierarchy, or scale, of theories (etc) in terms of what is more fundamental? Or, in Ken G's words, how do you work out what's higher up on the continuum?
It is possible to rank theories by the level of confidence that scientists have in their claims. Theories at the core of scientific knowledge have high levels of confidence, while those at the frontier have lower levels.

Ken G has implied that core knowledge might be overthrown, as has happened in the past. I suspect that this is unlikely, simply because the resources devoted to astronomy all tell the same consistent story about core truths of physics and the universe. If so, this suggests that new knowledge is more likely to add to existing knowledge in areas that have not yet been sufficiently studied, rather than to call in question the basic understanding that science now has.

BTW, what do you mean by "and that observation is consistent with a cosmological constant."?
I'm sorry Nereid, I probably shouldn't comment on matters that I understand so little. The wiki on the cosmological constant (http://en.wikipedia.org/wiki/Cosmological_constant) comments "This is the cosmological constant problem, the worst problem of fine-tuning in physics: there is no known natural way to derive the tiny cosmological constant used in cosmology from particle physics." It seems the consistency of a possible cosmological constant with observation is a tough problem.

It seems, to me, that you're reducing everything (well, all astronomical knowledge beyond pure observations) to just one test; namely, model consistency. Is that so?
Model consistency has a logical elegance that serves as a basic test in epistemology. I am not a working scientist, so don't know the practical importance against other factors, but consistency is central to assumptions about what may be productive areas of research. Only those areas that can be analysed by methods that will show consistency with existing knowledge, or reveal a material inconsistency, are productive of new knowledge. An interesting thing in astronomy is that scientists see things they can't explain, like galactic rotation speed, and so have to imagine possible physics in order to work out how observation can be consistent with theory. I would be interested in how important you see model consistency in ranking of reliability of knowledge about stars.

Ken G
2010-Jul-28, 04:03 AM
I would have thought that consistency is a stronger condition than sameness, as the same laws might conceivably apply inconsistently, in a way that we cannot yet detect.I think both you and Nereid are making valid points-- consistency can be a subset of sameness, but sameness can also be a subset of consistency. In short, the two sets are saying fundamentally different things, and we are most interested in their intersection.


GR provides a consistent story about the distance of galaxies. Other theories do not. Hence it is not arbitrary to prefer a theory that is consistent with observation over possible others. Yet don't forget this consistency is only achieved by introducing two new parameters, the dark matter and the dark energy. For some, that's a "cheat", or at the very least, it means that we are not testing GR by asking it to tell a consistent story, we are testing our consistent story by asking it to conform to the structure of GR. For others, it's a kind of combination-- we extrapolate our confidence in GR as a foundation of cosmology, and then we are satisfied that it only took a surprisingly few new parameters to get the story to be consistent. Either way, the new parameters are serving to focus our new observational efforts, so the science is working.


It is possible to rank theories by the level of confidence that scientists have in their claims. Theories at the core of scientific knowledge have high levels of confidence, while those at the frontier have lower levels. I'm not sure I'd call it "levels of confidence", but rather, "levels of holy-cow" if they turned out to be wrong. It is always fraught with peril to make claims about "confidence"-- how confident was Renaissance Europe that the Earth didn't move, after 13 centuries of successful application of Ptolemy's model? How confident was Newton that his laws gave the correct explanation of Kepler's laws (instead of, e.g., Einstein's)? How confident were physicists in 1900 that Maxwell's equations would only hold in an aether frame? How confident was Einstein that physical systems would display a property of local realism? These were all cases where mainstream science was pretty darn confident it has something fundamentally right, and it turned out that it was only effectively or approximately right, while instead being fundamentally wrong. It's not that the observations themselves are later found to be wanting, it's always the interpretive matrix we subject them to that can undergo profound reconstruction, virtually overnight.


Ken G has implied that core knowledge might be overthrown, as has happened in the past. I suspect that this is unlikely, simply because the resources devoted to astronomy all tell the same consistent story about core truths of physics and the universe.But that exact same statement has been true many times in the history of science. Eddington had a theory of the cooling of Earth that said the Earth was millions of years old, and a theory of the contraction of the Sun that said the Sun was also millions of years old. It was a triumph of consistency, and it would have held on for who knows how long, if not for the totally independent geological and biological evidence to the contrary. Do we have such independent evidence on cosmological scales?


If so, this suggests that new knowledge is more likely to add to existing knowledge in areas that have not yet been sufficiently studied, rather than to call in question the basic understanding that science now has.Perhaps. But everyone has always thought that, for millennia.

It seems the consistency of a possible cosmological constant with observation is a tough problem. Quite so-- indeed, some theorists have resorted to anthropic arguments (our universe is selected from very very many as one of few that life can flourish in) to "explain" the fine tuning. In my view, that is a tantamount to giving up on the possibility that science can answer the question at all.


I am not a working scientist, so don't know the practical importance against other factors, but consistency is central to assumptions about what may be productive areas of research.I don't think it matters if someone is a working scientist or not, because frankly, working scientists don't worry much about the consistency of the vast edifice of science. They typically chew on a small part of the larger consistent whole, so they rely on the body of science as a whole to provide them with that consistent basis. It is largely taken for granted, and few step back and ask questions like just how consistent is it all really, and what does it mean that it has the consistency it does, or doesn't. How much is my own research affected by the disunion of quantum mechanics and relativity? Squatola. That's more something that philosophers do, than working scientists-- working scientists tend even to scoff at the value of the enterprise of wondering about consistency. It's a bit like Darwinian evolution into an environmental niche-- the "working scientists" are like the survivable species, and they don't worry too much about why they are survivable or what other species might be even more survivable, they are just busy surviving, and they have faith that the process will progress.

George
2010-Jul-28, 05:02 AM
I think both you and Nereid are making valid points-- consistency can be a subset of sameness, but sameness can also be a subset of consistency. In short, the two sets are saying fundamentally different things, and we are most interested in their intersection. This is interesting, but I'm not sure I understand it. Perhaps the H-R diagram will serve as an example since it is the main sequence consistency that is of special interest because of how stars of the same luminosity and temperature align themselves.


I'm not sure I'd call it "levels of confidence", but rather, "levels of holy-cow" if they turned out to be wrong. :) I like your view best but I suspect the OMG factor is likely directly proportional to the confidence erroneously given it.


It is always fraught with peril to make claims about "confidence"-- how confident was Renaissance Europe that the Earth didn't move, after 13 centuries of successful application of Ptolemy's model? How confident was Newton that his laws gave the correct explanation of Kepler's laws (instead of, e.g., Einstein's)? How confident were physicists in 1900 that Maxwell's equations would only hold in an aether frame? How confident was Einstein that physical systems would display a property of local realism? These were all cases where mainstream science was pretty darn confident it has something fundamentally right, and it turned out that it was only effectively or approximately right, while instead being fundamentally wrong. It's not that the observations themselves are later found to be wanting, it's always the interpretive matrix we subject them to that can undergo profound reconstruction, virtually overnight. Yes, "confidence" does seem to be the wrong word to use, since it isn't a desireable feeling one should hold, especially if the foundation cracks.

Perhaps degrees of "expectation" is a better qualifier for an anticipated result. BBT, for instance, comes with far greater expectations for consistency now that so many different lines of evidence seem to support it.


But that exact same statement has been true many times in the history of science. Eddington had a theory of the cooling of Earth that said the Earth was millions of years old, and a theory of the contraction of the Sun that said the Sun was also millions of years old. It was a triumph of consistency, and it would have held on for who knows how long, if not for the totally independent geological and biological evidence to the contrary. Do we have such independent evidence on cosmological scales?
Perhaps. But everyone has always thought that, for millennia. I'd feel a little better if we knew what gravity is or isn't. ;)


...They typically chew on a small part of the larger consistent whole, so they rely on the body of science as a whole to provide them with that consistent basis. It is largely taken for granted, and few step back and ask questions like just how consistent is it all really, and what does it mean that it has the consistency it does. That's more something that philosophers do, than working scientists-- working scientists tend even to scoff at the value of the enterprise. it's a bit like Darwinian evolution into an environmental niche-- the "working scientists" are like the survivable species, and they don't worry too much about why they are survivable or what other species might be even more survivable, they are just busy surviving, and they have faith that the process will progress. I see it differently because science has the foundational strength of objectivity that philosophy lacks. These foundations are concrete pours spread out all over the Isle of Science and some are stronger than others. Of course, there are a number of rubble piles here and there where the cracks overcame them, but much is being built upon the stronger, tested foundations. [This analogy quickly fails since scientists love to see the foundations crack, and are rewarded for it, as opposed to the building contractors I know. :)]

Ken G
2010-Jul-28, 03:03 PM
This is interesting, but I'm not sure I understand it. Perhaps the H-R diagram will serve as an example since it is the main sequence consistency that is of special interest because of how stars of the same luminosity and temperature align themselves.OK, using that example, we can look at the possibilities. Maybe the laws of physics are different in different places, so stars that have the same luminosity and temperature but are in different places are actually quite different from each other. If we could not tell the consistent story that the H-R diagram allows us to tell, we might have to entertain that possibility-- but we are able to tell a consistent story using the same laws everywhere, so that's what we do. Alternatively, if the H-R diagram did not allow us to tell a consistent story using the same laws, we might try asserting that the laws are the same, but they are inconsistent with each other, so which ones we need to invoke depend on some other contextual issues. That may seem far-fetched, until you recognize that we already have that problem when cobbling together stories told by quantum mechanics and those told by general relativity, but fortunately only on scales that we never actually observe.

In terms of what we actually do observe, we are only interested in the intersection of what is consistent in our laws, and what allows those laws to be the same everywhere. As long as we get away with that approach, it will define physics.

I like your view best but I suspect the OMG factor is likely directly proportional to the confidence erroneously given it.Exactly. (And I too prefer "OMG" to "holy-cow," let's use that.)


Yes, "confidence" does seem to be the wrong word to use, since it isn't a desireable feeling one should hold, especially if the foundation cracks.
And it runs somewhat counter to the way science works. Science progresses via a kind of balancing act between confidence and skepticism-- too much skepticism and you can't use science, too much confidence and you can't advance science.


Perhaps degrees of "expectation" is a better qualifier for an anticipated result. BBT, for instance, comes with far greater expectations for consistency now that so many different lines of evidence seem to support it. True, but I would point out that what that really means is that the "space of observations" has expanded to the point that most new observations would be expected to be "similar" to previous ones. So it's not so much that we know the theory is right, it's that we know the theory applies well to a wide "space" of observations, so it's more unlikely to find an observation outside that space. That's a comfort, and a challenge, to observers.

[This analogy quickly fails since scientists love to see the foundations crack, and are rewarded for it, as opposed to the building contractors I know. ]And that difference is quite important to bear in mind.

George
2010-Jul-28, 07:08 PM
OK, using that example, we can look at the possibilities. Maybe the laws of physics are different in different places, so stars that have the same luminosity and temperature but are in different places are actually quite different from each other. If we could not tell the consistent story that the H-R diagram allows us to tell, we might have to entertain that possibility-- but we are able to tell a consistent story using the same laws everywhere, so that's what we do. Perhaps the old horse and cart analogy will help clarify what I am fairly sure you are stating. If I have it, the horse here is the "sameness" assumed in the laws and the cart of consistency simply tags along as we would expect.


Alternatively, if the H-R diagram did not allow us to tell a consistent story using the same laws, we might try asserting that the laws are the same, but they are inconsistent with each other, so which ones we need to invoke depend on some other contextual issues. Here we seem to be looking for consistency, letting this horse pull the cart of sameness. Is this close? [You have a gift for seeing Nature's more subtle inflections. ]


That may seem far-fetched, until you recognize that we already have that problem when cobbling together stories told by quantum mechanics and those told by general relativity, but fortunately only on scales that we never actually observe. Now you're getting at the root of all of this; it's not so much sameness but saneness. Do I hear a Feynman quote in the background? :)


In terms of what we actually do observe, we are only interested in the intersection of what is consistent in our laws, and what allows those laws to be the same everywhere. As long as we get away with that approach, it will define physics. Yes. But some landscapes are trodden more than others and the expectation is much greater that no "holy cows" or "Ooops" will be forth coming.


And it runs somewhat counter to the way science works. Science progresses via a kind of balancing act between confidence and skepticism-- too much skepticism and you can't use science, too much confidence and you can't advance science. Nicely put.

[I suppose, as a logic barometer, one should have some confidence in one's skepticism, as well as, skepticism of one's confidence, as long as both personal assays are not conducted simultaneously, I'm fairly confident about this. :) ]


True, but I would point out that what that really means is that the "space of observations" has expanded to the point that most new observations would be expected to be "similar" to previous ones. So it's not so much that we know the theory is right, it's that we know the theory applies well to a wide "space" of observations, so it's more unlikely to find an observation outside that space. That's a comfort, and a challenge, to observers. That is certainly wise, yet there still seems to be some justification in assigning some probability differences between various views or models, knowing that things are "subject to change with or without notice" especially for science since the "change" has always been there unobserved.

Extending the terrain/theory analogy, consider the well-mapped and eternal beauty of Yellowstone National Park compared to, say, Europa.

Ken G
2010-Jul-28, 07:39 PM
If I have it, the horse here is the "sameness" assumed in the laws and the cart of consistency simply tags along as we would expect.Actually, I'm trying to avoid answering which is the cart and which the horse, in favor of saying that they are just two different horses that are both pulling the cart of scientific understanding. We can visit the places both horses are comfortable, and we simply have faith we will eventually get everywhere we want to go, but we don't know where those horses can be made to go until we try.

That is certainly wise, yet there still seems to be some justification in assigning some probability differences between various views or models, knowing that things are "subject to change with or without notice" especially for science since the "change" has always been there unobserved.It's the importance of the latter part I would stress-- we seem to have no shortage of the former.


Extending the terrain/theory analogy, consider the well-mapped and eternal beauty of Yellowstone National Park compared to, say, Europa.And yet, some of science's best moments are when the seasoned hiker makes a new discovery and thinks, "suddenly Yellowstone seems more like Europa"-- until he/she becomes a seasoned hiker in the newer vein.

George
2010-Jul-28, 08:30 PM
Actually, I'm trying to avoid answering which is the cart and which the horse, in favor of saying that they are just two different horses that are both pulling the cart of scientific understanding. We can visit the places both horses are comfortable, and we simply have faith we will eventually get everywhere we want to go, but we don't know where those horses can be made to go until we try. Thanks, I now see the two trails you see.


It's the importance of the latter part I would stress-- we seem to have no shortage of the former. I certainly respect your reluctance given your sound logic for caution, which explains why a probability quantification is unlikely.

Of course, for each probability comes improbability. A 40% chance of rain is a 60% chance of no rain. Perhaps a clever [I]improbability quantification would better serve both science and the public. [Suddenly the "George Equation" title seems a bit more apropos than before. :)]


And yet, some of science's best moments are when the seasoned hiker makes a new discovery and thinks, "suddenly Yellowstone seems more like Europa"-- until he/she becomes a seasoned hiker in the newer vein. [Darn, I thought for sure you'd throw on your vulcanist cap for this one.]

peterf
2010-Jul-29, 06:57 AM
nereid, ken g, et al:
i have been reading this thread with great interest and was wondering if you guys would be interested to include a discussion that goes beyond classic scientific materialism?
it is your thread, nereid, and you specifically mention astronomical questions in your opening statement. i have no intentions of "hijacking" this thread and i have no specific agenda, except a life-long fascination with the question "what can and do we know?".
it seems obvious to me that consciousness has very specific properties that no other object of scientific materialism has and i find it puzzling that few people seem to contemplate that fact, or even want to investigate it. most people seem to be happy to just deal with space and time and matter...
i don't really want to open a new thread, since in my experience that tends to attract the "wrong crowd"...

what do you guys think?

Ken G
2010-Jul-29, 09:08 AM
It sounds like a new thread, but the fact is, we know so little about consciousness and its paradoxes that scientific discussions usually cannot get too far with it. I don't think it is avoided because it is viewed as nonscientific or woo-woo, it's just that there's very little solid information outside of the much more focused questions of neuroscience. You already allude to the problem-- we can assume there is such a thing as knowing, and ask about its properties, but if we do not take knowing as axiomatic, and ask how do we know that there is such a thing as knowing, well, you see the problem there! I would say that "knowledge" is a label we hang on certain information, based on whether or not it waddles and quacks like a duck. "Knowing" is nothing but a brand name, and we are just interested in the reputation of the various brands. To understand what knowing actually is, from the standpoint of the mental processes that generate it, gets painfully self-referential!

peterf
2010-Jul-29, 09:25 AM
It sounds like a new thread, but the fact is, we know so little about consciousness and its paradoxes that scientific discussions usually cannot get too far with it. I don't think it is avoided because it is viewed as nonscientific or woo-woo, it's just that there's very little solid information outside of the much more focused questions of neuroscience. You already allude to the problem-- we can assume there is such a thing as knowing, and ask about its properties, but if we do not take knowing as axiomatic, and ask how do we know that there is such a thing as knowing, well, you see the problem there! I would say that "knowledge" is a label we hang on certain information, based on whether or not it waddles and quacks like a duck. "Knowing" is nothing but a brand name, and we are just interested in the reputation of the various brands. To understand what knowing actually is, from the standpoint of the mental processes that generate it, gets painfully self-referential!

1. i don't think we can separate know-ing from the know-er.

2. do i understand you correctly in that you make the assumption that knowing is a mental process as in a neurological process as in a brain-based process?

Ken G
2010-Jul-29, 01:37 PM
1. i don't think we can separate know-ing from the know-er.
Neither do I-- and that's the problem. If we could, then we could know what knowing is.


2. do i understand you correctly in that you make the assumption that knowing is a mental process as in a neurological process as in a brain-based process?
Yes. If the organ that pumps blood is the heart, the organ that knows the heart pumps blood is the brain. So we can study knowing from a neurological perspective, or from a behavioral perspective, or a philosophical perspective, but they're all just pieces. It will be impossible to know what knowing is.

peterf
2010-Jul-29, 02:35 PM
Neither do I-- and that's the problem. If we could, then we could know what knowing is.

what if knowing is not an object?
knowing is consciousness is self-awareness. it is not an object to anything. it is the ultimate subject.


