By coincidence, Not Exactly Rocket Science linked today to an article about a new idea on the subject: electrons are unusual elementary particles in that they have a consistent spin. Electrons that are spit out in various reactions, such as beta decay (in which a neutron leaves a nucleus and decays into a proton, a neutrino, and an electron, previously bound together by the nuclear "weak force"), appear to exert a consistent, predictable twisting action on some biological molecules.
The researchers found that left-handed bromocamphor was just slightly more likely to react with right-handed electrons than with left-handed ones. The converse was true when they used right-handed bromocamphor molecules. At the lowest energies, the direction of the preference flipped, causing an opposite asymmetry.
In all cases the asymmetry was tiny, but consistent, like flipping a not-quite-fair coin. “The scale of the asymmetry is as though we flip 20,000 coins again and again, and on average, 10,003 of them land on heads while 9,997 land on tails,” says Dreiling.Over evolutionary periods of time, even tiny assymmetries can add up.
Yes, that's right. And it's why we observe things in physics, at medium scales, that function in a 'lawlike' way -- so much so that during the Early Modern period they were thought of as iron laws of universal force.
ReplyDeleteWell, it turns out that's not right: it's just that the 'medium scale' stands atop nigh-infinite smaller transactions that we can't observe. What we're calling laws of nature are really just the most probable outcomes of much less-lawlike processes.
What we're calling laws of nature are really just the most probable outcomes of much less-lawlike processes.
ReplyDeleteIf the imbalance is consistently extant, there's a law governing it; we just can't, yet, articulate it, including the possibility that we've misstated our currently "known" law of nature.
Eric Hines
The article seems a bit confused towards the end, where it starts asking where there would be sources of polarized electrons. The molecule itself, by reacting with one helicity more than another, will change the average polarization of the electrons hitting it.
ReplyDeleteElectrons (and quarks and neutrinos and any other spin-1/2 elementary particles) have two pure helicity states--typically an unmeasured electron will be in a linear combination of the two states, though. Throw such a random assortment of electrons at a pile of this stuff, though, and the survivors will tend to be more often of one helicity. That's a very cool discovery, and there are probably other examples somewhere.
Spoken like a Kantian, Mr. Hines!
ReplyDeleteSince the concept of a causality carries with it that of laws in accordance with which must be posited, through that which we call a cause, something else, namely its result; therefore freedom, even though it is not a quality of the will in accordance with natural laws, is not for this reason lawless, but rather it has to be a causality in accordance with unchangeable laws, but of a particular kind; for otherwise a free will would be an impossibility. Natural necessity was a heteronomy of efficient causes; for every effect was possible only in accordance with the law that something else determined the efficient cause to causality; what else, then, could the freedom of the will be, except autonomy, i.e., the quality of the will of being a law to itself? (Ak: 4:446-7)
On this view, everything must be determined by a law: the only question is whether the law is imposed by something acting on something else, or whether it is possible for the law to come from within.
But if the law is only the law of probability, the events aren't really lawlike in this way at all: smallest-scale events are undetermined, and thus could happen either way. Even at the level of electrons, we're seeing the playing-out of probabilities happening at smaller scales, which don't show any evidence of being lawlike, but instead probabilistic irreducibly.
James:
ReplyDeleteI tend to think that what we call 'laws of nature' are really three different things. There are laws that come from the nature of a thing ("an electron will have a negative charge"). There are laws that come from the interaction of multiple things (like the electron/molecule example). And then there may be some laws that are general -- perhaps the law of conservation is one of these.
But what do we say, then, about the undetermined behaviors? Well, they have something to do with the nature of the thing; the range of potentials is limited by it. But the law can't be a determining law, if there is irreducible probabilistic behavior (as there seems to be).
I'm not sure we're in a position to know whether probabilistic behavior is irreducible, only that it's not reduced at the moment. You have to wonder at any process that produces a predictable probabilistic distribution rather than true randomness: is there really not an underlying mechanism we can't see?
ReplyDeleteBut I have no reason to think the question will be answered any time soon. Einstein and Bohr argued about it for their whole professional lives. Bohr thought it was a waste of time to try to "peek under the veil." He may have been right, but the jury is still out. So far all the clever explanations for what's under the veil haven't come to much. All we really know is that the only tools we have at present are probabilistic, and they work pretty well.
it's just that the 'medium scale' stands atop nigh-infinite smaller transactions that we can't observe. What we're calling laws of nature are really just the most probable outcomes of much less-lawlike processes.