Yes. If the organ that pumps blood is the heart, the organ that knows the heart pumps blood is the brain. So we can study knowing from a neurological perspective, or from a behavioral perspective, or a philosophical perspective, but they're all just pieces. It will be impossible to know what knowing is.

if you regard the brain as a material process - however subtle and complex - how do you explain consciousness or self-awareness?
do the characteristic properties of consciousness (the most important and descriptive being awareness of itself) appear anywhere in the properties of matter or forces or energies or in our mathematical description of them?
are they even hinted at?
i cannot see it anywhere! not even as a hint, as a possibility for extrapolation.
the materialistic pov has no room for consciousness. it is incomplete. we need a much better model for reality, one that must include consciousness.

Ken G
2010-Jul-29, 03:35 PM
what if knowing is not an object?
knowing is consciousness is self-awareness. it is not an object to anything. it is the ultimate subject."Knowing" is a word. Thus it requires a definition that is supplied entirely by us. So the question is never "what is knowing", the question is always, "how do we choose to define knowing, and for what purpose." Science has specific purposes, so defines knowing in regard to those purposes, which has to do with, what will testably turn out to give satisfactory predictions, while at the same time conveying a sense of logic and simplicity? That's all "knowing" means in science, and it has the odd property of being changeable. When we ask a question, and choose among possible answers, we undergo a process quite similar to selecting a product in a supermarket-- that's "knowing" in science. It sounds like you are talking about other kinds of knowing-- more visceral. But that just means selecting a different definition based on your own goals in addressing the concept.


if you regard the brain as a material process - however subtle and complex - how do you explain consciousness or self-awareness?I have no idea. I certainly accept that a better understanding of consciousness may have to go outside what is physically verifiable in the brain, but that's all science has access to.


do the characteristic properties of consciousness (the most important and descriptive being awareness of itself) appear anywhere in the properties of matter or forces or energies or in our mathematical description of them?One might say they appear in the properties of functioning brain matter, and we don't have the mathematics for it yet. Or, one might say that description will never completely describe consciousness. Those statements might both be right.

peterf
2010-Jul-30, 03:53 AM
I certainly accept that a better understanding of consciousness may have to go outside what is physically verifiable in the brain, but that's all science has access to.

if you understand science within the current materialistic paradigm, then yes.

but many contemporary scientists (myself included) suggest a different paradigm. a paradigm where consciousness and not matter is primary and fundamental.
i would go as far as saying that consciousness creates matter. this has nothing to do with theology. no god is needed in this paradigm. in fact, it has a lot to do with quantum physics.
this paradigm can explain the paradoxes we see in qp.
as you know, qp does not regard objects as finally defined "things". rather, objects are regarded as possibilities "waiting" for actualization.
and how does that actualization happen? through an observer. through consciousness giving it attention!

is this just idle and random speculation (as i hear many saying on this forum)?
not any more than the classical materialistic paradigm.
if you want to go beyond solipsism you have to postulate axioms.
the primacy of matter axiom fails to explain many phenomena and creates paradoxes left and right.
the primacy of consciousness axiom is a much more versatile one. and consciousness itself is certainly not an axiom. we all share it. it is self-evident. even to question it requires a consciousness to question it.

this is just a very rough and broad outline of the consciousness paradigm. i am happy to expand if there is interest.

Ken G
2010-Jul-30, 05:43 AM
if you understand science within the current materialistic paradigm, then yes.

but many contemporary scientists (myself included) suggest a different paradigm. a paradigm where consciousness and not matter is primary and fundamental.The problem is deeper than just the paradigm-- how do you do experiments on it? Science does not deal in materialistic paradigms because it has a belief that reality is materialistic, it does because objective experiments are material in nature. To use nonmaterialist paradigms in science, you actually have to be doing something other than science, you have to change what science is. It would make more sense to invent a new word for it, then to change the definition of the old word. That would make it harder to get funding though.

as you know, qp does not regard objects as finally defined "things". rather, objects are regarded as possibilities "waiting" for actualization.
and how does that actualization happen? through an observer. through consciousness giving it attention!
That is certainly one way to frame it, but the question is, do you make any different testable predictions? Does that perspective offer a different direction for tackling problems like the unification of gravity and quantum mechanics? It's a different philosophical stance that commits to a different focus of interest, but it's not clear that the resulting science is any different. I don't think philosophies of science are unimportant, because we never know from what idea a new theory will spring, but the proof is in the pudding. In my view, science clearly stems from an interaction between consciousness (or intelligence, that connection is unclear) and its surroundings, it's a conversation with nature, moreso than a description of nature. Yet that is what it has always been, we just didn't usually call it that-- so recognizing that is what it has always been doesn't change it, it just helps us understand what we've been doing all this time.

if you want to go beyond solipsism you have to postulate axioms.
the primacy of matter axiom fails to explain many phenomena and creates paradoxes left and right.Indeed, I would even say it is merely an illusion that there is any such axiom of science. If science needs matter, poof, there's matter, and if it needs fields, poof, there's fields. Whereever the conversation takes us. Yes, there was some resistance to the importance of fields early on, but their importance was pretty quickly established. Now it's to the point that fields often seem to supercede matter-- we talk about virtual particles and matter-creating fields! So I don't think there's any such axiom-- science is whatever works. But we put the focus on what we are observing, not what our minds are doing, simply because what our minds are doing is part of the structure of the whole endeavor. You use a hammer on a nail, not on another hammer. (And yes, when all you have is a hammer, everything looks like a nail-- important to recognize, but impossible to escape.)


the primacy of consciousness axiom is a much more versatile one. and consciousness itself is certainly not an axiom. we all share it. it is self-evident. even to question it requires a consciousness to question it.
But what does the axiom do? There's nothing wrong with the axiom, it's always been there whether we call attention to it or not. Recognizing that doesn't change how science works, and I don't see how it inspires new directions, other than the standard experiment/theorize/test approach we've always used.

this is just a very rough and broad outline of the consciousness paradigm. i am happy to expand if there is interest.
You might get interest, but it sounds like that should be its own thread. Your challenge will be to demonstrate a difference between such a paradigm, and the standard one. From where I'm sitting, the primacy of matter or consciousness is still the same thing in different clothes-- everything we've ever done has come from the conversation our consciousness is having with nature, no matter what we call it. That's why the same science has been done by die-hard rationalists as that done by die-hard empiricists, the two fit together with hardly a seam in sight. You want to do it from the idealist perspective (a la Berkeley), which is less common, but I would expect would be just as seamless. In my opinion, the real benefit of such an endeavor is not to adjudicate which perspective is "right", but merely to expand our ability to recognize the possible angles. Who knows what insights it could lead to, but it's not an either/or situation. It's certainly a laudable goal to liberate from certain unrecognized prejudices.

Jerry
2010-Jul-30, 05:57 AM
Strange twist on the topic, Peter, but helpful none the less.

A page or two back someone pointed out logic underpins everything. Is the conscious thought process physical? As near as we can tell, if we pull the plug on the brain, the thought process ends. We can also say all brain functions are binary. If knowledge is a burnt finger, then the concious thought process evolved towards a goal of self-preservation.

You can teach a computer program to develop self preservation skills that look for all the world like self-awareness. The most difficult part is tuning the feedback so that the self-protection skill does not become overwhelmingly conservative. The impossible task, is programming an appropriate response to an event that lies completely outside of event contigencies anticipated by the programmer.

I think this applies to science as it is practiced today on many levels. We have two basic sets of rules that have to be renormalized all over hell to tie them together. So adding new parameters is an adaptive skill; just like learning to forage higher in a tree or elevate the Mario brothers to a new level. It is crack filling. We fill cracks because it is the safe solution, not necessarily the best one, or the one that will advance knowledge.


The Curies and others provided us with a solution to the burning problem of the sun; but the solution killed them. A burnt finger is the realization that our senses provide only hints about reality. Moving outside of our comfort zone can introduce us to new opportunities, but it can also kill us. So most of us just work on filling the cracks. It is a conscious choice based upon self-preservation. It is just barely science.

pzkpfw
2010-Jul-30, 07:06 AM
this is just a very rough and broad outline of the consciousness paradigm. i am happy to expand if there is interest.

No.

After moderator discussion, I'd ask that this line be halted here. This is very much an ATM claim, and is off topic for the ATM section of this forum. There are other places to discuss this concept - but not at BAUT.

Robert Tulip
2010-Jul-30, 08:59 AM
what if knowing is not an object?
knowing is consciousness is self-awareness. it is not an object to anything. it is the ultimate subject.
Knowing is a relation between the knower and the known. The object is the known. In astronomy, knowledge is objective, disclosing facts in the universe. The discussion here is about what can be known about stars. This is about the boundaries of objective knowledge, using scientific method. The main sequence (http://en.wikipedia.org/wiki/Main_sequence) is well attested as objective knowledge. Astronomy refutes the claim that "consciousness is self-awareness" and has no object, by reference to objective knowledge of the universe.

peterf
2010-Jul-30, 11:57 AM
No.

After moderator discussion, I'd ask that this line be halted here. This is very much an ATM claim, and is off topic for the ATM section of this forum. There are other places to discuss this concept - but not at BAUT.

:clap:
you really are a funny bunch, aren't ya?

Nereid
2010-Jul-31, 09:45 PM
Here we see the 'Rumsfeld epistemology', ranging from the 'known knowns', through the 'known unknowns' to the 'unknown unknowns'. There are no 'unknown knowns' because this category is logically inconsistent.

Astronomy is well placed to demarcate between these categories of knowledge.

My impression is that the theory of distance that supports general relativity is accurate, a 'known known'. This theory of distance rests on an assumption that the laws of physics are consistent throughout the universe, and that observation is consistent with a cosmological constant.

Theories of stellar physics rest on knowledge of a cosmological constant. I do not have the grasp of physics to say what likelihood there may be that an 'unknown unknown' would invalidate the foundation of knowledge provided by modern observation of the universe. It seems this frontier of knowledge would be seen through analysis of inconsistencies in the standard models. Postulation of explanations such as mass that does not engage with light (dark matter), and also possibly string theory, pushes towards both the 'known unknown' and the 'unknown unknown'.(bold added)

There are unknown knowns, in the sense that one level is known, but its deeper level is not.

A good example is the conservation of energy ... it took quite a while before it became known that this is, "just", a reflection (cough!) of a translation-in-time symmetry. The profundity of symmetries - they are much deeper and broader concepts than naive conservation laws - was not recognised (or discovered) until many centuries after the corresponding conservation laws were first written.

Nereid
2010-Jul-31, 09:49 PM
For example, what are the unknown unknowns concerning our ideas of distance (wrt stars we cannot visit, in person, in our lifetimes)? And are these unknowns any more, or less, "high up on the continuum" than unknown unknowns in chains leading to effective temperature or surface gravity?Nice question. [Are these second order unknowns?] This seems theory specific. Is the physics associated with stellar temperature better, or more robust, than another?(bold added)

Answering the question with a question: how could you go about making an objective, verifiable, robust determination of the rank ordering of various bits of physics, in terms of their robustness?

And if you can't do that, doesn't that reduce all astronomical knowledge to just two levels of knowing, i.e. pure observations and theoretical-derived knowledge?

Nereid
2010-Jul-31, 09:55 PM
[...]

The principle that the laws of physics are the same (a stronger condition than merely being consistent) throughout the observable universe is a basic one; take that away then surely there is no astronomical knowledge worth anything at all.
This axiom of identity is a foundation of knowledge, starting from the observation that there is one universe in which the same laws of physics are universally consistent. I would have thought that consistency is a stronger condition than sameness, as the same laws might conceivably apply inconsistently, in a way that we cannot yet detect.

[...]
Looks like mere semantics; "the same laws might conceivably apply inconsistently" is, to me, hopelessly muddled thinking ... if a law is applied inconsistently, it's not the same law, by definition!

(more later)

Nereid
2010-Jul-31, 10:15 PM
However, if there are levels within astronomical knowledge (beyond a division into pure observation and everything else), or a continuum (per a previous post of yours), then aren't we smuggling in judgements about the unknown unknowns?To me, this "continuum" we are talking about is not a continuum of likelihood of being right or wrong, because we never really get to assess such a likelihood.
Agreed.


It's a continuum of violence to our current ideas, that's all-- perhaps even a continuum of violence to the entire process of doing science. If someday we discover that the laws of physics themselves are undergoing gradual aribitrary changes, then the discipline of physics itself would be rocked.
Agreed, and as that's the bedrock of astronomical knowledge - other than pure observation - it does define one other level (of astronomical knowledge). FWIW, if I've understood some of the ATM ideas of a certain, prolific, BAUTian, it is precisely this kind of violence that his ATM ideas are based on (more later).


If that never happens, we will continue to assume that the laws of physics stay constant, and place that at the "top" of the "continuum." The next tier might include models that demonstrably helped us achieve amazing feats, like getting to the Moon or accelerating particles to TeV energy. The models might be wrong, and might be later shown to have only a kind of coincidental connection with these feats of accomplishment (such as using Newton's laws to fly to the Moon-- I'm not sure if GR was a requirement for getting there), but we can be assured that they would continue to be useful in the same contexts for which they were once useful.
An interesting idea - a level in the continuum is based on historical success in going beyond a few km above sea level. This would - it seems - give GR a status different from, say, theories of particle physics, and thus models of super-massive black holes would be higher on the continuum than models of neutron stars, say, with models of white dwarfs somewhere in between.


Then lower in the "continuum" might be models we could alter overnight without any significant impact on anything we've demonstrably accomplished, like if dark energy is reintepreted as being unnecessary, or if it stems from something other than a cosmological constant. The discovery of dark energy led to no dramatic changes in any other area of physics, and neither would its un-discovery! It is a part of the "house of cards" on which no other cards have yet been placed, so that's the sense to which it is "low" in the continuum.
Aha, the 'upsets how many apple carts?' criterion! :)

Of course this leads to some interesting apparent dilemmas; for example, CDM may well upset the GR apple cart, the particle physics apple cart, the quantum mechanics apple cart, some combo of the preceding, or none at all! Until we know which one it does upset, if any, how can we decide where in the continuum CDM, as astronomical knowledge, should be placed?



By putting distance and radius higher up on the continuum than effective temperature and surface gravity, aren't we making judgements about the unknown unknowns?Only if we interpret our placement as an assessment of likely correctness. If we instead look at how many other things we'd have to have wrong, and what kind of fundamental impact it would have on the self-consistent stories we tell, then we are not necessarily assessing unknown unknowns, we are looking at knowns-- we know what our models help us understand, we just don't know if that makes them true. We never get to know that, it's a known unknown.
How, then, does this differ from the criterion of consistency of models, extended to the whole of the astronomical knowledge domain?

Besides, the unknown unknown that makes distance and radius, of stars say, as astronomical knowledge wrong may, in the end, not do any particular violence to more than a tiny, obscure, part of contemporary physics (it may involve only a minor tweak to GR, say). The point is, until we know what the unknown is, we have no way of assessing this ... hence putting distance and radius higher on the continuum does, in fact, amount to smuggling in some judgements about the unknown unknowns.



For example, aren't we saying that whatever the unknown unknowns about distance are, they will prove to be less disruptive (I'm sure there are better words!) than the unknown unknowns about effective temperature?
Yes, but how disruptive something would be can be assessed by how many cards we rely on to be placed on top of it. In principle, anyway.
Surely not!

The whole point of an unknown unknown (well, one of them anyway) is that we have no way of making the kind of assessments you suggest ... if we did, it wouldn't be an unknown unknown (by definition).

Nereid
2010-Jul-31, 10:22 PM
CDM is pure astronomical knowledge; what puts it lower on the continuum than, say, super-massive black holes (which are also pure astronomical knowledge)?CDM is like a fart in a crowd.
Huh? I don't understand how this has any relevance to the, so far, perfectly good discussion in this thread.

Would you care to elaborate?



And, to make a bridge to the Earth, were neutrinos just like CDM is today, in the period between 1931 and 1957 (or 2002)?
Or 2010. The lack of agreement between the solar neutrino count and that expected from the solar engine may be telling us that we are missing something even more fundamental than a winking neutrino.
Of course it might.

And some tiny, apparently utterly trivial - and up-to-now entirely unnoticed - astronomical observation may be telling us that we are missing something incredibly profound about the foundations of quantum mechanics.

Again, how do any of these hypothetical possibilities relate to astronomical knowledge?

(Oh, and btw, solar neutrino observations are beyond the scope of astronomical knowledge, as I defined it in the OP; namely, 'beyond the solar system'.)


More fundamental would be a new radius for the proton; based upon the orbital of a muon. On large and small scales, our measurements are only as good as the weakest of our theoretical understandings multiplied by the inaccuracy of our measuring sticks: unknowns multiplied by unknowns.
What has this got to do with astronomical knowledge?


Which brings us back to black holes, the sump where particle physics and relativistic physics collide and bleed profusely.
You're kidding, right?


We can see the blood, be we are kidding ourselves if we think we have the blow by blow description.
I have no idea what you're saying; would you care to clarify please?

Nereid
2010-Jul-31, 10:27 PM
Perhaps one could characterize that stance as the "extremely pessimistic" one, a kind of foil against the "overly optimistic" one that generally prevails. I might not agree with each of the opinions expressed, but I can support the overall point-- it's impossible to use science to know what you don't know, all you can do is seek consistency, and use the cracks in that consistency as a guide to where the next big discoveries may come.
I disagree, quite strongly.

Jerry's point is, as I understand it, that there is only one kind of astronomical knowledge ... pure observation.

He is saying, in no uncertain terms, that the universal sameness of the laws of physics cannot be used to make decisions about astronomical knowledge; therefore, there is no astronomical knowledge (except, perhaps, pure observation). IOW, the most speculative ideas about multiverses (say) are equivalent to estimates of the distance and radius of some (fairly close) star, in terms of what we can know, in astronomy.

Ken G
2010-Jul-31, 11:01 PM
I disagree, quite strongly.

Jerry's point is, as I understand it, that there is only one kind of astronomical knowledge ... pure observation.
My interpretation is that he is merely expressing doubt that our current theories are the final word.