ReplyDeleteSounds like crazy humans making up together rational social and national decisions, as well as economic policies.
I'm not sure we're in a position to know whether probabilistic behavior is irreducible, only that it's not reduced at the moment. You have to wonder at any process that produces a predictable probabilistic distribution rather than true randomness: is there really not an underlying mechanism we can't see?
ReplyDeleteIt's been a general wonder of mine for decades.
We observe it at every level, though, including our own. Traffic patterns are in a very real sense a set of free choices: you could go home a different way, and maybe you often do go one of several ways depending on whim. But somehow millions of individual choices turn into a highly predictable system if you raise the scale of observation -- say by putting traffic cameras everywhere, and graphic volume over time.
We get -- I think -- the same kind of false sense of 'lawlike' behavior when we talk about economics. By abstracting the scale, we end up seeing some things that look very much like governing laws. That gives us the sense that we know the rules, and can manipulate the system.
But we really can't, because what is really there isn't a "system" with "laws," but a bunch of free choices being made at the smallest level by independent actors. Attempting to treat all that as if it were a system with laws leads invariably to unexpected results, not to the predicted outcomes: and we see this all the time, do we not?
So yes, in a way it's proper to say that we can't know yet, or don't know yet; but in another way, we can say that we have reason to think it may be so. The analogy between smallest-level physics and independent choices by human beings is controversial, but if we do in fact have free will it is proper: then we, like the probabilistic functions, are really not determined by an exterior law. We may be influenced, but ultimately the choice is ours.
Thus we too (if the probabilistic models are right, and if there is free will, both of which people fight about) are undetermined; and yet if you abstract a bit from the free choices, you will similarly 'see' a lawlike system that seems for all the world to be deterministic and governed.
By the one, one objection I hear to that analogy is that we do, in fact, have the ability to make predictive judgments about physics in a way that we don't about economics.
ReplyDeleteBut we aren't trying to manipulate the "system" in physics -- we just want to predict it, at most to interact with it on its own terms effectively. In physics proper, we want to be able to predict what will happen; in engineering, we want to be able to insert something into the system in such a way that we get a desired outcome. But we don't expect to change the system.
In economics, some of the lesser predictability comes from the fact that we are much closer to the 'free' actions -- there aren't that many levels between the free individual choices and the 'system' we want to predict, whereas between us and quanta are very many smoothing interactions. But I think much of our problem is that we want to change the way the system works, as by Keynesian pump-priming, rather than just interacting with it.
Does that mean Kant is in good company? [g]
ReplyDeletewhich don't show any evidence of being lawlike, but instead probabilistic irreducibly.
This assumes laws must be deterministic. This hasn't been...determined. It's a perfectly fine law that holds that the Universe doesn't care about outcomes at an irreducible level of individuality, but at some aggregation (achieved smoothly--but of necessity "smooth" is itself an aggregation--or via some step function, or chaotically) there must be a particular outcome.
But we really can't, because what is really there isn't a "system" with "laws," but a bunch of free choices being made at the smallest level by independent actors.
Yet we do this all the time. Despite the probabilistic nature of the "choices" of quanta, we successfully shoot the moon, and the planets of our solar system--at a level of abstraction, the matter becomes a highly predictable engineering problem.
And so it is in economics: given individual free choice (even stipulating that free choice is driven by rationally acted on self interest, with that definition of self interest itself highly variable), at a level of abstraction, we can identify a law relating price to the relationship between supply and demand. That there remain problems in acting on this indicates one or more of a number of as-yet unsolved problems, not the inapplicability of some law: we've not abstracted to the correct level, we don't yet understand sufficiently free choice or self interest, we don't yet understand sufficiently the relationship between price and the interaction of supply and demand, we haven't measured accurately supply or demand, etc.
That measurement problem, incidentally, goes to a problem T99 raised earlier: universal constants at one time suggested a law of motion/gravity/mass that held the Universe would expand to a point and then begin collapsing in on itself. New (I don't say better) measurements of those constants now suggest a law of motion/gravity/mass that has the Universe expanding until it dies a heat death, or time stops. We don't know the law; that doesn't mean the law doesn't exist.
Eric Hines
But I think much of our problem is that we want to change the way the system works....
ReplyDeleteI think it's more an effort to interact with the system of economics through laws we don't yet adequately understand: your Keynesian example, for instance. In economics, and in systems generally that involve human behavior, we're still at the orrery stage of navigation--if we've come that far.