He is saying, in no uncertain terms, that the universal sameness of the laws of physics cannot be used to make decisions about astronomical knowledge; therefore, there is no astronomical knowledge (except, perhaps, pure observation). IOW, the most speculative ideas about multiverses (say) are equivalent to estimates of the distance and radius of some (fairly close) star, in terms of what we can know, in astronomy.I'll let him speak for himself, but I didn't get that-- I just got him as saying we might be quite surprised which of our current models don't survive the next millennium. Neither would I. However, I might not agree with him which ones will!

Nereid
2010-Aug-01, 12:07 AM
... Then lower in the "continuum" might be models we could alter overnight without any significant impact on anything we've demonstrably accomplished, like if dark energy is reintepreted as being unnecessary, or if it stems from something other than a cosmological constant. The discovery of dark energy led to no dramatic changes in any other area of physics, and neither would its un-discovery! It is a part of the "house of cards" on which no other cards have yet been placed, so that's the sense to which it is "low" in the continuum.The problem with a big pile of maybes is that the fundamental tiers can be buried by an undecypherable maze.
That's a good point, when applied to any particular individual.

However, if there is a community of individuals - astrophysicists, say - to whom the "maze" is not the least bit "undecypherable", then this point has no impact on their astronomical knowledge, does it?



Only if we interpret our placement as an assessment of likely correctness. If we instead look at how many other things we'd have to have wrong, and what kind of fundamental impact it would have on the self-consistent stories we tell, then we are not necessarily assessing unknown unknowns, we are looking at knowns-- we know what our models help us understand, we just don't know if that makes them true. ..
Help us understand, or help us engineer? Newton's explaination helped us get to the moon because the mathematical model is a close enough depiction of how a large rocket will behave in near space. The Eygptions had a good enough handle on gravity to build a level pyrimid. In both cases, the theory was not as important as knowing how to use a plumb bob and a square. (Not to mention feeding a very large workforce with visions of granduer.)
I think this reflects a profound misunderstanding of space missions, even ones to LEO.

For example, building a working GPS requires a rather good grasp of GR, not to mention an acceptance that the laws of physics - in general - are the same here on the surface of the Earth as in LEO. Of course, a lot of engineering is needed too.


Our comfort zone for theories shouldn't necessarily extend as far as our comfort zone for engineering.
And how is this relevant to this thread?

Remember, the scope is astronomical knowledge, defined as starting from the outer boundaries of the solar system. By definition, such knowledge is beyond the reach of mere engineering (in the sense you have used it).


On the other hand, I am reasonably comfortable when we look at the spectra of a star and say 'that is a red dwarf, just like our sun, and we can tell that star is moving away from us or towards us at a quantifiable rate because of the small shifts in the spectral lines.' In this case our theory is supported by both geometry and chemistry.
No, the theory is not supported by either geometry or chemistry (actually, atomic physics).

The theory requires acceptance that the laws of physics are the same, throughout the universe - take away that, and you can know nothing about anything astronomical (beyond pure observations).

Accept that fundamental principle, and you can build models - i.e. theoretically-based mathematical constructs - which are consistent with the observations, and which use geometry and atomic physics. This, then, is just the 'model consistency' that was introduced - and discussed at length - early in this thread.


I see three distinct cases:

1) Our observations agree with locally measurable features, such a spectral lines; and we have at least two reasonably independant avenues of investigation. The identification of galaxies as being similar to our own fits in this group because we can study both the spectra and the basic kinetical relationships.
I don't see how this follows; for example, "locally measurable features" extend only to ~30 au, and the only "locally measurable" star is the Sun. Ergo, everything beyond the solar system is, by definition, beyond the scope of "locally measurable features"



2) We can use at least one well-understood local tool; but we are extrapolating beyond others. Black holes fit here, because we can observe doppler behaviour consistant with an accretion ring.
By definition, all astronomical knowledge (using the definition in the OP) is of this kind, including estimates of the distance and radius of a star; they all involve "extrapolating beyond" "at least one well-understood local tool".


The iffy part is that the theory behind black holes is not fleshed out as well as we would like it to be.

3) Our observations are at odds with our predictions. This is what we find in the invention of Dark Energy and Dark Matter (not to mention the surface chemistry of Titan). In these cases, we are making observations that are at odds with the simplist predictions of how we expect matter to behave. If we had found baryonic dark matter; DM would move into class 1. But we didn't. The confidence in the current model is based upon the elimination of the usual suspects; not a detection whimps either locally or in any spectral signature.
So neutron stars and white dwarfs fit into this category too, especially magnetars?

And, in any case, isn't this just repeating the 'consistency of models' concept?



I am very uncomfortable with how DM and DE are presented to the public; and even more so the funding institutions. I wish we heard more of the 'this doesn't make sense' than the 'we aim to prove'.
How is this relevant to a discussion focussed - intentionally - on the epistemology of astronomical knowledge?

Nereid
2010-Aug-01, 12:27 AM
I'm more confortable with dark matter and dark energy as things that do "make sense", but I agree with your other points about how to describe "the continuum" here, and also how to avoid sounding more sure than we really have any right to be. We can certainly say that DM and DE are "surprises" relative to our previous experience, but surprises relative to previous experience can still make sense, and indeed that is a pretty fair description of much of the process of making sense that has always been the history of science. The error tends to be in over-extrapolating the "trueness" of the result, rather than, as you put it, the effectiveness of the engineering.(bold added)

Do you really? :surprised

The post you're referring to (#89) contains an all-but-explicit rejection of the fundamental principle (universality of laws of physics) that everyone else participating in this thread has (apparently) accepted - use of "tools", the primacy of "engineering" over "physics", confusion over the physics involved ("chemistry" vs "atomic physics"), misuse of "extrapolation", etc, etc, etc.

Perhaps you could clarify?

Nereid
2010-Aug-01, 12:36 AM
(continued)
[...]
BTW, what do you mean by "and that observation is consistent with a cosmological constant."?
I'm sorry Nereid, I probably shouldn't comment on matters that I understand so little. The wiki on the cosmological constant (http://en.wikipedia.org/wiki/Cosmological_constant) comments "This is the cosmological constant problem, the worst problem of fine-tuning in physics: there is no known natural way to derive the tiny cosmological constant used in cosmology from particle physics." It seems the consistency of a possible cosmological constant with observation is a tough problem.

[...]
Thanks.

Back in post #82, you used "cosmological constant" twice:

"This theory [GR] of distance rests on an assumption that the laws of physics are consistent throughout the universe, and that observation is consistent with a cosmological constant" (bold added), and

"Theories of stellar physics rest on knowledge of a cosmological constant".

Re the first: what observations are you referring to?

Re the second: I didn't know that they did; may I ask where you got this idea from?

In both cases, I'm using the meaning of "cosmological constant" in the sense of the (later) post of yours (that I'm quoting).

(more later)

Nereid
2010-Aug-01, 12:39 AM
nereid, ken g, et al:
i have been reading this thread with great interest and was wondering if you guys would be interested to include a discussion that goes beyond classic scientific materialism?
it is your thread, nereid, and you specifically mention astronomical questions in your opening statement. i have no intentions of "hijacking" this thread and i have no specific agenda, except a life-long fascination with the question "what can and do we know?".
it seems obvious to me that consciousness has very specific properties that no other object of scientific materialism has and i find it puzzling that few people seem to contemplate that fact, or even want to investigate it. most people seem to be happy to just deal with space and time and matter...
i don't really want to open a new thread, since in my experience that tends to attract the "wrong crowd"...

what do you guys think?
Sorry for the delay in responding, but I, for one, would very (very, very, very) much like to restrict the scope of this thread to epistemology wrt astronomical knowledge, as defined in the OP.

Consciousness is, indeed, a fascinating topic ... but it has almost nothing to do with astronomical knowledge (beyond the obvious, basic, requirement ...).

Nereid
2010-Aug-01, 01:12 AM
(continued)
[...]

But beyond that principle, to invoke GR, or any theory of distance, is just as arbitrary as any other, isn't it?GR provides a consistent story about the distance of galaxies. Other theories do not. Hence it is not arbitrary to prefer a theory that is consistent with observation over possible others. The scales of astronomical distance appear quite robust.
IOW, model consistency?



I mean, how do you establish a hierarchy, or scale, of theories (etc) in terms of what is more fundamental? Or, in Ken G's words, how do you work out what's higher up on the continuum?It is possible to rank theories by the level of confidence that scientists have in their claims.
So, this case, epistemology inevitably includes "the opinions of experts" as an essential element?

If so, this makes sociology at least an element of epistemology, doesn't it? I mean, objectively (and independently verifiably) "the level of confidence that scientists have in their claims" is at least somewhat tied to things like seniority, number of published papers, size of ego, and so on.


Theories at the core of scientific knowledge have high levels of confidence, while those at the frontier have lower levels.
Which pushes questions of hierarchies of astronomical knowledge, in an epistemological sense, into the territory of how do we know which theories are "at the core of scientific knowledge" (and which are "at the frontier"), and assumes a one-to-one relationship between "levels of confidence" and position on a (one dimensional?) cline from core to frontier. And, incidentally, assumes no psychological or sociological dimension to scientists' confidence.

Have I got it right?


Ken G has implied that core knowledge might be overthrown, as has happened in the past. I suspect that this is unlikely, simply because the resources devoted to astronomy all tell the same consistent story about core truths of physics and the universe. If so, this suggests that new knowledge is more likely to add to existing knowledge in areas that have not yet been sufficiently studied, rather than to call in question the basic understanding that science now has.
But doesn't this beg the question of how to determine the likelihood "that core knowledge might be overthrown "?

I mean, epistemologically speaking, we don't want any classification of astronomical knowledge - into a hierarchy, or a continuum, or a series of discrete levels - to depend on the unknown unknown of such a likelihood, do we?


[...]
It seems, to me, that you're reducing everything (well, all astronomical knowledge beyond pure observations) to just one test; namely, model consistency. Is that so?
Model consistency has a logical elegance that serves as a basic test in epistemology. I am not a working scientist, so don't know the practical importance against other factors, but consistency is central to assumptions about what may be productive areas of research.
I think we need to be careful to distinguish among the many things that "consistency" might be applied to.

In this case, "model consistency" is much, much, much narrower than "consistency with the laws of physics" (whether they are the same, universally, or not), to take just one example.

Would you care to say more about this? Specifically, what sorts of consistency are you using - as a basis for classifying astronomical knowledge, epistemologically, into hierarchies, continua, levels, etc?


Only those areas that can be analysed by methods that will show consistency with existing knowledge,
Is this self-referential? Or are you referring to distinct classes of knowledge, results from lab-based experiments, say, vs pure astronomical observations?


or reveal a material inconsistency,
I'm sorry, what does this mean?


are productive of new knowledge. An interesting thing in astronomy is that scientists see things they can't explain, like galactic rotation speed, and so have to imagine possible physics in order to work out how observation can be consistent with theory.
Which is, it seems to me, a pretty darn good definition of model consistency (in just a few words!) :)


I would be interested in how important you see model consistency in ranking of reliability of knowledge about stars.
I have a very open mind about this ... so far, from this thread, it seems to me that model consistency is lower in the continuum (or hierarchy, or levels, or as a dimension, or ...) than pure observations, and also lower than the fundamental principle of the universality of the laws of physics.

However, wrt what others have posted, this cannot be a unique classifier ... because then CDM models would rank equally, epistemologically, with models of the distance and radius of a star (per my earlier example), and almost no one seems to accept that.

Nereid
2010-Aug-01, 04:16 AM
A quick summary.

Everyone who has participated in this thread so far seems to agree* that astronomical knowledge, as defined in the OP, is no wider than scientific knowledge, and perhaps no wider than physics knowledge.

If so, then all kinds of things 'come along for free'; such as the centrality of consistency (however defined); the role of theories; and the primacy of objective, independently verifiable observations.

Beyond that, we all also seem to agree that there are different levels of astronomical knowledge (or, less restrictively, a hierarchy, a continuum, a multi-dimensional continuum, and so on).

There, I think, the consensus ends.

All but one (more?) of us have also accepted the centrality of the universality of 'the laws of physics', though there is some on-going discussion over 'same' vs 'consistent'; the one outlier seems to hold 'engineering' (whatever that is) to be central, and the universality of the laws of physics to be optional.

Most of us also agree - more or less - that 'model consistency' is a good criterion for ranking astronomical knowledge, but there's been little discussion on what this criterion actually is, much less on how it can be applied, consistently and objectively.

Fair summary? Did I miss any significant consensuses?

* well, no one has explicitly disagreed!

Ken G
2010-Aug-01, 05:14 AM
I don't see how this follows; for example, "locally measurable features" extend only to ~30 au, and the only "locally measurable" star is the Sun. Ergo, everything beyond the solar system is, by definition, beyond the scope of "locally measurable features"
Again I'm uncomfortable speaking for Jerry, but as he hasn't yet responded, I took his meaning of "locally measurable features" to mean laboratory experiments. So we have what we see astronomically, and what we have in the lab, and he is much happier when the two reach a concordance, than when all we have is the astronomical observations and a big ? about what they mean because we have no laboratory analog (like dark matter and dark energy, and perhaps even event horizons of black holes and whatnot, though certainly the theory of general relativity is accessible in the laboratory).

A good example of this general issue is two spectral lines first detected in the Sun-- the lines of "coronium" and "helium" (both named for the Sun where they were first detected). If you look at a periodic table today, you still find helium, because it was later found in the lab to be a bona fide new element. But there's no coronium-- that line turned out to be a line of iron that was ionized much more than anyone expected the Sun to be able to do. If we didn't have laboratory access to these questions, how would we ever know that helium really was a new element, and coronium wasn't? Is dark matter the coronium, or the helium, of the 21st century?

Robert Tulip
2010-Aug-01, 05:24 AM
(continued) Thanks. Back in post #82, you used "cosmological constant" twice: "This theory [GR] of distance rests on an assumption that the laws of physics are consistent throughout the universe, and that observation is consistent with a cosmological constant" (bold added), and "Theories of stellar physics rest on knowledge of a cosmological constant". Re the first: what observations are you referring to? Re the second: I didn't know that they did; may I ask where you got this idea from? In both cases, I'm using the meaning of "cosmological constant" in the sense of the (later) post of yours (that I'm quoting). (more later)
Nereid, thanks for picking up on these comments. How I imagine the cosmological constant, in my ignorance of physics, is as a factor that derives from observation of the acceleration of expansion of the universe, on the one hand, and as a residual in quantum mechanics on the other hand. The inconsistency of relativity and quantum physics seen in the problem of the cosmological constant illustrates the limits of knowledge. Apologies if my use of these terms is wrong or imprecise. What prompted my comment was the supposition a cosmological constant could affect perception of distance. It seems the cosmological constant that derives from the observation of accelerating expansion is part of the house of cards that explains astrophysics, including the size of the universe. I have little idea where it sits in the lower or upper levels of the house, and how the inconsistency with observations at quantum level matters for knowledge of stars. I have the impression that astronomical knowledge of distance is robust, and was wondering aloud about whether different values of a cosmological constant might affect the accuracy of perception of distance. Perhaps this is just a wrong way to picture how a cosmological constant affects the relation between GR and expansion?

Ken G
2010-Aug-01, 05:42 AM
Most of us also agree - more or less - that 'model consistency' is a good criterion for ranking astronomical knowledge, but there's been little discussion on what this criterion actually is, much less on how it can be applied, consistently and objectively.

Fair summary? Did I miss any significant consensuses?
Perhaps there are some more elements to supplement the "consistency" issue that came up. We can all agree that consistency is part of what we are searching for, and we would not be satisfied with less. But I would hesitate to equate "degree of consistency" with the "hierarchy" or "continuum" we've talked about-- to me, consistency is just the first "hurdle", you need that to even be in the race.

To the degree of consistency (which better be pretty darn high to have any confidence in an explanation), I would add two more types of hierarchies or continua: influence and correctness. By influence, I mean how many other things would need to be wrong for this to be wrong (so it's how deeply is it rooted, how far does its tentacles extend). By correctness, I mean how unchanged will it be in the physics and astronomy a millennia from now, assuming continued advancement at the current rate.

I would argue that the consistency, the influence of the consistency, and the correctness of that influence, are three almost completely different things, not a single continuum that correlates across the board. Since Nereid's purpose for the thread seems mostly focused on the last of these, I'd offer the following analogy to describe why it is so separate: call it the parable of Dr. Jekyll and Mr. Hyde:

If a guy named Joe Jekyll commits a series of senseless murders, and newspaper accounts of the man go like "he kept to himself, didn't seem to have many friends. He always seemed depressed or angry about something, like a pressure cooker ready to explode", then we think "it's no wonder he finally popped, it's too bad no one intervened ahead of time." That's like a theory with cracks in consistency and little influence that was eventually overturned, and everyone said "no wonder." It's also the way history tends to paint overturned theories (like, say, geocentrism), because it's what we expect to happen, what we'd like to believe.

But just like with mass murderers, overturned theories don't always work that way. Sometimes the mass murderer had lots and lots of friends, many independent interactions of a purely positive and respectable nature, a genuine pillar of the community. Then the report goes "everyone who knew Dr. Jekyll was shocked by the news, he was the last person they would ever expect to be a mass murderer." Stories like that are much more disturbing, and it is likewise just as difficult to accept the way some well established scientific theories (like geocentrism) could be overturned when they had such a widespread positive influence. But of course the answer is, all those people who commented on his personality only met Dr. Jekyll-- they simply had never met Mr. Hyde. So the combination of consistency and influence is an uncertain guide to correctness.

peterf
2010-Aug-01, 06:14 AM
Sorry for the delay in responding, but I, for one, would very (very, very, very) much like to restrict the scope of this thread to epistemology wrt astronomical knowledge, as defined in the OP.

Consciousness is, indeed, a fascinating topic ... but it has almost nothing to do with astronomical knowledge (beyond the obvious, basic, requirement ...).

no worries, nereid. one VERY would have been sufficient btw, as i stated clearly that i - of course - would respect your decision as op.

besides, a mod who openly displays his military fetish in his avatar had already made it very clear that he doesn't want to hear about consciousness (probably confusing it with conscience)...

Nereid
2010-Aug-01, 08:38 AM
I don't see how this follows; for example, "locally measurable features" extend only to ~30 au, and the only "locally measurable" star is the Sun. Ergo, everything beyond the solar system is, by definition, beyond the scope of "locally measurable features"Again I'm uncomfortable speaking for Jerry, but as he hasn't yet responded, I took his meaning of "locally measurable features" to mean laboratory experiments.
By all means let's wait for a response from him.


So we have what we see astronomically, and what we have in the lab, and he is much happier when the two reach a concordance, than when all we have is the astronomical observations and a big ? about what they mean because we have no laboratory analog (like dark matter and dark energy, and perhaps even event horizons of black holes and whatnot, though certainly the theory of general relativity is accessible in the laboratory).
But that merely makes the 'drawing of the line' even more arbitrary!

We have no laboratory analogue of nuclear degenerate matter (hence neutron stars should be added to the list), and, arguably, electron degenerate matter of the kind expected in white dwarfs.

Nor do we have a laboratory analogue of a core collapse supernova, or a GRB, or ...


A good example of this general issue is two spectral lines first detected in the Sun-- the lines of "coronium" and "helium" (both named for the Sun where they were first detected). If you look at a periodic table today, you still find helium, because it was later found in the lab to be a bona fide new element. But there's no coronium-- that line turned out to be a line of iron that was ionized much more than anyone expected the Sun to be able to do. If we didn't have laboratory access to these questions, how would we ever know that helium really was a new element, and coronium wasn't? Is dark matter the coronium, or the helium, of the 21st century?
I'm glad you mentioned this - and you can add nebulium to the list too - because it illustrates the limitations of 'laboratory access'.

There are plenty of forbidden lines, regularly observed in astronomical spectra, that have never (AFAIK) been observed in any laboratory, the [OIII] 500.7 nm one for example. Yet no one (AFAIK) puts these in the same category as CDM, or supermassive black holes, or neutron stars. But why not? Isn't calling the 500.7 nm line an [OIII] transition just as much an extrapolation from what's measured in any lab as a neutron star is? If not, why not?

Nereid
2010-Aug-01, 08:51 AM
(continued) Thanks. Back in post #82, you used "cosmological constant" twice: "This theory [GR] of distance rests on an assumption that the laws of physics are consistent throughout the universe, and that observation is consistent with a cosmological constant" (bold added), and "Theories of stellar physics rest on knowledge of a cosmological constant". Re the first: what observations are you referring to? Re the second: I didn't know that they did; may I ask where you got this idea from? In both cases, I'm using the meaning of "cosmological constant" in the sense of the (later) post of yours (that I'm quoting). (more later)Nereid, thanks for picking up on these comments. How I imagine the cosmological constant, in my ignorance of physics, is as a factor that derives from observation of the acceleration of expansion of the universe, on the one hand, and as a residual in quantum mechanics on the other hand. The inconsistency of relativity and quantum physics seen in the problem of the cosmological constant illustrates the limits of knowledge. Apologies if my use of these terms is wrong or imprecise. What prompted my comment was the supposition a cosmological constant could affect perception of distance. It seems the cosmological constant that derives from the observation of accelerating expansion is part of the house of cards that explains astrophysics, including the size of the universe. I have little idea where it sits in the lower or upper levels of the house, and how the inconsistency with observations at quantum level matters for knowledge of stars. I have the impression that astronomical knowledge of distance is robust, and was wondering aloud about whether different values of a cosmological constant might affect the accuracy of perception of distance. Perhaps this is just a wrong way to picture how a cosmological constant affects the relation between GR and expansion?
Thanks for the clarification.

I think you have confused two things; namely, observations which can be conveniently given the shorthand 'dark energy' and the cosmological constant.

The former, as pure observations, are just as much the bedrock of astronomical knowledge as type Ia supernovae are, or hard GRBs ... no matter what theories of physics we may have, they remain pure astronomical observations.

The size (value) of any cosmological constant (lambda) - in the relevant GR equations - is not derivable from anything in GR, so even if future astronomical observations turn out to be consistent with lambda of value X, all you'd have is (yet another) model consistency*.

It does seem that models of the universe based on particle physics - and perhaps quantum mechanics in general - have a glaring model inconsistency (120 OOM!!) wrt the cosmological constant.

How should we handle this, in terms of astronomical knowledge? I do not know ... and I think it might be very interesting to discuss! For example, in terms of model inconsistency we may need to consider the really big gorilla in the room; namely, the mutual incompatibility of QM and GR.

* and let's not overlook the fact that there may be GR-based models which explain the dark energy observations with a zero value lambda (google 'void cosmology'), or which introduce new fields, etc.

Ken G
2010-Aug-01, 01:08 PM
We have no laboratory analogue of nuclear degenerate matter (hence neutron stars should be added to the list), and, arguably, electron degenerate matter of the kind expected in white dwarfs.

Nor do we have a laboratory analogue of a core collapse supernova, or a GRB, or ...Right, I think that would be his point, all these models are relatively "low" in the continuum because they only predict our indirect astronomical observations, they aren't corroborated directly. We might point out that they were quite successful in anticipating various observed phenomena (pulsars, classical novae, neutrino fluxes from supernovae, GRB afterglows, etc.), but there could always be coincidences involved when the "influence" dimension doesn't extend to include direct laboratory experiments, and there may be a selection effect that focuses our attention on efforts to increase influence that succeeded, while ignoring efforts that failed on the grounds that they'd eventually "work out". A classic example of that is Eddington's view of the age of the Earth/Sun, where he was able to use his understanding of the evolution of the Sun to correctly infer the heat flux from the interior of the Earth, but he made two false assumptions (that the Sun had no fusion, and the Earth had no magma convection), and the errors effectively cancelled out! So that was a strike against consistency when influence is relatively low-- efforts to expand the influence into geology and biology met with huge difficulties that presaged the errors in the assumptions.

Of course, a counterargument can be leveled using this same individual, Eddington, when he objected to Chandrasekhar's idea that white dwarfs would become gravitationally unstable at 1.4 solar masses, essentially on the grounds of the low "influence" of the idea. Eddington felt that some other physics would appear to make that not happen, and Chandrasekhar had no answer to that because his idea did not connect with any phenomena that was directly observed (the problem dark matter has today). But the "influence" of Chandrasekhar's idea increased dramatically with the discovery of pulars, similar to how the influence of dark matter increased dramatically with the discovery of halos of gravitational lensing, and the contrasting gas dynamics of the Bullet Cluster. So sometimes an idea with initially weak influence experiences drastic increase in influence with time (and the idea "catches on" more and more), and sometimes an idea with initially weak influence is contradicted when it attempts to expand its sphere of consistency and influence.

But what does this have to do with correctness? That's a lot harder to say, because we've seen that some ideas whose correctness was initially judged low by established astronomers (like Eddington and the Chandrasekhar mass) later gain in infiuence, and presumably correctness, but we've also seen ideas whose correctness was judged very high (based on its vast influence) that later turned out to have very low correctness (like the geocentric model). One might argue that these examples are cherry-picked, but I'd ask them to assess the influence and judged correctness in the year 1900 of the following assertions:
1) gravity is a force
2) particles follow trajectories
3) time is absolute
That's the problen-- one would be hard pressed to find three statements with higher influence and more securely judged correctness in all of physics in 1900, but we now regard that high influence as merely conditional on our never having met Mr. Hyde.


Isn't calling the 500.7 nm line an [OIII] transition just as much an extrapolation from what's measured in any lab as a neutron star is? If not, why not?This gets deeply into the tricky scientific issue of what constitutes a "similar observation." This is rarely discussed, but I think is quite crucial. To me, a physical theory always does one thing-- it takes a set of observations, and finds a way to unify their results suitably well to extend them to confident predictions about similar observations. Interpolating between similar observations is essential for science to work at all, but is also easy and highly reliable-- we live in a "smooth" universe. But extrapolating, as you mention, is another story, because it requires knowing how far the concept of "similar observations" extends. Newton's orbital mechanics had never been extended to observations of bodies moving at, say, 1000 km/s, but it would have worked dandy. But 100,000 km/s, well, we all know the problem there. If we have a theory that predicts all the spectral lines we see in the laboratory, and also predicts forbidden lines we only see in space, is that extrapolation or interpolation? I think that question also could be extended to everything else on your list-- if it is interpolation, we should be confident in the predictions, but if it is extrapolation, we just might not have yet met Mr. Hyde. The trouble is, of course, those darn unknown unknowns make it often hard to know how far we are extrapolating.

Nereid
2010-Aug-01, 09:00 PM
We have no laboratory analogue of nuclear degenerate matter (hence neutron stars should be added to the list), and, arguably, electron degenerate matter of the kind expected in white dwarfs.

Nor do we have a laboratory analogue of a core collapse supernova, or a GRB, or ...Right, I think that would be his point, all these models are relatively "low" in the continuum because they only predict our indirect astronomical observations, they aren't corroborated directly. We might point out that they were quite successful in anticipating various observed phenomena (pulsars, classical novae, neutrino fluxes from supernovae, GRB afterglows, etc.), but there could always be coincidences involved when the "influence" dimension doesn't extend to include direct laboratory experiments, and there may be a selection effect that focuses our attention on efforts to increase influence that succeeded, while ignoring efforts that failed on the grounds that they'd eventually "work out". [...](bold added)

I'll leave it to later, when (if?) Jerry responds, to address whatever he posted.

Wrt to this part of your post, *all* astronomical knowledge (other than pure observations) lacks direct corroboration! :surprised

Why? If for no other reason than that we have not, directly, observed anything over distances greater than ~30 au (and, by definition, the astronomy of this thread all refers to (much) greater distances). Further, accepting the universality of the laws of physics, as a foundation principle, doesn't make this go away ... if you accept that principle, then all predicted indirect astronomical observations (actually model outputs) are equivalent, epistemologically speaking.

This, of course, gets back to my example of the distance and radius of a star (other than the Sun) - why should it be higher in the continuum (or otherwise different, in an epistemological sense) than the surface gravity and temperature of the same star? We didn't actually finish discussion of that ...

(to be continued)

Nereid
2010-Aug-01, 09:06 PM
(continued)
[...]
Isn't calling the 500.7 nm line an [OIII] transition just as much an extrapolation from what's measured in any lab as a neutron star is? If not, why not?This gets deeply into the tricky scientific issue of what constitutes a "similar observation." This is rarely discussed, but I think is quite crucial. To me, a physical theory always does one thing-- it takes a set of observations, and finds a way to unify their results suitably well to extend them to confident predictions about similar observations. Interpolating between similar observations is essential for science to work at all, but is also easy and highly reliable-- we live in a "smooth" universe. But extrapolating, as you mention, is another story, because it requires knowing how far the concept of "similar observations" extends. Newton's orbital mechanics had never been extended to observations of bodies moving at, say, 1000 km/s, but it would have worked dandy. But 100,000 km/s, well, we all know the problem there. If we have a theory that predicts all the spectral lines we see in the laboratory, and also predicts forbidden lines we only see in space, is that extrapolation or interpolation? I think that question also could be extended to everything else on your list-- if it is interpolation, we should be confident in the predictions, but if it is extrapolation, we just might not have yet met Mr. Hyde. The trouble is, of course, those darn unknown unknowns make it often hard to know how far we are extrapolating.
This being so, are not an association of a line at 500.7 nm with a [OIII] transition, and an association of {insert appropriate set of pure astronomical observations here} with a gravitationally bound body comprised almost entirely of nuclear degenerate matter, equivalent, epistemologically?

(to be continued)

Ken G
2010-Aug-01, 09:26 PM
(continued)
This being so, are not an association of a line at 500.7 nm with a [OIII] transition, and an association of {insert appropriate set of pure astronomical observations here} with a gravitationally bound body comprised almost entirely of nuclear degenerate matter, equivalent, epistemologically?
That depends on whether we think we have good reason to count it as interpolation or extrapolation. This doesn't answer the question, it clarifies it in some sense. If you think you understand the atomic physics involved, based on a set of "similar observations" that nevertheless occur at much higher density, you cannot be sure that "Mr. Hyde" doesn't show up at lower density. But we have other astronomical observations of similarly low densities that involve the same lines we can observe at higher densities, and we see no evidence of any problem. So we begin to think that we are interpolating, not extrapolating, to the forbidden lines (although we cannot know for sure).

On the other hand, if the "pure astronomical observations" are something that is happening in neutron degenerate matter, it sounds more like we are extrapolating, and the potential for a "Mr. Hyde" is much larger. Indeed, that's why we have so many models of neutron stars, and aren't sure which one is yet the best.

Nereid
2010-Aug-01, 10:51 PM
This being so, are not an association of a line at 500.7 nm with a [OIII] transition, and an association of {insert appropriate set of pure astronomical observations here} with a gravitationally bound body comprised almost entirely of nuclear degenerate matter, equivalent, epistemologically?That depends on whether we think we have good reason to count it as interpolation or extrapolation. This doesn't answer the question, it clarifies it in some sense. If you think you understand the atomic physics involved, based on a set of "similar observations" that nevertheless occur at much higher density, you cannot be sure that "Mr. Hyde" doesn't show up at lower density. But we have other astronomical observations of similarly low densities that involve the same lines we can observe at higher densities, and we see no evidence of any problem. So we begin to think that we are interpolating, not extrapolating, to the forbidden lines (although we cannot know for sure).

On the other hand, if the "pure astronomical observations" are something that is happening in neutron degenerate matter, it sounds more like we are extrapolating, and the potential for a "Mr. Hyde" is much larger. Indeed, that's why we have so many models of neutron stars, and aren't sure which one is yet the best.
Does the interpolating/extrapolating criterion apply generally, to astronomical knowledge?

And, epistemologically speaking, does the interpolating/extrapolating criterion yield an objective, independently verifiable, binary result? or is it more of a cline?

Let's take some of the tests of GR: we have the GPB results, and astronomical observations of pulsars.

To what extent - or extents - do the astronomically-based tests of GR (the ones to do with pulsars) fit when we shine the epistemological interpolation/extrapolation light on them?

Ken G
2010-Aug-02, 02:40 AM
Does the interpolating/extrapolating criterion apply generally, to astronomical knowledge?
I'd extend it to all scientific knowledge of every kind-- anything that relies on the concept of "similar observations" (even if it does not overtly recognize that reliance).


And, epistemologically speaking, does the interpolating/extrapolating criterion yield an objective, independently verifiable, binary result? or is it more of a cline?
The distinction is artificial, it's not black and white-- just another way to talk about our "continuum."


Let's take some of the tests of GR: we have the GPB results, and astronomical observations of pulsars.

To what extent - or extents - do the astronomically-based tests of GR (the ones to do with pulsars) fit when we shine the epistemological interpolation/extrapolation light on them?
Because of GPB, and other laboratory-type experiments, GR has a lot of independent avenues to build up "influence." At the same time, that expands the domain of "similar observations" that GR appears to work for. That helps it feel like we are interpolating, not extrapolating. But when we introduce dark matter and dark energy, those begin to feel more like extrapolation. We'd sure like some independent avenues to study them-- to protect us from the kinds of coincidences and selected interpretations that bit Eddington!

Robert Tulip
2010-Aug-02, 07:47 AM
"the same laws might conceivably apply inconsistently" is, to me, hopelessly muddled thinking ... if a law is applied inconsistently, it's not the same law, by definition!
Please excuse my muddling here, I was thinking partly on the analogy of statute law, where application of law is inconsistent in different jurisdictions for the same statute. Law in physics is obviously different. A law such as conservation of mass-energy applies consistently.

Sorry if it is another muddled question, but could a variable in the laws of physics, such as mass, vary across the universe in ways that we cannot detect?

I'm wondering here about the chance of a similar change as the shift from classical physics to relativity, and the chance that a piece of the puzzle as yet undiscovered would provide coherence and require revision of current theory. My prejudice is to say the room for such revision is small, due to the sheer consistency of scientific findings.

A good summary of the main laws of physics is at http://physics.about.com/od/physics101thebasics/p/PhysicsLaws.htm.

Ken G
2010-Aug-02, 03:19 PM
Please excuse my muddling here, I was thinking partly on the analogy of statute law, where application of law is inconsistent in different jurisdictions for the same statute. Law in physics is obviously different. A law such as conservation of mass-energy applies consistently. Are you sure they are so different? I think they are very much the same as the "jurisdictions." We have the jurisdiction of Newtonian mechanics (low speed, weak gravity, large scales), the jurisdiction of quantum mechanics (low speed, weak gravity, small scales which can be extrapolated to large scales because they were built on the back of large-scale physics, but which no one ever actually uses them in that context), the jurisdiction of general relativity (any speed, any gravity, though no one ever uses it unless the speed and gravity are fairly substantial, large scales, with untestable extension to small scales because gravity is too weak until you get to scales so small we can't test them). Also those only work for simple systems, for complex systems we have the jurisdictions of thermodynamics and statistical mechanics, which are neither consistent nor inconsistent with the other jurisdictions, they are just completely different .


A good summary of the main laws of physics is at http://physics.about.com/od/physics101thebasics/p/PhysicsLaws.htm.
And note that those "laws" are not a consistent set. Physicists routinely navigate an inconsistent theoretical landscape-- they do it simply by picking and choosing what "laws" they wish to apply, based on considerations of expediency. So I'm not sure what Nereid means that the same law cannot be applied inconsistently-- that seems to me to be quite a routine application of the body of physical laws. Perhaps it has to do with the meaning of the word "law", which is actually much closer to the human version than we might recognize: it is something our calculations obey whenever we choose to make it so, noting that we may face a penalty, or we might not, if we choose calculations that do not obey some law. Above all, the key point is that calculations that we do obey laws, not reality. We have no idea what reality is doing, we know only what we are doing, and how well it works-- in the various jurisdictions.

Nereid
2010-Aug-02, 07:11 PM
"the same laws might conceivably apply inconsistently" is, to me, hopelessly muddled thinking ... if a law is applied inconsistently, it's not the same law, by definition!Please excuse my muddling here, I was thinking partly on the analogy of statute law, where application of law is inconsistent in different jurisdictions for the same statute. Law in physics is obviously different. A law such as conservation of mass-energy applies consistently.