Eric Hines
That's possible, but it's a very different view from mine. If you're right, some adequately powerful computer might be able to completely manipulate human beings at some point in the future, such that whole economic systems would be manipulable like puppets.
ReplyDeleteIf I'm right, the real problem is that we're too close to the free choice for the 'laws' to be smooth.
Also, the laws don't really exist. The things exist. The illusion of laws is produced by the properties of the things, which (because they have those properties) interact in ways that -- over time -- create lawlike effects. That long interaction of things like electrons and molecules is what explains regularities like the one Tex's OP was talking about. That's the smoothing process.
Since we don't have a long enough chain in between humans and economics, then, we can't make reliable predictions. It's like the coin toss example: if you're right there, you really can't predict if the next one (or three or ten) will be heads or tails. It's only when you've removed to the distance of nigh-infinite coin tosses that the lawlike effect becomes reliable.
It's still not a law, though. It's an abstraction.
If you're right, some adequately powerful computer might be able to completely manipulate human beings....
ReplyDeleteWe already do this--education and propaganda, for instance. We're just not perfect at it (but we're getting pretty good) because you're also right: we're too close to the free choice for the 'laws' to be smooth.
In the context of understanding things well enough to predict outcomes/influence behavior, I think the distinction between abstraction and law is an overcomplexification.
Eric Hines
Well, we can't both be right. :) If I'm right, it's not just that we don't understand well enough or that we don't have adequate processing power. It's that the lawlike effect isn't a law. What that means is that control doesn't flow down from the abstraction to the genuinely free event: I can't 'manipulate the laws of the economic system' and control the behavior of the people.
ReplyDeleteI can offer them incentives to do what I want. I can try to force them to do what I want. But the thing that looks like a law ("Supply and Demand," say) is not really a governing law. You can no more force me to behave as you wish by manipulating economic laws than you can force electrons to jump one way or the other by changing the laws of physics.
So even very strong candidates for laws of nature -- gravity, say, which is as strong a candidate as 'supply and demand' -- may turn out to be merely lawlike effects of smaller-scale interactions that aren't themselves governed by deterministic laws.
I leave open the possibility that there may be lawlike effects that really do have the force of laws, because there is nothing whose nature doesn't include them. This would be things like the law of conservation of energy and matter: if we're right in our construction of it, or even approximately right, it appears to have no exceptions. (To put it in Aristotelian terms, it would be a law of matter; the others arise from forms of organizing matter, or the interactions of such forms.)
It's that the lawlike effect isn't a law. What that means is that control doesn't flow down from the abstraction to the genuinely free event
ReplyDeleteCouple things here. The first is what I was getting at about overcomplexification. If I can make useful predictions/manipulations from it, it's law enough. The second thing is that probabilistic/nondeterministic law. I can't use supply/demand, for instance, to force you in particular to do anything, but I can use that law to get populations to do.
force electrons to jump one way or the other by changing the laws of physics.
This gets at what is a law--whether that even matters--and when there is knowledge. I've never believed that there is anything in the Universe that's beyond man's ability to know. This includes changing those laws of electron behavior or finding a way to cross the light-speed barrier or.... This might begin by cheating in some sense through manipulating probabilities at quantum levels. Could we succeed at any of that, since we're by nature arbitrary, the Universe would be shown to have been arbitrary all along. But then, whatever it is I'm doing to achieve one of those things must be law enough to produce predicted (my arbitrarily desired) results, or I haven't actually done anything reliably. And so the Universe wouldn't actually be arbitrary (except perhaps in an Art Clarke-ian sense to an alien species) at all, but "lawfully" malleable.
Eric Hines
If you're right, some adequately powerful computer might be able to completely manipulate human beings at some point in the future, such that whole economic systems would be manipulable like puppets.
ReplyDeleteThat would be equivalent to God or approaching omniscience. Or a version of Deus ex Machina.
God works in mystery ways, probably because humans were too stupid to understand the manipulation even if it was explained.
You should read Gersonides' Wars of the Lord, Ymar.
ReplyDeleteMr. Hines:
I've never believed that there is anything in the Universe that's beyond man's ability to know.