Sorry if it is another muddled question, but could a variable in the laws of physics, such as mass, vary across the universe in ways that we cannot detect?

I'm wondering here about the chance of a similar change as the shift from classical physics to relativity, and the chance that a piece of the puzzle as yet undiscovered would provide coherence and require revision of current theory. My prejudice is to say the room for such revision is small, due to the sheer consistency of scientific findings.

A good summary of the main laws of physics is at http://physics.about.com/od/physics101thebasics/p/PhysicsLaws.htm.
Perhaps it's time to clarify what the foundational principle which all us - except, perhaps, Jerry - have accepted as necessary for there to be any astronomical knowledge at all, beyond pure observation.

As I am using the principle, which goes by the shorthand "the laws of physics are, everywhere and everywhen, the same", the "laws" are abstract, and they are similar to, but not the same as, theories (in physics).

In terms of what listed in the link in your post, a theory can be totally derivative of another, thermodynamics, say; and the tractability of applying a theory is irrelevant (so, for example, two of the Clay challenges (http://www.claymath.org/millennium/) - Navier-Stokes and Yang-Mills - have no bearing on this aspect of the discussion, unless either one were to be shown to be, formally, internally inconsistent).

Theories always come with a scope, whether that scope is explicit or implicit; the scope of a theory may change, in light of developments in the underlying math (think of how conservation of mass and energy, separately, became conservation of mass-energy, and then just the counterpart to a deep symmetry, that of invariance under translations in time), or observational or experimental results. As such, theories are restricted versions of laws, in that they do not apply everywhere and everywhen (and every other every too).

Now if you want to say that theories can be inconsistently applied, all I think you'd be saying is that their scopes are poorly defined (or not defined at all); once you've tightened up the scope statement - which is an inseparable part of the theory - the inconsistency goes away.

Further, under how I'm treating laws, inconsistency is impossible, by definition.

I hope that at least clarifies what I mean; obviously if, between us, we attach different meanings to key terms - law, theory, consistency, 'same', universal, etc - confusion and misunderstanding are guarranteed.

Nereid
2010-Aug-02, 07:14 PM
Please excuse my muddling here, I was thinking partly on the analogy of statute law, where application of law is inconsistent in different jurisdictions for the same statute. Law in physics is obviously different. A law such as conservation of mass-energy applies consistently.Are you sure they are so different? I think they are very much the same as the "jurisdictions." We have the jurisdiction of Newtonian mechanics (low speed, weak gravity, large scales), the jurisdiction of quantum mechanics (low speed, weak gravity, small scales which can be extrapolated to large scales because they were built on the back of large-scale physics, but which no one ever actually uses them in that context), the jurisdiction of general relativity (any speed, any gravity, though no one ever uses it unless the speed and gravity are fairly substantial, large scales, with untestable extension to small scales because gravity is too weak until you get to scales so small we can't test them). Also those only work for simple systems, for complex systems we have the jurisdictions of thermodynamics and statistical mechanics, which are neither consistent nor inconsistent with the other jurisdictions, they are just completely different .
As I said in the post I just wrote, I think you're confusing "laws" with "theories"; the latter always come with scope statements, the former are truly universal.



A good summary of the main laws of physics is at http://physics.about.com/od/physics1...hysicsLaws.htm.
And note that those "laws" are not a consistent set. Physicists routinely navigate an inconsistent theoretical landscape-- they do it simply by picking and choosing what "laws" they wish to apply, based on considerations of expediency. So I'm not sure what Nereid means that the same law cannot be applied inconsistently-- that seems to me to be quite a routine application of the body of physical laws. Perhaps it has to do with the meaning of the word "law", which is actually much closer to the human version than we might recognize: it is something our calculations obey whenever we choose to make it so, noting that we may face a penalty, or we might not, if we choose calculations that do not obey some law. Above all, the key point is that calculations that we do obey laws, not reality. We have no idea what reality is doing, we know only what we are doing, and how well it works-- in the various jurisdictions.
As you can now see - I hope! - that is not at all what I meant.

Nereid
2010-Aug-02, 07:16 PM
BTW, for "scope" in above, "domain of applicability" may be a better descriptor; in any case, I treat them as equivalent.

Ken G
2010-Aug-02, 11:23 PM
As I said in the post I just wrote, I think you're confusing "laws" with "theories"; the latter always come with scope statements, the former are truly universal.
You know of a universal law? I'm all ears! Remember, they used to call it "the law of universal gravitation" back in Newton's day. The lesson had to be learned yet again-- laws are like brands of cereal, that we choose for their attributes and reputation. Nothing more.


As you can now see - I hope! - that is not at all what I meant.There are several important clarifications that I think are working making:
1) the difference between internal consistency, and what we might call "external" consistency. All laws of physics attempt to be pretty close to internally consistent, and whether or not they succeed generally involves issues of rigor that are beyond most physicists interest level. They are certainly close enough to internally consistent for most people's satisfaction. But external consistency is essentially never true, where by that I mean consistency with the other laws. That means "the laws", taken as a unit, are fundamentally inconsistent. This is the common tradeoff for reaching for greater completeness-- the sacrifice of consistency. Even in Newton's day, the laws of thermodynamics were not strictly consistent with Newton's laws, but statistical mechanics interpreted the differences as involving pathologically unlikely scenarios when the systems were complex enough.
2) the idea that a theory "comes with a scope." That's true on a subjective, human kind of level-- it's not like the theory comes with a postulate "as long as the speed is less than 100,000 km/s." It is up to the scientist to decide what "law" they will invoke in their calculation-- and they suffer the consequences of any errors they make as a result. That is both because the laws are not exact, and because they require idealizations to apply in practice. A large part of theoretical science is figuring out what laws, and what idealizations, are "good enough" for the purpose at hand. But the main point here is again: a law is an instruction for doing a calculation, it is not something that reality itself is doing. So it's not really a "law of nature" at all, there's no such thing-- but we allow ourselves the metaphor because it satisfies our aesthetic sense.

Nereid
2010-Aug-03, 02:50 PM
As I said in the post I just wrote, I think you're confusing "laws" with "theories"; the latter always come with scope statements, the former are truly universal.You know of a universal law? I'm all ears!No, and no one does.


Remember, they used to call it "the law of universal gravitation" back in Newton's day. The lesson had to be learned yet again-- laws are like brands of cereal, that we choose for their attributes and reputation. Nothing more.
The lesson I think that got learned, going from Newton and his universal laws, is that they are not laws, but theories, and that theories come with domains of applicability.

Theories may be good approximations to some, as yet undiscovered, law, within their domains of applicability.



As you can now see - I hope! - that is not at all what I meant.
There are several important clarifications that I think are working making:
1) the difference between internal consistency, and what we might call "external" consistency. All laws of physics attempt to be pretty close to internally consistent, and whether or not they succeed generally involves issues of rigor that are beyond most physicists interest level.
It may not be of much interest to most physicists, but it should be of great interest to us, in discussing astronomical knowledge and epistemology! :)

Besides, you're continuing to use "laws" as a synonym for "theories"; I think the sooner we start making a clear distinction between the two, the sooner we'll get to some good, common ground wrt astronomical knowledge and epistemology, at least wrt what I've called the foundational principle.


They are certainly close enough to internally consistent for most people's satisfaction.
And we, the people discussing this topic here in this thread, should be among the people who are not "most"; don't you think?


But external consistency is essentially never true, where by that I mean consistency with the other laws.
Not laws, theories. And there is, by definition, no "external inconsistency" problem with theories. If two theories are mutually inconsistent all that tells us is that their respective domains of applicability have not been determined accurately enough. If there is, as yet, no way to test hypotheses which aim to examine a region of parameter space in which two such theories seem to be mutually inconsistent, then the best that can be said is that the respective domains of applicability are poorly constrained (in that region of parameter space).


That means "the laws", taken as a unit, are fundamentally inconsistent. This is the common tradeoff for reaching for greater completeness-- the sacrifice of consistency. Even in Newton's day, the laws of thermodynamics were not strictly consistent with Newton's laws, but statistical mechanics interpreted the differences as involving pathologically unlikely scenarios when the systems were complex enough.
2) the idea that a theory "comes with a scope." That's true on a subjective, human kind of level-- it's not like the theory comes with a postulate "as long as the speed is less than 100,000 km/s." It is up to the scientist to decide what "law" they will invoke in their calculation-- and they suffer the consequences of any errors they make as a result. That is both because the laws are not exact, and because they require idealizations to apply in practice. A large part of theoretical science is figuring out what laws, and what idealizations, are "good enough" for the purpose at hand. But the main point here is again: a law is an instruction for doing a calculation, it is not something that reality itself is doing.
Again, that's a theory, not a law.


So it's not really a "law of nature" at all, there's no such thing-- but we allow ourselves the metaphor because it satisfies our aesthetic sense.
Not in the domain of astronomical knowledge as epistemology; if there are no laws, then there is no foundational principle, and thus the only two kinds of astronomical knowledge worth having are pure observation and consistent models. That'd make astronomy a pretty bleak pursuit, wouldn't you say?

Ken G
2010-Aug-03, 04:46 PM
No, and no one does.
Then I am not clear on the distinction you draw between "law" and "theory". I would say the standard meaning is that a theory invokes, or strings together, some set of laws, and whether one is "using a theory" or just solving a problem, one is selecting laws for some practical situation. That same law may appear in different theories in different ways, and any particular use of it might not clearly fall in one theory or another, or maybe it will-- it doesn't matter. When one invokes a law, one never knows if it will hold or not, or if the error it introduces will be decisive or not, but one can make progess by looking to how that same law has been successfully applied to "similar observations." A complete set of such observations can sometimes be tied together into some larger "theory", but often, that step is neither necessary, nor identifiably ever happens.

The lesson I think that got learned, going from Newton and his universal laws, is that they are not laws, but theories, and that theories come with domains of applicability.
But in that usage, there would be no such thing as a law, we may as well strike the term from the lexicon. Instead, what "law" generally means is a kind of "mathematical snippet" incorporated into various theories. As such, "Kepler's laws" are incorporated into, and explained by, the various "Newton's laws", which in turn are sewn together into a larger whole one might call the "theory of classical mechanics." But when one adopts one of Kepler's laws in an application, or one of Newton's, one does not stipulate "I am invoking the theory of classical mechanics", because one usually is not starting from the postulates of that theory-- one is simply invoking a law, with some confidence that it will apply to the desired accuracy, based on knowledge of similar situations. Again, it's choosing a cereal box, based on experience with that brand.


Theories may be good approximations to some, as yet undiscovered, law, within their domains of applicability.
I would argue that theories do not actually come with domains of applicability, they are self-contained mathematical structures designed to treat a wide array of calculations. It is always up to the user to decide if the theory, or any of its laws, are usable in a given application, and that is more determined by the accuracy and pedagogical goals, than by some clearly defined "domain" that comes "with the theory." It's a small distinction, but it underscores the point that laws and theories are ours, not the universe's.


It may not be of much interest to most physicists, but it should be of great interest to us, in discussing astronomical knowledge and epistemology! This is on the issue of rigor, and that is definitely of interest in epistemological issues, but occupies a very different place in the "what do we know" question. Rigor is a strange kind of knowing-- it is built right it, it is always tautological. When you rely on rigor, you only "know" the result because you defined yourself as knowing it. There is only meaningful knowing in regard to the real world when you cross the bridge from rigor into meaning, and you never know if the rigor survives that crossing-- you lose the knowing in favor of having something that means something in the real world. That brings us to Plato vs Aristotle-- Plato says we are in a cave, dimly aware of the "real" things that come from reason and rigor, while Aristotle says that where we live is the reality, and the rigor and concepts are what we create to understand that reality. So Plato says that all horses mimic to some degree the ideal "Horse", whereas Aristotle says that the ideal "Horse" is the way we mentally organize all those horses out there. I am taking Aristotle's stance on scientific epistemology.


Besides, you're continuing to use "laws" as a synonym for "theories"; I think the sooner we start making a clear distinction between the two, the sooner we'll get to some good, common ground wrt astronomical knowledge and epistemology, at least wrt what I've called the foundational principle.I'm still a bit confused what distinction you are making there. Are you making the common distinction that laws are components of theories, or are you making a more fundamental distinction that laws are universally true, and none yet exist, while theories have domains of applicability?


And we, the people discussing this topic here in this thread, should be among the people who are not "most"; don't you think?
Not necessarily-- when I said "most people", I meant "most educated people interested in scientific epistemology", not most people off the street. This is still on the matter of internal consistency (rigor), and the problem is, it is not at all clear that scientific knowledge can, or even should, be internally consistent at the level of a mathematician's rigor. Again, that's the issue of meaning vs. tautology. The greatest mystery in physics is why mathematical rigor works at all (no one uses it in their daily lives, for example, or even other areas of "softer" science, frankly, so why does it work so well in physics?). Indeed, I would argue that rigor works so well in physics it tends to beguile us into not even noticing the difference between tautology (truth by proving) and meaning (truth by testing).


Not laws, theories. Laws can be externally inconsistent. Newton's law of universal gravitation is externally inconsistent with Einstein's equivalence principle, both of which I would count as laws in the standard usage.

And there is, by definition, no "external inconsistency" problem with theories. If two theories are mutually inconsistent all that tells us is that their respective domains of applicability have not been determined accurately enough. I wouldn't put it quite that way, again because theories don't actually have "domains of applicability", usages of theories do. If two theories are inconsistent, then they are so in a rigorous sense in every application, but we can mitigate that problem by selecting one or the other based on the context of our usage. The theory of classical mechanics based on Newton's laws is inconsistent with quantum mechanics because they make different predictions. The latter are much more accurate in some contexts, yet are virtually identical in others, or are impossible to even apply in practice for a third group. The user must navigate those inconsistencies, and whether they have succeeded depends on their goals.


Not in the domain of astronomical knowledge as epistemology; if there are no laws, then there is no foundational principle, and thus the only two kinds of astronomical knowledge worth having are pure observation and consistent models. That'd make astronomy a pretty bleak pursuit, wouldn't you say?I wouldn't, actually-- I'd say that's just exactly what astronomy is: observations and consistent models for understanding them. Not just astronomy, all science. Where comes the need to tell ourselves it is something else, despite all the perfectly clear historical evidence to the contrary? We allow ourselves many metaphors in science, like "laws" and "truth" and what has been "proven", and those metaphors are fine-- metaphors are common to how humans think. We can use the metaphors, without forgetting they are metaphors, and astronomy does not suffer from the clarity of the distinction.

Nereid
2010-Aug-03, 09:25 PM
Ken G, it seems I am not making myself clear at all; let me try again.

First, this whole thread is about epistemology, not physics, nor astrophysics; as such we will have to deal with concepts which do not match well with standard terms in physics.

In particular, we can - and perhaps should - allow philosophical concepts, such as "laws" (per my definition); while they may be meaningless in physics, in epistemology they may be very helpful.

Second, the focus of what I've written is what I've called the foundational principle in the epistemology of astronomical knowledge; namely the universality of the 'laws of physics'. Of course, I'm perfectly happy to discuss what this principle is, or should be, in terms of its detailed specification; however, my starting point is that without this principle, there are only two kinds of astronomical knowledge - pure observation and consistent models (caveats apply, of course). If that is so, then there's little to discuss (IMHO).

Third, whether theories come, in physics, with domains of applicability - explicit or not - is somewhat irrelevant; what this thread is about (to repeat) is epistemology, which is (to repeat again) not physics.

Does this clarify things?

And what say ye, dear lurkers?

Ken G
2010-Aug-03, 10:59 PM
First, this whole thread is about epistemology, not physics, nor astrophysics; as such we will have to deal with concepts which do not match well with standard terms in physics.
Yet, I would say that epistemology here is not separable from physics, instead, we have the epistemology of physics. In other words, there is no such thing as Epistemology in some universal sense-- each subarea decides what "knowing" means. So here, we are talking about what physics and astronomy mean by "knowing", and what they mean by strongly suspecting, and right on down a continuum of possible commitment to the models. I'm saying that oftentimes, the epistemology of physics and astronomy is more a question of the "influence" of various ideas, than any real basis for confidence they represent Truth, again in some universal sense. This is typical in the sciences-- scientific epistemology is more like an effective version of "knowing", moreso than the real mccoy (one can know in mathematics, but at the expense of meaning something real-- the connection there is like a tightrope act that we negotiate so routinely that we forget it's even there).


In particular, we can - and perhaps should - allow philosophical concepts, such as "laws" (per my definition); while they may be meaningless in physics, in epistemology they may be very helpful.If we want an epistemology that works in physics, we need to define the meaning of a "law" in a language that physics can manipulate. It's incoherent to say that there is some meaning of a "Law" that is outside physics, and then ask what we have learned from physics that informs us about such a Law. What we can do, however, is use physics to talk about laws, and then wonder to what extent those laws correspond to some Law we have no language for. However, if we do that, then we are done with the evidence that physics and astronomy can bring to bear when we are done discussing the laws of physics and astronomy. Is that what you mean by the thread being about epistemology?


Second, the focus of what I've written is what I've called the foundational principle in the epistemology of astronomical knowledge; namely the universality of the 'laws of physics'.Universality of laws is a meta-law, but not a Law. All the language we'd need to establish laws, we'd also need to establish their universality-- the only difference is that a meta-law is a law that relates to laws. Both types of law come from us, whereas a Law comes from nature. What I'm calling a law is like a landscape painting or a map, it's not the territory.
Of course, I'm perfectly happy to discuss what this principle is, or should be, in terms of its detailed specification; however, my starting point is that without this principle, there are only two kinds of astronomical knowledge - pure observation and consistent models (caveats apply, of course). If that is so, then there's little to discuss (IMHO).I don't understand what "third kind" of astronomical language you are looking for. The meta-laws apply to the "consistent models" we will make. I can easily make a consistent model that explains everything we see, it says "what happens in any situation is what happens in that exact situation." Not worth much, because it obeys none of the meta-laws of physics and astronomy (universality in regard to place and time, as you are saying, but also as many other symmetry principles as are not expressly broken, and Occam's razor, mathematical logic, etc.). But none of those constraints are anything but the way we choose to do science, they are the meta-laws of physics and astronomy, going into the endeavor. They don't tell us anything about the universe, they tell us about ourselves-- what we are trying to accomplish with the exercise. That they work is another matter-- and an amazing one at that.