That's an interesting belief. I can think of two real challenges to it. The first is Kant's distinction (since we were just talking about him) between noumena and phenomena. Kant argues that everything you could know by experience is changed from what it really is by a necessary process of unification in your mind (your knowledge of the apple in your hand is really some mysterious combination of separate sensory impressions of light and feeling, which your brain has unified and further interpreted the light as 'color' and the feeling with the light as 'shape,' in a three-dimensional space that it has constructed around you). As a result, what you can know is only your experiences, which are unique to your mind -- they don't exist outside of it. And the things in themselves, those you can't know at all.
Kant meant for this to be helpful, as he was responding to an even more skeptical problem raised by Hume. It's a pretty thorny problem for the idea that there is any ability to know anything outside the mind, though, let alone everything.
The second is Heisenberg's claim that there is a necessary limit to the ability to know pairs of things -- especially position and velocity -- with equal precision. Roughly, the more precisely you know one, the less precisely you know the other, and so forth.
So if you can precisely know the position of a particle, you can't know it's velocity. In that case, there is something you couldn't know, contra your hypothesis.
But if you can't know the position of a particle precisely, then again there is something you can't know, contra hypothesis.
How do you propose to think about these problems?
Kant, Heisenberg, et al., are simply articulating a form of a law with an inadequate articulation, an incomplete specification, or a wrong one.
ReplyDeleteOr I am.
Kant expressed a position I also argued for some years ago in another thread in the Hall (pre-Blogger "upgrade" that trashed your archives) that what we know is only what we know; I had taken it a step further, insisting that what we know is flawed from imperfect sensors and processors. But flawed isn't unknowable; it's only hard to know today. Flaws are correctable.
So it is with Heisenberg. We can't know both position and velocity--but only within the limits of our current understanding of the Universe. Indeed, in an earlier era, the question wasn't even applicable. We'll find a better understanding tomorrow.
So it is with all the other "unknowables." They're unknowable only within our current level of understanding. I have no reason to believe there's an outer bound to our ability to understand. Kant's, Heisenberg's, et al. arguments are founded on such a boundary existing.
The only way to determine which of us is right on that score is empirically. So far, every claimed or implied limit to knowability has been exceeded. Of course, that doesn't mean it always will be, but....
Eric Hines
Heisenberg's Principle lead inevitably to the whole issue of how people can observe something without interfering with it. Thus it brings to mind the Prime Directive and time travel paradoxes, where merely traveling back in time automatically changes the time stream.
ReplyDeleteBy observing velocity, this causes the placement to be changed. And vice a versa.
Thus to break that principle requires an observation that is entirely free. Free of interference. Passive instruments are not precise, however.
It may be possible, but it's a difficult concept to engineer. Twin quantum particles may help. Or it may not.
Flaws are correctable.
ReplyDeleteKant's argument is a little more dangerous than that. It's not that we have a flawed perception of the world. It's that anything we'd do to correct it would still, of necessity, have to be organized by our minds in order for us to understand it. But in organizing the information, the mind is of course changing it.
So the problem isn't that we don't have good enough technology. It's something more specific about what it takes to have an experience as a conscious being of the sort we are. His technical term for it (in case you want to read more) is "transcendental apperception." But all he means by that is this business of what a mind has to do in order to make sense of the many things presented to it, and to organize those things into an experience of the sort a mind can have.
So it is with Heisenberg.
Not if he's right! :) In any case, a part of the claim is that measuring a particle means interacting with it. That, too, means changing it in some way. To measure a hair I might put it under a microscope and bounce light beams off it (e.g., turn on the light) so that I can see it to measure it. The hair won't move much because of the impact of the light beams. But at the particle level, we're often dealing with things smaller than photons. That's why we need electron microscopes for smaller things. But there are things smaller than electrons. A 'bounce' to determine position is going to change velocity. A 'bounce' that determines velocity is going to change position.
A few little comments:
ReplyDeleteThe branch of physics called statistical mechanics shows how, with a few simple assumptions about a system of particles, you can make predictions about large scale behavior. For example, trying to follow the collisions of every single gas molecule in a jar exceeds any computing system we could make. But with enough molecules involved, you can derive things like the old familiar PV=nRT. It isn't exact, of course, but when you're dealing with such enormous numbers of molecules as you find in a auto engine cylinder, the deviations from ideal are completely negligible. If you only have a few hundred, the deviations start to matter.
Heisenberg's principle seems to be pretty fundamental. It even extends to things you wouldn't expect, like the uncertainty in energy in protons in a nucleus. Because you know the protons are pretty much confined to the nucleus, Heisenberg's rule says their momenta are uncertain in inverse proportion to the nuclear size--and if you probe the nucleus you find that they do in fact possess such a large uncertainty in momenta.
That there should be a minimum angular momentum isn't obviously necessary, but there it is, and it may be built into the structure of spacetime/momentum space. Some researchers are studying what you get if space/time/momentum isn't continuous but grid-like.
If there's something smaller than an electron, I'm not familiar with it. Taking into account the uncertainties associated with probing, it seems to be a point particle. Presumably the muon and tau and neutrino are similarly pointlike, but we haven't been able to study them as thoroughly : the former two decay quickly and we can't get neutrinos to hold still. The neutron, by contrast, isn't pointlike and has a kind of size.
By observing velocity, this causes the placement to be changed. And A 'bounce' to determine position is going to change velocity.
ReplyDeleteNo, the one leaves the other indeterminate, nothing in the other is changed that we can tell. And it doesn't take an active bounce; a passive measure has the same effect. Don't know means don't know. If Heisenberg is right.
Kant's argument is a little more dangerous than that. ... So the problem isn't that we don't have good enough technology.
There's more to it than mere technology, though tech may be involved at a later stage. It's a matter, first, of whether Kant is right at all--a matter of whether he has accurately described the experiential nature. The business of "anything we do to correct" requires organizing, which changes the thing, is minor--the correction itself changes the thing, by design. If the correction is...correct...it will, of necessity, account for the change from the organization.
Yet if Kant is right, then Heisenberg must be wrong: his uncertainty principle is only what his own mind has done to organize his understanding--a purely internalized manipulation of the thing, which has changed the thing away from what he's describing with his principle.
Yet, further, it may be that each has simply misunderstood the underlying principle or law, and so have not articulated a true thing. Which Kant admits he has done (whether he intended it or not) with his claim that that everything you could know by experience is changed from what it really is by a necessary process of unification in your mind and which Heisenberg, and others, have conceded, as well, if more convolutedly with their claim that the act of measuring changes the thing: in both cases, the claim is that to know is to change and so to undo or prevent the knowing.
That's my rub. The change--if it occurs; this is, in fact, simply another estimate of ours concerning the nature of the Universe, limited by our currently limited understanding--is predictable, and so the knowing can be left intact. It just has to be led, as it were, by the nature of the change.
Eric Hines
If there's something smaller than an electron, I'm not familiar with it.
ReplyDeleteNor is anyone, as far as I know. But since we're talking about an area we have no access to directly -- the electron microscope only works on things that electrons can bump -- we aren't sure what's down at the Planck area. It's theoretical at most below 10 to negative 35th. People talk about 'quantum foam,' but maybe that's just a concept.
...a passive measure...
A passive measure still requires something to encounter something else. Otherwise, there's nothing to measure! What we can measure are events.
It's a matter, first, of whether Kant is right at all--a matter of whether he has accurately described the experiential nature.
That's the right place to start, yes.
If the correction is...correct...it will, of necessity, account for the change from the organization.
You're asking after knowledge, though. How would you know the correction was correct, even if it were? You'd have to have a standard independent of the experience of your mind against which to judge the experience of your mind. But every piece of data you bring to check the experience of your mind is an experience of your mind. And so too for everyone else you talk to about it, for whom it's all experiences of their mind; etc.
Yet if Kant is right, then Heisenberg must be wrong...
Well, that's not clear either. Maybe Heisenberg is correct in the sense of the last paragraph. But we can't know that he's correct, if Kant's right: we can just know that our experience of his statement does seem to work with our experiences of the data. He might be correct about the universe or not; the problem is that you can't know the noumena at all. We can't only have coherent knowledge of our phenomena.
The change--if it occurs; this is, in fact, simply another estimate of ours concerning the nature of the Universe, limited by our currently limited understanding--is predictable, and so the knowing can be left intact. It just has to be led, as it were, by the nature of the change.
I'm not sure I understand what you mean by this. The change refers to whatever the mind does in transcendental apperception? One part of the problem is that we don't really know what it does -- indeed, that's another thing we can't know, because we can't measure the output against the input, since the output is our only standard for guessing what the input might have been.
James, why do you stress how small the electron is? I mean, does that bear on the helicity issue? Speaking of which, I assume that the "spin" is metaphorical, and not something like the "English" on a billiard ball that might sort of impart a direction to the targeted molecule?
ReplyDeleteYou're asking after knowledge, though. How would you know the correction was correct, even if it were?