So questions that deal with epistemology are, "why did we choose these meta-laws", and "why do these meta-laws work so well", and "how reliable do we expect this particular law, that follows the meta-laws, to be in some situation", but not "the universe obeys a universal Law that we express as a meta-law of science", as that puts the cart before the horse. Our fingerprints are on what we are doing here, we cannot pretend the universe got here first and we didn't disturb anything.


Third, whether theories come, in physics, with domains of applicability - explicit or not - is somewhat irrelevant; what this thread is about (to repeat) is epistemology, which is (to repeat again) not physics.But is it the epistemology of physics, or do you have some other epistemology in mind? There is epistemology of various religions, epistemology of existentialism and various philosophical stances, epistemology of personal revelation, etc. There certainly isn't one thing we could call epistemology, unless we want to talk about all those things.

Len Moran
2010-Aug-04, 07:06 AM
Not in the domain of astronomical knowledge as epistemology; if there are no laws, then there is no foundational principle, and thus the only two kinds of astronomical knowledge worth having are pure observation and consistent models. That'd make astronomy a pretty bleak pursuit, wouldn't you say?




I wouldn't, actually-- I'd say that's just exactly what astronomy is: observations and consistent models for understanding them. Not just astronomy, all science. Where comes the need to tell ourselves it is something else, despite all the perfectly clear historical evidence to the contrary? We allow ourselves many metaphors in science, like "laws" and "truth" and what has been "proven", and those metaphors are fine-- metaphors are common to how humans think. We can use the metaphors, without forgetting they are metaphors, and astronomy does not suffer from the clarity of the distinction.



Second, the focus of what I've written is what I've called the foundational principle in the epistemology of astronomical knowledge; namely the universality of the 'laws of physics'. Of course, I'm perfectly happy to discuss what this principle is, or should be, in terms of its detailed specification; however, my starting point is that without this principle, there are only two kinds of astronomical knowledge - pure observation and consistent models (caveats apply, of course). If that is so, then there's little to discuss (IMHO).



I don't understand what "third kind" of astronomical language you are looking for. The meta-laws apply to the "consistent models" we will make. I can easily make a consistent model that explains everything we see, it says "what happens in any situation is what happens in that exact situation." Not worth much, because it obeys none of the meta-laws of physics and astronomy (universality in regard to place and time, as you are saying, but also as many other symmetry principles as are not expressly broken, and Occam's razor, mathematical logic, etc.). But none of those constraints are anything but the way we choose to do science, they are the meta-laws of physics and astronomy, going into the endeavor. They don't tell us anything about the universe, they tell us about ourselves-- what we are trying to accomplish with the exercise. That they work is another matter-- and an amazing one at that.


When I read the first quote above from Neried, I thought to myself – well isn’t this exactly what science is, but I hesitated to say as much because I wasn't sure what she meant by laws as opposed to consistent models. As I say, to my mind observation and consistent models are exactly what science is.

The responses from Ken G are very much closer I think to what actually is the case. But then Nereid certainly seems to want some kind of accessable universal “ground of things” that makes astronomy something that isn’t a “pretty bleak pursuit” as she says.

So I also am intrigued as to the nature of this “third kind” of language that would restore astronomy to something that Nereid would find to be more worthwhile. I can understand macroscopic laws as underpinning empirical reality, and I can understand that some of these laws could be said to have some “connection” with mind independent reality, but I wouldn't say that laws have some kind of universality such that we can think of them as existing in that form outside of any involvement (however you choose to philosophically frame that involvement) on our part. The "connection" from macroscopic laws to mind independent reality is a philosophical enquiry that I consider to be beyond the scope of science, but that doesn't mean to say that the science we practice is meaningless, it instead means that the enquiry is at all times referenced to our involvement. Observation, intersubjective agreement and consistency of models seems a pretty good combination to me for "understanding" the nature of our reality, (understanding in the sense that we attach importance to groups of impressions that are relatively stable such that we raise those impressions to representations of elements of reality, without implying that we can through science establish the nature of that reality as it "exists" within mind independent reality).

Ken G
2010-Aug-04, 03:30 PM
The way I navigate that tricky terrain is to imagine that science is a conversation between us and nature. We ask the questions, in a language that both we and nature can understand (which is a trick that has required millennia to master), and nature answers, in a language both we and nature can understand. But the interesting thing about that conversation is that there are many parts of our question that we think we understand, but don't notice that nature doesn't, and there are many parts of the answers that nature understands, but we do not. The communication is more like two people who speak different languages trying to communicate by pointing and grunting, yet somehow, some extremely accurate and sophisticated material has crossed that gap. We have learned to communicate with nature, but what are we still missing by not asking the right questions, or by not recognizing the answers when we get them?

The example here is the concept of a "law". If we tried to communicate to a dog what our laws of society mean, how much could a dog really understand? The dog would think we mean whatever laws mean to dogs, like instincts and growls, not courts and jails. So what we mean by a "law of nature" is really "what humans imagine laws of nature are." They are ultimately our laws, not nature's, but there is certainly some kind of communication going on.

Disinfo Agent
2010-Aug-04, 03:53 PM
I would argue that theories do not actually come with domains of applicability, they are self-contained mathematical structures designed to treat a wide array of calculations. It is always up to the user to decide if the theory, or any of its laws, are usable in a given application, and that is more determined by the accuracy and pedagogical goals, than by some clearly defined "domain" that comes "with the theory." It's a small distinction, but it underscores the point that laws and theories are ours, not the universe's.While agreeing with what you've said, I would put it in a slightly different way: theories (and laws) are ours, but their domains of applicability are for the universe to decide. One can never know the domain of applicability of a theory in advance - at least not completely. What you do is try the theory in various particular domains, and see if it succeeds or fails.

Ken G
2010-Aug-04, 04:44 PM
While agreeing with what you've said, I would put it in a slightly different way: theories (and laws) are ours, but their domains of applicability are for the universe to decide. We can agree it is a kind of joint agreement-- we have to say what our goals are, but only the universe can say if we met them or not.


What you do is try the theory in various particular domains, and see if it succeeds or fails.Right, but what constitutes success is defined by us.

Disinfo Agent
2010-Aug-04, 07:04 PM
I think your point is that sometimes a theory is devised to solve only a certain class of problems, and is not expected to solve other kinds of problems. For example, Newton's theory of gravity was developed to describe the motion of macroscopic objects. That it did not explain the motion of the electron around the nucleus was, in a sense, not a failure of the theory, but the result of misapplying it to a problem it had never been meant to solve.

But I'm still not sure I would agree. Especially since so often in science a tool we make to solve a limited class of problems turns out to be useful in a broader context. I'm thinking of wave theory, which was created for mechanical waves, but extends rather neatly to quantum mechanics. Or, to keep within astronomy, Kepler's laws, originally devised for the planets in our solar system, but (I assume) easily extended to the moons of Jupiter.

Or am I misunderstanding you?...

Robert Tulip
2010-Aug-04, 09:39 PM
what we mean by a "law of nature" is really "what humans imagine laws of nature are." They are ultimately our laws, not nature's, but there is certainly some kind of communication going on.

This discussion of epistemology in astronomy is opening many interesting questions. I will later respond to some of the points raised by Nereid.

This comment from Ken G raises the status of scientific law. My view is that modern science has in fact discovered real scientific laws of nature, shown by the observation that when laws are tested they deliver consistent results.

An example from outside astronomy is the law of evolution, that life evolves by cumulative adaptation. Biologists such as Richard Dawkins have said that they expect that extraterrestrial life may be very different from life on earth, but would be similar in evolving by the law of evolution. The argument here is that biology has discovered a universal law for life, moving outside the realm of imagination into the realm of reality. Laws of physics are considered to have similar domain of applicability.

Obviously we cannot know what happens in distant star systems while we are physically confined to the solar system, but observation justifies the inference that laws of physics are consistent across the universe, ie that they disclose reality rather than mere imagination. The qualitative difference from earlier epistemology is that the frontiers of consistency are now so wide, and astronomy has so much objective knowledge, that questions of imagination can be confined to the boundary problems where we lack the tools to see consistent application of law. Where earlier epistemology relied heavily on imagination, astronomy has provided the evidence to base epistemology on observation. This does not of course exclude the possibility that there are laws of nature as yet undiscovered that will reveal further consistency, through a unified theory.

George
2010-Aug-04, 11:27 PM
...science is a conversation between us and nature. I know you've stated this before, but in case I haven't mentioned it, I find this statement very appealing and apt. "Conversation" suggests an active engagement, which is something we can all relate to and we all want to do. It is also a more personalized approach with Nature as she does seem to enjoy revealing answers to those who ask the best questions, and she's not shy about it, either.

I suspect the metaphor could reasonably be extended into helping us understand the distinctions between theory and laws, perhaps. The general views and beliefs (theory) others have can be revealed through conversation, and the specific every-day behavioral practices (laws) can also be discovered, perhaps even before the broader views of the individual are discovered. Of course, we can draw erroneous conclusions about the broader viewpoints of the person, but asking the right questions should eventually give us a good idea of many of the person's broader views. Fortunately, Nature is less fickle than we are, but she is also not human and has broader interests that we have yet to be able to understand enough to even begin asking her the right questions.


We have learned to communicate with nature, but what are we still missing by not asking the right questions, or by not recognizing the answers when we get them? Yes, historically, we have missed much. However, if we look for elegance, we often find favor.


The example here is the concept of a "law". If we tried to communicate to a dog what our laws of society mean, how much could a dog really understand? The dog would think we mean whatever laws mean to dogs, like instincts and growls, not courts and jails. So what we mean by a "law of nature" is really "what humans imagine laws of nature are." They are ultimately our laws, not nature's, but there is certainly some kind of communication going on. The dog that is told "No!" by its master is hearing a command or "law", but the dog may not comprehend the broader picture or context (theory) for which it is given. :)

Ken G
2010-Aug-05, 01:48 AM
I think your point is that sometimes a theory is devised to solve only a certain class of problems, and is not expected to solve other kinds of problems. I'm not saying a theory only works for the problems it is designed for, I'm saying it is only designed for the problems it works for. What it is "designed for" is kind of a moving target, it doesn't really matter the intention of its original crafter. The intention that counts is the intention of whomever would choose to use it, and in what context they would make that choice. No theory comes with instructions on when to use it, instead, we have a body of "similar observations" on which the theory has a proven reputation. Just like a brand name in the supermarket-- if you buy a cereal made by Kellogg's, you think you have some idea what you are getting, but if you buy a frozen pizza by Kellogg's, you may have strayed from the "domain of applicability" of that brand.

Ken G
2010-Aug-05, 01:56 AM
This comment from Ken G raises the status of scientific law. My view is that modern science has in fact discovered real scientific laws of nature, shown by the observation that when laws are tested they deliver consistent results.Is that not just a selection effect? When we go to the store, we try on clothes, and select those that fit us. But we do not conclude those clothes, by virtue of fitting us, were made for us.


An example from outside astronomy is the law of evolution, that life evolves by cumulative adaptation. Biologists such as Richard Dawkins have said that they expect that extraterrestrial life may be very different from life on earth, but would be similar in evolving by the law of evolution. This is an interesting example, because it reveals the hidden assumptions behind Dawkins' position. I would say it is yet another example of the "similar observations" phenomenon. Dawkins assumes that looking at another world is a "similar observation" to looking at ours, in regard to how life develops. That is quite likely true, for a lot of planets out there. But for all of them? Not necessarily-- we may someday study another planet and "meet Mr. Hyde", by which I mean, see life developing in a way that is not a similar observation to our own. For example, we may see life that is cultivated by intelligence, the way dog breeds on Earth were-- then Dawkins' expectation would be wrong in that particular case. Or we may find a way for life to convey itself that does not rely on DNA or any other information storage from generation from generation, but instead has managed to work out some kind of Lamarckian scheme. We always think we'll meet Dr. Jekyll, right up until we stumble onto Mr. Hyde-- and then it makes our head swim how fast we change our expectations.


The argument here is that biology has discovered a universal law for life, moving outside the realm of imagination into the realm of reality. Yes, and that same argument has proven wrong countless times.

Ken G
2010-Aug-05, 02:37 AM
The dog that is told "No!" by its master is hearing a command or "law", but the dog may not comprehend the broader picture or context (theory) for which it is given. I like that metaphor to a point, but where I differ is in the implication that nature follows theories at all. I would say that theories are still very much how we think about nature. Take something like the theory of classical mechanics. We can express this theory in many mathematically equivalent ways, but one unifying principle is the principle of least action. Of all the possible paths that connect two points in space and time that a particle visits, it must follow the path of least action. Now, this appears to be true to some high level of accuracy and idealization, but does that mean that nature is actually calculating actions on the fly, and saying, oops, I can't take that path, there's another with less action? I doubt nature has even the slightest idea what "action" is, it just does whatever is natural to do, and we cook up the concept of action to try and make sense of it-- and like the clothes we try on in the store, we find one that fits. By that reasoning, nature also has no idea what a "theory" is, nature is just thinking "whatever works for you, I think you're very cute when you try to think." And we respond, "hah, got you right to 12 decimal places", to which nature replies "pretty smart, for an ape. Now do it with five particles instead of two. I do it with 1080 of what you think of as particles, and if I had to do it your way, I'd have no idea what I was doing."

Len Moran
2010-Aug-05, 05:57 AM
Newton’s first law states that an object will travel in a straight line unless acted upon by a force, and he conceived of that law in terms of observation, but when did he ever observe a body travelling in a straight line in the absence of a "force"?

The law (in practice) can be thought of being applicable in terms of a very small section of a travelling bodies trajectory, but, in principle, given sufficient resolution, we could detect a curvature on any part of its path in the absence of a local force - can anyone think of a real life object that is known to travel in a straight line if the path of that object was extrapolated over a long distance? So I can’t see how we could ever test this law - how do we know that a body would travel in a straight line in the absence of any interacting force?

Newton first conceived of this law on the basis of what is observed; he assumed that bodies actually attract each other via a gravitational “force” and that this force was an all pervading one, affecting every object in space - and in the absence of this force, objects would travel in a straight line. But no one ever tested that law; it is based entirely on the premise that there is a “force” of gravity.

I know GR complicates the above, and I don’t give the example in terms of that model, I just think it illustrates somewhat the tenuous nature of “laws”. That's not to say of course that the law is meaningless, it just means that it is not intrinsic to nature in the sense that many would like it to be, but it serves us well in life, what more do we want from it?

Ken G
2010-Aug-05, 06:09 AM
So I can’t see how we could ever test this law - how do we know that a body would travel in a straight line in the absence of any interacting force?Good point, and it gets right to the heart of what scientific principles are: ways to organize our thinking in regard to highly idealized situations that never actually occur, in order to make approximate predictions in situations that do. To the extent that the idealization is a good approximation, the prediction will be accurate, but there is no law that does not require idealization. And that's not a bug, it's a feature-- it's another meta-law of science, the idealizations are just what we are looking for.

Galileo was an innovator in this area-- he said that all objects fall the same, in obvious contradiction to the fact that they do not. But his contribution was recognizing that the reason they do not is because of a complication, air resistance, that he could idealize as being absent, even though he could not actually remove it (he mitigated it by rolling things down inclined planes, but that didn't get rid of it altogether). It speaks to the role of idealization in the language of physical laws, which again tells us something about how we like to think about nature, moreso than nature itself. Nature doesn't idealize, if there's air resistance, then things do not fall the same.

That's not to say of course that the law is meaningless, it just means that it is not intrinsic to nature in the sense that many would like it to be, but it serves us well in life, what more do we want from it?Yes, that would seem to be just exactly what we are doing when we seek scientific principles.

Disinfo Agent
2010-Aug-05, 10:36 AM
The lesson I think that got learned, going from Newton and his universal laws, is that they are not laws, but theories, and that theories come with domains of applicability.

Theories may be good approximations to some, as yet undiscovered, law, within their domains of applicability.I agree with what you say here, Nereid, but it seems to conflict with other things you've written. If laws, to you, are a philosophical concept, then why do you speak of 'laws (in physics)' and 'laws of physics'?


[...] this whole thread is about epistemology, not physics, nor astrophysics; as such we will have to deal with concepts which do not match well with standard terms in physics.

In particular, we can - and perhaps should - allow philosophical concepts, such as "laws" (per my definition); while they may be meaningless in physics, in epistemology they may be very helpful.It's not very clear to me what you mean by 'laws'. Could you explain your definition further?


Obviously we cannot know what happens in distant star systems while we are physically confined to the solar system, but observation justifies the inference that laws of physics are consistent across the universe, ie that they disclose reality rather than mere imagination. The qualitative difference from earlier epistemology is that the frontiers of consistency are now so wide, and astronomy has so much objective knowledge, that questions of imagination can be confined to the boundary problems where we lack the tools to see consistent application of law. Where earlier epistemology relied heavily on imagination, astronomy has provided the evidence to base epistemology on observation. This does not of course exclude the possibility that there are laws of nature as yet undiscovered that will reveal further consistency, through a unified theory.

Newton’s first law states that an object will travel in a straight line unless acted upon by a force, and he conceived of that law in terms of observation, but when did he ever observe a body travelling in a straight line in the absence of a "force"?

The law (in practice) can be thought of being applicable in terms of a very small section of a travelling bodies trajectory, but, in principle, given sufficient resolution, we could detect a curvature on any part of its path in the absence of a local force - can anyone think of a real life object that is known to travel in a straight line if the path of that object was extrapolated over a long distance? So I can’t see how we could ever test this law - how do we know that a body would travel in a straight line in the absence of any interacting force?The universality of a particular hypothesis in physics (or of a theory, law, etc.) is a working assumption, not something that we prove empirically first. The underlying principle is that in science we assume that our hypotheses are universal until proven otherwise. This is for pragmatic reasons: a hypothesis with (lots of) exceptions is useless, because it doesn't allow you to make reliable predictions.