ReplyDeleteThat's at the center of it. We don't know at any particular stage of our knowledge/understanding evolution how close we are. All we can do is keep approximating the true aspect until we reach it. Or discover we've been running down a blind path.
Eric Hines
But how do you know you're approximating the truth if you have no access to the truth that isn't filtered by apperception? You've got nothing to check your phenomena against except other phenomena; the best you can hope for is that your phenomena will cohere.
ReplyDeleteJames, why do you stress how small the electron is?
ReplyDeleteI think he's responding to my claim that there are things smaller than electrons. It depends on what you mean by "thing," but if you mean something like a particle, we can't be sure there is. Just what's going on at smaller scales is conceptual/theoretical, in part because the smallest "thing" we know of is already what we use in our best microscopes. How to make a microscope that could 'see' below the scale of the smallest particles?
Well, if there are smaller particles, we might be able to do it. But if there is some basic 'stuff' down there -- something like Aristotelian prime matter -- it may be something we can't really know anything about. (That, by the way, was the Aristotelian conclusion: that prime matter is essentially unintelligible and therefore unknowable.)
how do you know you're approximating the truth if you have no access to the truth that isn't filtered by apperception?
ReplyDeleteHow do I know I'm approximating truth better with an astrolabe than with an orrery? How do I know I'm approximating truth better with interplanetary devices that let me get among the planets than with an astrolabe?
prime matter is essentially unintelligible and therefore unknowable
If the foundation of our knowledge is unknowable, then nothing is knowable. And so Aristotle had no clue of whether he was right.
Eric Hines
How do I know I'm approximating truth better with an astrolabe than with an orrery? How do I know I'm approximating truth better with interplanetary devices that let me get among the planets than with an astrolabe?
ReplyDeleteWell, the way you think you know is that your phenomena are more coherent. But they're still phenomena; you don't know anything still about the noumena producing them.
If the foundation of our knowledge is unknowable, then nothing is knowable.
The foundation of knowledge is form, though. To know a thing is to grasp its form. Form is (literally) perfectly intelligible. The form is the order into which the matter has been put. In grasping that, you can know not only what a thing is, but what it is for -- and you can work out its essence, the part of its form that really determines what it is.
Aristotle himself doesn't talk about prime matter (at least, most scholars don't think he does). But later Aristotelian thinkers believed that -- although prime matter couldn't exist without a form -- it must be possible to talk about it, to separate it conceptually if not actually from the form. So what is this stuff out of which the universe is made? What is it that takes on various forms?
To see the problem, a table may be made out of wood. But wood is a form too: it's a material based on an organic molecular structure of a particular kind. And the molecules are forms of order, too; and the carbon is carbon and not gold because of its atomic structure. The proton is a proton and not a neutron because of its structure and charge, which are forms; and it turns out that, actually, anything we could in principle describe is a kind of form.
So what is being put into these various forms? What is the basic stuff out of which reality is structured?
That's the question, and it may not have an answer. Any answer you give to it is describing its structure, which is another sort of form. And that, by definition, can't be what you're wanting to talk about: you're questing after the prime matter, not the forms. We already know the forms are intelligible. What we want to know is what they are informing.
The foundation of knowledge is form, though.
ReplyDeleteThat's one approximation.
And finally, it may not have an answer.
Of course it has an answer. That we can't--yet--figure it out doesn't mean it doesn't exist.
Eric Hines
The more radical idea is that it may not have an answer at all -- not because there's nothing there, but because the way we could describe it is in terms of form, and the stuff we want to describe is (conceptually) formless. To say what it is is to say what form it takes.
ReplyDeleteWhich, of course, it doesn't: that's why it can't exist without form. But it's interesting that the practical reality of it turns out not to be conceptually separable in a way we can talk about. Language attaches to form; all we can say about this is that it manifestly has a capacity to assume form.
Which, actually, is in line with not only the scientific problem we were discussing, but also the Buddhist conception of the basic stuff of reality. I always find it interesting that such radically different modes of inquiry lead to similar conclusions.
If you don't like Kant, try Wittgenstein. He comes at the problem from the other side, and says that of course we can know whether an astrolabe works: we have something external to our minds to agree upon, against which we can each check our internal impressions. What we can't know, for him, is what is purely internal: our feelings are not knowable. If I say that I am experiencing feeling X, you can't check to see if I'm right; and I can't bring back what I remember as earlier examples of X to check them, either. So I am speaking nonsense, Wittgenstein says: I'm just telling you that I agree with myself about what I am feeling.