George
2010-Aug-05, 01:23 PM
I like that metaphor to a point, but where I differ is in the implication that nature follows theories at all. I would say that theories are still very much how we think about nature. Take something like the theory of classical mechanics. We can express this theory in many mathematically equivalent ways, but one unifying principle is the principle of least action. Of all the possible paths that connect two points in space and time that a particle visits, it must follow the path of least action. Now, this appears to be true to some high level of accuracy and idealization, but does that mean that nature is actually calculating actions on the fly, and saying, oops, I can't take that path, there's another with less action? Yes, but this is not exceptional to the metaphor, which suggests we are receiving incomplete answers of her general behavior due to the shortcomings of our questions asked. We will never know her true behavior, but the better our questions the better we have some ideas of that behavior even if it is circumstantial to other behavior we have not seen yet. But there are times when her dance steps can be well-established (=laws) even if we can't quite hear all the music she is responding to, and never will.


I doubt nature has even the slightest idea what "action" is, it just does whatever is natural to do, and we cook up the concept of action to try and make sense of it-- and like the clothes we try on in the store, we find one that fits. Right, but the "action" I mention is simply a way to describe her behavioral respone that comes with every question we toss her. For instance, she is actively "talking" to those in Geneva conCerned. :)


By that reasoning, nature also has no idea what a "theory" is, nature is just thinking "whatever works for you, I think you're very cute when you try to think." And we respond, "hah, got you right to 12 decimal places", to which nature replies "pretty smart, for an ape. Now do it with five particles instead of two. I do it with 1080 of what you think of as particles, and if I had to do it your way, I'd have no idea what I was doing." Yep, she can be a little sassy at times. These behaviors are harder to put into theory form, no doubt. ;)

George
2010-Aug-05, 01:42 PM
Newton’s first law states that an object will travel in a straight line unless acted upon by a force, and he conceived of that law in terms of observation, but when did he ever observe a body travelling in a straight line in the absence of a "force"?

The law (in practice) can be thought of being applicable in terms of a very small section of a travelling bodies trajectory, but, in principle, given sufficient resolution, we could detect a curvature on any part of its path in the absence of a local force - can anyone think of a real life object that is known to travel in a straight line if the path of that object was extrapolated over a long distance? So I can’t see how we could ever test this law - how do we know that a body would travel in a straight line in the absence of any interacting force? Remove all the letters from an Alpha-Bits cereal and you will have a bowl of milk. Newton asked the right questions. Removing known forces acting upon a body will produce a straighter trajectory proportional to the forces that are removed. The asymptote that appears strongly suggests true straightness is the result. This aymptote was a significant answer from nature that has been vey helpful. But, of course, your point must not be ignored as there are deeper answers to what is at play here. GR reveals that much and more.


I know GR complicates the above, and I don’t give the example in terms of that model, I just think it illustrates somewhat the tenuous nature of “laws”. That's not to say of course that the law is meaningless, it just means that it is not intrinsic to nature in the sense that many would like it to be, but it serves us well in life, what more do we want from it? "Law" is a great word of efficacy. Not only does it give direction but it also restricts liability. If one obey's the law of government, no liability can be assessed. If an engineer's design follows a law of physics and the law itself fails.... its Ken's fault! :)

George
2010-Aug-05, 01:57 PM
Good point, and it gets right to the heart of what scientific principles are: ways to organize our thinking in regard to highly idealized situations that never actually occur, in order to make approximate predictions in situations that do. To the extent that the idealization is a good approximation, the prediction will be accurate, but there is no law that does not require idealization. And that's not a bug, it's a feature-- it's another meta-law of science, the idealizations are just what we are looking for. Yep, and this is why, in practice, it is almost never "Ken's fault" since no "law" goes without such scrutiny first. Physicists understand the behavior of engineers even better than Nature's. :)


Galileo was an innovator in this area-- he said that all objects fall the same, in obvious contradiction to the fact that they do not. But his contribution was recognizing that the reason they do not is because of a complication, air resistance, that he could idealize as being absent, even though he could not actually remove it (he mitigated it by rolling things down inclined planes, but that didn't get rid of it altogether). It speaks to the role of idealization in the language of physical laws, which again tells us something about how we like to think about nature, moreso than nature itself. Nature doesn't idealize, if there's air resistance, then things do not fall the same. Galileo asked the right questions. As a youth he noticed that large hail hit the ground at the same time as small hail. Aristotle said the larger stones will fall faster proportionate to their size, thus the larger stones would have had to have began the drop from higher in the cloud than the small. [This scenario seems to be the one Galileo recognized most.] Galileo asked if this made any sense, and Nature said, "Well, it's about time someone asked this!". His recognition that Aristotle was not only wrong, but grossly wrong, which set the stage for change, and along came modern science, though there were others who contributed, obviously.

[Added: I am not sure that this hail observation wasn't the most important in all of modern science (to help establish modern science). Aristotle had to have his pedestal knocked-out from under him since "science", philosophy and even religion had infused his ideas. Galileo's bigger challenge was to come up with a way to knock this same pedestal out from underneath most otheres, especially those of influence. This created a need to recognize the benefits of observation and experimentation, as well as, get the conversational language (math) working. He had no choice but to demonstrate he was right, and it was his lack of demonstration that got him into trouble on Geocentricity (Tychonic model).]

Len Moran
2010-Aug-05, 05:26 PM
Remove all the letters from an Alpha-Bits cereal and you will have a bowl of milk. Newton asked the right questions. Removing known forces acting upon a body will produce a straighter trajectory proportional to the forces that are removed. The asymptote that appears strongly suggests true straightness is the result.


I see what you are saying here, and obviously, removing “known” forces (like a water jet for example) will change a trajectory in the lab, such that it will follow a straighter path with respect to the initial conditions.

But, purely in terms of Newton, he firstly considered that there is an instantaneous gravitational attraction between bodies. He deduced this (so they say) from an apple falling, but whatever went through his mind, he made an assumption that bodies are attracted to each other due to a gravitational force that pervades all of space. From that assumption he deduces a law that says any object not affected by this force will travel in a straight line. So the law, in this case is based entirely on an assumption that there is a gravitational force between bodies. Nothing wrong with that – the law works very well for us as part of a model that considers a gravitational force to exist between any number of objects.

But now consider another scenario (again forgetting (very respectfully!) about GR). If after Newton, but before Einstein, I presented a mathematical model that explained the curvature of motion, not in terms of forces, but in terms of the conservation of angular momentum, then I would have as much right as Newton to establish a law within my model that does not recognise gravity as being a force between bodies that said: all motion is elliptical in nature unless acted upon by a force.

So now we have two laws regarding the curvature of motion, one based on gravitational force, the other based on conservation of angular momentum, and both appear to contradict each other. But both laws work, so clearly, the common factor here is not an intrinsic law that exists outside of our reality. Rather both laws are entirely referenced to selections we make.

My scenario does actually involve a mathematical model that I know of and (according to the originator) matches all of Newtonian predictions and up to features reserved for GR (the precession of Mercury), so this is not a hypothetical case. I’m not at all suggesting that this model is a competitor to GR, or has anything new to offer, (or that I wish to displace Newton's first law!), and of course you are entitled to question the legitimacy of the model I mention. I only mention it because it gives a good illustration (well, I think it does anyway) of how, many laws within physics involve selections on our part.

I have read that there are some laws that are considered to be “Great” laws, and that they could conceivably thought to be examples showing some kind of intrinsic connection with mind independent reality. Not intrinsic in any familiar sense, more a case of saying that at the very core of nature (which from my philosophical perspective is outside of science) there is “something” that gives rise to these “Great” laws. The one example given of this type of law were Maxwell’s laws. I can’t expand on the legitimacy of that claim of “greatness”, but it was written by someone with some authority in these matters.

George
2010-Aug-05, 10:12 PM
But now consider another scenario (again forgetting (very respectfully!) about GR). If after Newton, but before Einstein, I presented a mathematical model that explained the curvature of motion, not in terms of forces, but in terms of the conservation of angular momentum, then I would have as much right as Newton to establish a law within my model that does not recognise gravity as being a force between bodies that said: all motion is elliptical in nature unless acted upon by a force. Interestingly, you might have been well received by the likes of Kepler or before. :) Prior to Newton, there was no great notion of gravity, though Copernicus used the word itself in de Revolutionibus, but not like we understand it. Further, without the notion of gravity, circular orbits were deemed to be the natural motions of celestial objects. [This was also held by Copernicus, unfortunately. So did Kepler, but he trusted Tycho's data accuracy more than the circular... view (I wan't say logic).] Your ellipitcal solution would have solved some real headaches.


So now we have two laws regarding the curvature of motion, one based on gravitational force, the other based on conservation of angular momentum, and both appear to contradict each other. But both laws work, so clearly, the common factor here is not an intrinsic law that exists outside of our reality. Rather both laws are entirely referenced to selections we make.

My scenario does actually involve a mathematical model that I know of and (according to the originator) matches all of Newtonian predictions and up to features reserved for GR (the precession of Mercury), so this is not a hypothetical case. I’m not at all suggesting that this model is a competitor to GR, or has anything new to offer, (or that I wish to displace Newton's first law!), and of course you are entitled to question the legitimacy of the model I mention. I only mention it because it gives a good illustration (well, I think it does anyway) of how, many laws within physics involve selections on our part. Agreed. I see a "law" as a "stamp of approval" by science that says you can bet your boots it will work, given reasonable boundary conditions. A law is not a theory, but something more like an algorithim -- but not the whole software program (=theory) -- that produces an output for a given input, and it works every time if the input isn't garbage.


I have read that there are some laws that are considered to be “Great” laws, and that they could conceivably thought to be examples showing some kind of intrinsic connection with mind independent reality. Not intrinsic in any familiar sense, more a case of saying that at the very core of nature (which from my philosophical perspective is outside of science) there is “something” that gives rise to these “Great” laws. The one example given of this type of law were Maxwell’s laws. I can’t expand on the legitimacy of that claim of “greatness”, but it was written by someone with some authority in these matters. I do like the idea that laws come in different sizes. :)

Ken G
2010-Aug-05, 11:10 PM
But now consider another scenario (again forgetting (very respectfully!) about GR). If after Newton, but before Einstein, I presented a mathematical model that explained the curvature of motion, not in terms of forces, but in terms of the conservation of angular momentum, then I would have as much right as Newton to establish a law within my model that does not recognise gravity as being a force between bodies that said: all motion is elliptical in nature unless acted upon by a force.That wouldn't quite work, because conservation of angular momentum is a consequence of Newton's laws, and can only predict a fraction of the full motion. By itself it could not replace Newton's laws, and asserting elliptical motion is also not enough because you need to know which ellipse-- they can all conserve angular momentum. You could add one more constraint: 1) conserve angular momentum, 2) all force-free motion is an ellipse with the largest object around being at one focus, 3) the square of the period is proportional to the cube of the semi-major axis, and then you could get all motion in a solar system from looking at any one of its objects. But then you just have Kepler's laws, which were viewed as wanting in the absence of Newton's-- it wasn't viewed as an equivalent theory because it wasn't as generalizable (it wouldn't handle binary stars, for example).

But your general point is well taken-- you are asking for a different-sounding set of laws that make all the same predictions. For that, you could use the principle of least action, or else go to GR and just say the predictions are indistinguishable in some domain. Either way, it makes the point that laws are what we make them, and there is some subjectivity involved in deciding what "the laws" of any given situation actually are.

Len Moran
2010-Aug-06, 06:53 AM
Agreed. I see a "law" as a "stamp of approval" by science that says you can bet your boots it will work, given reasonable boundary conditions. A law is not a theory, but something more like an algorithim -- but not the whole software program (=theory) -- that produces an output for a given input, and it works every time if the input isn't garbage.


That seems to me a very practical and useful definition of a law within physics. It kind of makes a nice distinction between a theory and a law.

It intrigues me how language (and hence our use of the language) is very much connected to defining theories, laws, hypothesis, etc. etc. And yet many would insist that some of these human definitions to be an intrinsic part of nature at it’s most fundamental level, separated entirely from our subjective involvement. And they may even suggest that if only we would all stop “naval gazing” these definitions would establish themselves quite naturally as fundamental aspects of nature without any action needed on our part!



That wouldn't quite work, because conservation of angular momentum is a consequence of Newton's laws, and can only predict a fraction of the full motion. By itself it could not replace Newton's laws, and asserting elliptical motion is also not enough because you need to know which ellipse-- they can all conserve angular momentum. You could add one more constraint: 1) conserve angular momentum, 2) all force-free motion is an ellipse with the largest object around being at one focus, 3) the square of the period is proportional to the cube of the semi-major axis, and then you could get all motion in a solar system from looking at any one of its objects. But then you just have Kepler's laws, which were viewed as wanting in the absence of Newton's-- it wasn't viewed as an equivalent theory because it wasn't as generalizable (it wouldn't handle binary stars, for example).


Yes, I re-read the non mathematical sections of the model I mentioned, and there do seem to be mentioned all sorts of conditions, and in fact it does actually say that:

"assumptions are made and so the model does not attempt to be a new axiomatic physical theory in any accepted sense but rather represents a synthesis of ideas - it is not intended to wipe out and replace Newtonian dynamics. One advantage of the model is that from these postulates, together with observational and empirical evidence, we are able to deduce the general equation for orbital motion without the assumption of Newton's inverse square law involving an in vacuo "gravitational" force".

If I had re-read this first I probably would not have used it to illustrate my point with regards to laws, but then I might not have posted at all, so as you say, it did serve a purpose!

George
2010-Aug-06, 09:17 PM
It intrigues me how language (and hence our use of the language) is very much connected to defining theories, laws, hypothesis, etc. etc. Until coming here, I never really was that interest. The credit, or error, is y'all's. *wink*

The "use" is very important to keeping on the right pathway. The occasional "abuse" of these definitions reveal the greater need to understand their true definition. "A theory is just a theory" is one example of abuse. I have suggested a few more verbal tools be added to the language tool box to eschew obfuscation. "Heliochromology", for instance, is one I have suggested, though this is educational in a more fun way than normal. "Theorino" was once suggested (Cougar) to help bridge the gap that has grown between theory and metaphysics. :)


And yet many would insist that some of these human definitions to be an intrinsic part of nature at it’s most fundamental level, separated entirely from our subjective involvement. One would like to hope there is that connection, but I have come to respect Ken's warnings of fleeting glory for those who suggest any solid contact with Nature. Modeling, mapping and conversations with Nature are descriptors useful to articulating the scientific process.

Jerry
2010-Aug-07, 03:01 AM
Is that not just a selection effect? When we go to the store, we try on clothes, and select those that fit us. But we do not conclude those clothes, by virtue of fitting us, were made for us.

Perfect! Using first principles, a neutron star-like object was predicted and eventually discovered. But there are many, many strange animals out there that require a lot of theoretical jostling to plug them into the puzzle.

Astrophysics is no different from any of the sciences: We make observations, develop hypothesis and test the ideas in either the laboratory or the field. What comes across as different to me, is the approach taken by many when the observations, when the tests, are at-odds with the theory .

Dark Matter is a fart in a crowd: Everyone can smell it, but nobody knows who caused it. We have tested for many of the usual suspects and drawn a blank. I don’t mind calling DM a place holder for something we do not understand; but it is unscientific to bank on the existence of WIMPs when it is equally plausible that we have a theory that smells.

I struggle to find an analogy in another field: On a galactic scale; the rotational velocities are inconsistent with our local experience; unless we propose a type of matter that isn't derived from first principles and is not found in our most sensitive local tests. Perhaps a geological equivalent would be to propose that the gigantic boulders in the snake river plain were nudged about by some kind of dark matter.

***

It is wrong to suppose that launching orbiting rockets confirm today’s iteration of gravitational theory. We developed rockets first, the theory came later. When rocket scientists design a rocket today, they don’t even use first principles. They use thermal, structural and chemical models developed through centuries of trial-and-error. It wouldn’t matter if the rationale behind the equations is wrong as long as the result is a close approximation of reality. This is why, to a rocket scientist, the plumb bob is more important than whether gravity is a force vector or a dimensional vector.

Here are some things we do know:

1) Our theories are not complete - there are gaping holes and bridges.
2) There is a chance that our theories are wrong; or do not extrapolate across galactic space the way our current theories describe.
3) There is no fundamental core theory that is infallible.
4) If observational evidence is at odds with a long standing theory; a single new variable can nil what would otherwise be classified as nullifying evidence.

The universality of the laws we have developed locally are challenged by observational evidence. We have to be very careful about introducing new variables to bail-out an idea that has run out of gas.

Len Moran
2010-Aug-07, 06:19 AM
The universality of a particular hypothesis in physics (or of a theory, law, etc.) is a working assumption, not something that we prove empirically first. The underlying principle is that in science we assume that our hypotheses are universal until proven otherwise. This is for pragmatic reasons: a hypothesis with (lots of) exceptions is useless, because it doesn't allow you to make reliable predictions.


I don’t quite understand the context of this – any expansion would be appreciated.

I can understand that any theory or law, when first established is an open book in terms of its applicability. But surely, that in itself means absolutely nothing doesn’t it? What addition to scientific knowledge (as opposed to scientific enquiry) is a theory or law that is awaiting verification and hence a determination of its domains of applicability?

String theory has not been tested; some say that in principle the bulk of it can never be tested. Now I’m not saying this renders the theory meaningless, but in terms of scientific knowledge (as opposed to scientific enquiry), it tells us nothing. Scientific knowledge (as I understand it in terms of empirical reality) has to involve observational verification; physics is an experimental science after all.

I made the point that no one has ever determined that an object will actually move in a straight line in the absence of a force, so in this sense Newton’s law is only applicable in terms of an assumption that an attractive gravitational force actually exists as “something” between objects. I only make the point to illustrate that the law in this case involved a human selection by Newton that bodies have this gravitational property. But without actually directly testing the law in the absence of any forces, I would not consider the law to be applicable within a model that did not take as its starting point the assumption that gravitational properties of attraction exist within objects. So I wouldn’t call the first law universal in applicability, and neither would I assume it’s applicability in the absence of this direct measurement in terms of scientific knowledge. The only thing that concerns me is it’s observationally verified domain of applicability.

George
2010-Aug-07, 05:51 PM
Dark Matter is a fart in a crowd: Everyone can smell it, but nobody knows who caused it. We have tested for many of the usual suspects and drawn a blank. I don’t mind calling DM a place holder for something we do not understand; but it is unscientific to bank on the existence of WIMPs when it is equally plausible that we have a theory that smells. But being down-wind from those two identifiable bullet-bean consumers (http://chandra.harvard.edu/photo/2006/1e0657/) has helped narrow the field, right?


The universality of the laws we have developed locally are challenged by observational evidence. We have to be very careful about introducing new variables to bail-out an idea that has run out of gas. Will the old variables work? I would assume all here agree that eyes must be wide open to consider every logical alternative to Zwicky's original proposal. Let no one become too spherical. :)

Jerry
2010-Aug-08, 05:59 PM
But being down-wind from those two identifiable bullet-bean consumers (http://chandra.harvard.edu/photo/2006/1e0657/) has helped narrow the field, right?
Absolutely not. MOND theoriest have taken this same event; made a few different assumptions; and declared victiory, too. I don't see how any one could be comfortable taking such an odd event a such great distance and saying 'look how this proves everything'. There are nebula that look all for the world like angels; but they do not prove the existance of the same.


Will the old variables work? I would assume all here agree that eyes must be wide open to consider every logical alternative to Zwicky's original proposal. Let no one become too spherical. :)Assuming no one here was a joint author of the bullet cluster paper that declares the dart matter issue is settled; why just look very closely at these colliding galaxies! There is non-baryonic dark matter snapping its teeth at us!

I like the fact that the simplest construct of string theory can be proven feasible, or all-but-elliminated by the Large Hadron Collider. I like the fact that the parameter space for the Higgs boson can be similarly constrained.

We 'know what we know' when our theories pass the tests we have derived for them. The gravity-B probe was designed to test for two aspects of GR; but the initial results, based upon the experimental design failed miserably. It is disheartening to me when the scientists modeled the discrepancies, and then proposed unanticipated and peculiar forces were blurring the GR signal. Shouldn't a conclusion that the experiment was well designed, and relativity failed be entertained? Can we go back with a similar but better controlled experiment and find out for sure?

The parameter space for gravitational waves is shrinking as we speak. If you read the daily abstracts, you know evidence that charge-based forces are important over larger scales than most thought possible is emerging. We shouldn't be reluctant to load-up our most distant probes with instrumentation that checks are retests fundamental assumptions; so we get a deeper dividing line between what is known, and what is absurd.

dayll
2010-Aug-08, 07:30 PM
Don't blush too much but the high quality of the minds that post on this site is impressive. Kind of humbling actually. But when human knowledge and comprehension is stacked up against what we know of the macro and micro realms that we are embedded in, our ignorance is staggering. Of course all we can do is to keep on grinding, building little edifices of what passes for TRUTH against the background of the great unknown that surrounds us. Probably/Maybe we are a tragic aberration--Consciousness being a freak thing, a doomed and overwhelmed endeavor that in the end reads like a Sartrean cosmic existential tragedy. Brave but doomed consciousness tragically going up against the unimpressed and unconscious jaganatha of the universe (whatever the "universe" may or may not be). But as Gatsby said--we beat on, boats against the current. -dayll

Ken G
2010-Aug-09, 04:19 AM
One would like to hope there is that connection, but I have come to respect Ken's warnings of fleeting glory for those who suggest any solid contact with Nature. You may be referencing the same scene, but I'm reminded of scene late in the movie Patton, where they describe a conquering Caeser who parades before him slaves and riches and exotic animals captured from his latest victory, but with the Caeser is a wise old slave who whispers to him a warning, that all glory is fleeting. Perhaps it could have been Newton or Einstein, instead of Caeser, for even if their contributions are more enduring, they still get superceded eventually.[/QUOTE]

Ken G
2010-Aug-09, 04:35 AM
Of course all we can do is to keep on grinding, building little edifices of what passes for TRUTH against the background of the great unknown that surrounds us. Probably/Maybe we are a tragic aberration--Consciousness being a freak thing, a doomed and overwhelmed endeavor that in the end reads like a Sartrean cosmic existential tragedy. Brave but doomed consciousness tragically going up against the unimpressed and unconscious jaganatha of the universe (whatever the "universe" may or may not be). I like the poetry of that, but personally I think that if all a billion years of time over a billion light years of distance did was to create one single consciousness that could look around at that vastness and say "wow", and then pop out of existence, that event would still be more significant than a billion supernovae or a million galaxies being eaten away at by supermassive black holes. All that other stuff has scope and energy and majesty, but is all a giant waste of time and space without the awareness to notice it, no matter how briefly. At least without evidence that there is some cosmic awareness that does not rely on tiny sentient lifeforms to fill in that part of the puzzle-- maybe all those lifeforms do is weakly and imperfectly tap into an awareness that is already there and doesn't need them to actualize itself, but to us it feels like something pretty important would be missing without us.

George
2010-Aug-09, 07:27 PM
You may be referencing the same scene, but I'm reminded of scene late in the movie Patton, where they describe a conquering Caeser who parades before him slaves and riches and exotic animals captured from his latest victory, but with the Caeser is a wise old slave who whispers to him a warning, that all glory is fleeting. Perhaps it could have been Newton or Einstein, instead of Caeser, for even if their contributions are more enduring, they still get superceded eventually. Yep, from some more distant thread I learned that's a favorite scene for both of us. It's such a poignant lesson in humility that I decided to use it even if it wasn't the perfect phrase for the point at hand. :)

George
2010-Aug-09, 07:55 PM
Absolutely not. MOND theoriest have taken this same event; made a few different assumptions; and declared victiory, too. From what little I know of MOND, they also require DM to account for the galaxy velocities within clusters. So maybe they are pleased. :neutral:


I don't see how any one could be comfortable taking such an odd event a such great distance and saying 'look how this proves everything'. . I assume the scientists aren't going hyperbolic with their views, but perhaps others in the media are.


There are nebula that look all for the world like angels; but they do not prove the existance of the same. Whenever objective measurements can be taken, science will have an impact on subjective claims, angelic or otherwise. Tying the data to DM or MOND is a more subjective step, but objective support is the basis to going from a pure idea to "knowing", though many more objective tests may be necessary.

I don't like the use of "knowing" when only our eyes (aided or otherwise) are the only instruments. If I put my hand in fire, I "know" its too hot because all of my senses are on max. sensitivy screaming me the final unintractable answer. I wanta hold concentrated DM in my hand then hit it with a hammer and not just watch what happens, but feel its resistance, hear the "squish" sound it makes, smell any odor it produces, put a little salt on it and see if tastes like chicken. Only if it hurts me enough will I qualify myself to claim a "knowing" event has been acquired. :)

Disinfo Agent
2010-Aug-10, 12:31 PM
I don’t quite understand the context of this – any expansion would be appreciated.

I can understand that any theory or law, when first established is an open book in terms of its applicability. But surely, that in itself means absolutely nothing doesn’t it? What addition to scientific knowledge (as opposed to scientific enquiry) is a theory or law that is awaiting verification and hence a determination of its domains of applicability?

String theory has not been tested; some say that in principle the bulk of it can never be tested. Now I’m not saying this renders the theory meaningless, but in terms of scientific knowledge (as opposed to scientific enquiry), it tells us nothing. Scientific knowledge (as I understand it in terms of empirical reality) has to involve observational verification; physics is an experimental science after all.

I made the point that no one has ever determined that an object will actually move in a straight line in the absence of a force, so in this sense Newton’s law is only applicable in terms of an assumption that an attractive gravitational force actually exists as “something” between objects. I only make the point to illustrate that the law in this case involved a human selection by Newton that bodies have this gravitational property. But without actually directly testing the law in the absence of any forces, I would not consider the law to be applicable within a model that did not take as its starting point the assumption that gravitational properties of attraction exist within objects. So I wouldn’t call the first law universal in applicability, and neither would I assume it’s applicability in the absence of this direct measurement in terms of scientific knowledge. The only thing that concerns me is it’s observationally verified domain of applicability.We can, in principle, imagine a conjecture put forth in science without any supporting evidence whatsoever. But in practice that never happens. (Or, if it occasionally happens, such conjectures are quickly laughed off in the more reliable sciences; certainly so in physics.)

Hypotheses worth discussing in science are always plausible extensions of what is known at the time. We always build on what came before. String theory illustrates this rather well. There is evidence in support of it: since ST is an extension of quantum mechanics and general relativity, every piece of evidence that supports QM or GR also supports string theory. The problem you allude to is not the lack of evidence in support of string theory, but the lack of discriminating tests allowing us to choose between string theory and other alternatives (thus far).

You might think of invoking Occam's Razor and saying that if all the evidence in favour of string theory is also evidence in favour of simpler theories (GR and QM) then string theory is redundant, and should be discarded. But it's actually more complicated than that: not only does string theory extend GR and QM, but it also resolves inconsistencies between the latter two. Even if no intrinsic evidence for string theory is ever found, ST may still be valuable for its unifying power. Incidentally, consistency and unification are not just logical-mathematical fetishes: they usually improve our understanding of the subject matter and point to useful extensions.

If my idea sounds too bold, consider that the current situation of string theory with regards to supporting evidence is not so different from that of general relativity. People sometimes say that relativity has been tested numerous times since it was proposed, and has always passed with flying colours, but that statement needs to be nuanced. It is certainly true of special relativity, but not of general relativity. From what I understand, precious few tests of general relativity have been performed that weren't also tests of special relativity. There have been a couple of them, nonetheless: this is why GR is treated as established science and string theory isn't. So, GR is a little better proven than string theory, but not by much.

Jerry
2010-Aug-11, 12:27 AM
There is a paper posted today that has application to this thread:

http://arxiv.org/abs/1008.1265

A Multi-color Optical Survey of the Orion Nebula Cluster. II. the H-R diagram


l" We adopt a lower distance for the ONC than previously assumed, based on recent parallax measurements. With a careful choice also of the spectral type-temperature transformation, we produce the new H-R diagram of the ONC population, more populated than previous works. With respect to previous works, we find higher luminosity for late-type stars and a lower luminosity for early types. We determine the age distribution of the population, peaking at 2-3 Myr, a higher age than previously estimated. We study the distribution of the members in the mass-age plane, and find that taking into account selection effects due to incompleteness removes an apparent correlation between mass and age.

In one paper we find authors challenging the HR relationship, the age, the derived luminosities of both early and late stars, the mass/age relationship and the parallax distance. Is nothing sacride?

loglo
2010-Aug-11, 01:35 PM
There is a paper posted today that has application to this thread:

http://arxiv.org/abs/1008.1265

A Multi-color Optical Survey of the Orion Nebula Cluster. II. the H-R diagram



In one paper we find authors challenging the HR relationship, the age, the derived luminosities of both early and late stars, the mass/age relationship and the parallax distance. Is nothing sacride?

They were just talking about stars in the Orion Nebula Complex and how they used new techniques and found the peak in the IMF was lower than that derived in some other papers (but agreeing with others). They then go on to explain why (reddening in the ISM and ONC itself, accretion disks, using a smaller derived distance from a newer study etc etc. I don't see one line about the HR diagram concept being invalidated let alone anything else you mentioned.

Robert Tulip
2010-Aug-12, 06:21 AM
(continued)
IOW, model consistency?
Returning to this earlier set of comments from *122 ( http://www.bautforum.com/showthread.php/105908-Astronomy-what-can-we-know?p=1770668#post1770668), Nereid asks first how model consistency supports knowledge of distance in astronomy. My intuition is to say ‘very strongly’ while deferring to the experts.

So, this case, epistemology inevitably includes "the opinions of experts" as an essential element? If so, this makes sociology at least an element of epistemology, doesn't it? I mean, objectively (and independently verifiably) "the level of confidence that scientists have in their claims" is at least somewhat tied to things like seniority, number of published papers, size of ego, and so on.
‘Confidence’ here is more a reference to consensus views. Established objective knowledge has higher epistemic status than personal beliefs. The subjective factors you mention are more about beliefs that have not been broadly accepted through peer review and empirical confirmation.


Which pushes questions of hierarchies of astronomical knowledge, in an epistemological sense, into the territory of how do we know which theories are "at the core of scientific knowledge" (and which are "at the frontier"), and assumes a one-to-one relationship between "levels of confidence" and position on a (one dimensional?) cline from core to frontier. And, incidentally, assumes no psychological or sociological dimension to scientists' confidence.

Have I got it right?
It depends if you accept a conservative or an adventurous view on the reliability of scientific confidence. Knowledge of stars has both a well-established core, and an adventurous frontier. Some claims that are now at the frontier will one day be accepted in the mainstream. With apologies for ignorance about debates in stellar physics, an example here may be the views of Halton Arp, where what you call a ‘psychological or sociological dimension to scientists' confidence’ has led to claims of knowledge that were later proved wrong. In the balance between caution and adventure, the test of ‘can we imagine this might be wrong’ is useful. Where there is enormous corroboration of a theory it makes sense to place it at the core of scientific knowledge.
But doesn't this beg the question of how to determine the likelihood "that core knowledge might be overthrown "? I mean, epistemologically speaking, we don't want any classification of astronomical knowledge - into a hierarchy, or a continuum, or a series of discrete levels - to depend on the unknown unknown of such a likelihood, do we?
It depends on the weight of opinion regarding whether a specific claim is open to doubt. Classification of claims as established, uncertain, dubious, false, etc should not include anything as established that is open to serious doubt. The gaps in model consistency for cosmology make the borders between these epistemic categories uncertain at some points. It should be possible to tabulate knowledge by level of confidence in a way that would discount the reliability of the subjective factors you mentioned above.

I think we need to be careful to distinguish among the many things that "consistency" might be applied to. In this case, "model consistency" is much, much, much narrower than "consistency with the laws of physics" (whether they are the same, universally, or not), to take just one example. Would you care to say more about this? Specifically, what sorts of consistency are you using - as a basis for classifying astronomical knowledge, epistemologically, into hierarchies, continua, levels, etc?
My view on this tough question, for what it is worth, just starts with the simple observation that we have one universe and that two contradictory claims cannot both be true. All true claims are mutually consistent. I don’t understand how ‘model consistency’ differs from ‘consistency with the laws of physics’. Surely the models developed in cosmology are an effort to explain how the universe obeys consistent physical laws? Where a claim has any inconsistency whatever with another well attested claim, doubt about its truth enters the equation.

Is this [Only those areas that can be analysed by methods that will show consistency with existing knowledge,] self-referential? Or are you referring to distinct classes of knowledge, results from lab-based experiments, say, vs pure astronomical observations?
What I was getting at was just that science is cumulative, that there is an enormous body of consistent scientific knowledge that provides the starting point for any new research.

I'm sorry, what does this [or reveal a material inconsistency,] mean?
I was thinking as an example that the observation of the Mercury perihelion revealed a material inconsistency in Newtonian mechanics.

[scientists see things they can't explain, like galactic rotation speed, and so have to imagine possible physics in order to work out how observation can be consistent with theory] seems to me, a pretty darn good definition of model consistency (in just a few words!) :)
So to repeat, in terms of your earlier question, and again revealing my ignorance, I’m still unsure about the difference between ‘model consistency’ and ‘consistency with the laws of physics’. Laws of physics enable model consistency.

... so far, from this thread, it seems to me that model consistency is lower in the continuum (or hierarchy, or levels, or as a dimension, or ...) than pure observations, and also lower than the fundamental principle of the universality of the laws of physics. However, wrt what others have posted, this cannot be a unique classifier ... because then CDM models would rank equally, epistemologically, with models of the distance and radius of a star (per my earlier example), and almost no one seems to accept that.Cold Dark Matter postulates unseen theoretical entities in order to save the model consistency of cosmology. Unobserved ‘gapfillers’ obviously have a lower epistemic status than things we can see. Observational support for CDM (http://astronomyonline.org/Cosmology/DarkMatterProject.asp) seems to make it more than a ‘gapfiller’.

Jerry
2010-Aug-15, 08:34 PM
Returning to this earlier set of comments from *122 ( http://www.bautforum.com/showthread.php/105908-Astronomy-what-can-we-know?p=1770668#post1770668), Nereid asks first how model consistency supports knowledge of distance in astronomy. My intuition is to say ‘very strongly’ while deferring to the experts.
Agreed, with certain caveates. Every part of the distance ladder relies upon the accuracy of the lower rungs. When we find errors in parallax determinations, this impacts the absolute scaling of cepheids, surface brightness, T-F relationships and so-ons. With increasing distances, secondary and tertiary calibrating parameters, such as metallicity, evolution and dust extinction play greater and greater roles. Finally, in redshifted space, absolute velocities, gravitational lensing and other relativistic corrections add to the uncertainty.


‘Confidence’ here is more a reference to consensus views. Established objective knowledge has higher epistemic status than personal beliefs. The subjective factors you mention are more about beliefs that have not been broadly accepted through peer review and empirical confirmation. It depends if you accept a conservative or an adventurous view on the reliability of scientific confidence. Knowledge of stars has both a well-established core, and an adventurous frontier.
I think we should be vary wary of 'consensus confidence' unless the consensus arrives from a 'triangulation' of methods that are as independent as possible. By this I mean Papers such as http://arxiv.org/abs/1008.1265, where the authors use parallax and knowledge of the H-R relationship in harmony with broad spectral observations to better characterized both the dust extinction, star distribution and H-R diagrams of the Orion Nebula Cluster.


I don’t understand how ‘model consistency’ differs from ‘consistency with the laws of physics’. Surely the models developed in cosmology are an effort to explain how the universe obeys consistent physical laws?
There are two differences: Astrophysical models always require a few assumptions - normal dust compostion and reddening, for example. If the reddening law is different in a different place because of grain size or composition, the model assumption can be wrong, but not the basic physics.

The second case is where auxillary hypothesis and no strong local analogs are required to complete the model. Models using Dark Matter, Inflation, Dark Energy and population III stars lack fundamental consistancy with first principles. In these instances, a 'consensus of experts' shouldn't add weight to the acceptability of the models. No one knows what lies under muddy water.