Jihad needs scientists
in [Design]
|
From: Daniel Mandic on 12 Oct 2006 13:47 Eeyore wrote: > The first stored-program electronic computer was British. > > Graham Also the first radar-apparature for defense purposes! Best regards, Daniel Mandic
From: Daniel Mandic on 12 Oct 2006 13:47 Eeyore wrote: > I have GCE Oxford and Cambridge Board 'O levels' in English Language > and English Literature. Wheew, that's pretty. I guess Cambridge Board 'O levels', is the more technic English lessons... Do you know some Liverpool :) (could you understand?) Best Regards, Daniel Mandic
From: John Larkin on 12 Oct 2006 13:50 On Thu, 12 Oct 2006 14:07:00 GMT, <lucasea(a)sbcglobal.net> wrote: > >"John Larkin" <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote in message >news:12fri2961usl6qoiqotcc442va44b4q1i3(a)4ax.com... >> On Thu, 12 Oct 2006 03:47:25 GMT, <lucasea(a)sbcglobal.net> wrote: >> >>> >>>"John Larkin" <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote in >>>message >>>news:fedri21s4toqia7j4a8o2ejko440n73gp9(a)4ax.com... >>>> On Thu, 12 Oct 2006 03:18:02 GMT, <lucasea(a)sbcglobal.net> wrote: >>>> >>>> >>>>>>>F = m1^1.0000000032 * m2^0.999999982 * (g-0.000000000016), >>>> >>>> Ah, gravitational force is independent of distance, as Occam's Razor >>>> predicts. That certainly is a useful simplification. >>> >>>Oops, my mistake. Please answer the same questions as if I had typed it >>>right-- >>> >>>Please explain why Newton chose F = m1 * m2 * g / d^2, when F = >>>m1^1.0000000032 * m2^0.999999999982 * (g-0.00000000000016) / >>>d^2.000000000013 would have fit the available data just as well. And if >>>you're going to put it off to Kepler's law, which is the basis for >>>Newton's >>>law, then why did Kepler get something so simple, like Period = k * >>>radius^1.5? Why didn't he say Period = k * radius^1.50000000000038? >>>After >>>all, if needless complexification is the norm in science, then he >>>shouldn't >>>have come up with anything so simple. >>> >>>Or are you being deliberately obtuse to cling to an untenable position? >> >> Obtuse? I'm not the one who swore there was no "2" in Newton's law of >> gravitation, as you did in the part you snipped. > >I didn't swear, don't put words in my mouth as a strawman. It was a >mistake. I suppose you never admit to making those. > > >> My answer remains, 2 >> is fundamentally correct, and Newton did not determine it >> experimentally. > >There's nothing fundamental about it--it's based on Kepler's observation of >a 2/3 exponent. Anyway, what about all the other numbers? > Simple mental experiments show that m1*m2 works in that form, at least in the non-relativistic case. > >> Occam's Razor is folklore. It is unproven, proves nothing, cannot >> separate good theories from incorrect ones and, if trusted, shuts off >> thought. In mathematics, only formal proof validates a theorem; in >> physics, only experiment. Being simple isn't enough. > >Of course it isn't. However, I'm going beyond the data--for any given set >of data, there are an infinite number of explanations. How do you choose >between them? Of course, one way is to get more data. However, it's not >possible to collect an infinite amount of data to disprove all of those >possible theories. The other way to select between theories is to choose >the simplest, and then look for data to invalidate it. Once you do, then >you choose a more complex theory, and repeat the process. However, >fundamental to all of this, at each stage, is choosing the simplest theory >that fit the data. Ask any experimental scientist, and you'll get laughed >out of the room if you tell them that "Ockham's razor is folklore." It's arguable that testing, for falsity, the simplest theories first is an efficient/lazy way to proceed through the list. But that doesn't make the simplest theory the most likely one to be true, which is what OR claims. "It often happens that the best explanation is much more complicated than the simplest possible explanation because it requires fewer assumptions. In light of this, the popular rephrasing of the razor - "The simplest explanation is the best one" - can lead to a gross oversimplification when the word simple is taken at face value." http://en.wikipedia.org/wiki/Occam's_razor#Science_by_Razor_alone.3F is about right. As an engineer, and out of experience, I tend to be suspect of simple explanations of mysterious behavior, because the ultimate explanations are often of astounding subtlety, like two or three pathological behaviors hiding behind one another, masquerading as a single simple problem. The more complex a system, the lower the probability that *any* one explanation is likely. What I object to is the common invocation of OR to "prove" that some theory must be right, or even is most likely to be right, because it's simple. John
From: T Wake on 12 Oct 2006 14:01 "John Larkin" <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote in message news:uu8ri25d43ms9k07u8ued1dc2rde6fqmru(a)4ax.com... > On Wed, 11 Oct 2006 20:30:49 GMT, <lucasea(a)sbcglobal.net> wrote: > >> >>"Eeyore" <rabbitsfriendsandrelations(a)hotmail.com> wrote in message >>news:452D52B3.95C3D22D(a)hotmail.com... >>> >>> >>> T Wake wrote: >>> >>>> I think you have mixed metaphors by bringing in the real >>>> world engineering solutions - yes there may be a convoluted answer >>>> needed, >>>> but that it is still the simplest, working, solution which is used. >>> >>> Engineers pretty much invariably like the simplest solution ! >> >> >>And I'll wager that even John goes for the simplest solution *that is >>consistent with all the data he has*. > > No. See my post above. Given imperfect data, I'd be inclined towards a > more flexible - more complex - solution to give me wiggle room to > adjust for the unexpected. For example, if I'm not sure about what > noise spectrum a data acquisition board might see, I might use more > opamps, or not use a circuit where a single component value adjusts > both gain and bandwidth, or bandwidth and loop stability. > > Sometimes more complex circuits are more reliable and easier to > understand/tune than simpler ones. The same thing happens in software: > the simplest possible solution to a given algorithmic requirement (and > there must be one) would likely collapse if you tried to change > anything, and is likely unintelligible as well. Clarity and > flexibility are more important than simplicity. I suspect you haven't fully understood the utility of Ockhams Razor. In your example you are still going for the simplest answer given the situation. You are not adding complexity for the sake of it, you are doing it for a purpose. The Razor, while not a "Law of Nature" (not that anything in Science really _is_) still serves a purpose. The fact you do it without thinking is more down to your skill and aptitude at the job in hand. If a total stranger to the field was given two solutions, all other things being equal (i.e. they both have the same level of flexibility) the simpler solution is the better one. You instinctively pick the simpler one so this thought process may not be as obvious for you. >>F = m1^1.0000000032 * m2^0.999999982 * (g-0.000000000016), but he didn't, >>despite the fact that that equation would certainly have given the same >>results, to within the experimental accuracy of the day. He didn't. Why? >>Because of Ockham's razor--there's no need for the added complexity, so >>why >>add it? > > That's not the reason he picked 2. He picked 2 because it's correct. > Because the integral of x is precisely x^2/2 and can't be anything > else. 2 is not a precise number. It has a massive area of uncertainty. The numbers and theories Newton expounded were excellent in their day. This does not mean the Razor does not apply to the thought processes he used.
From: T Wake on 12 Oct 2006 14:09
"John Larkin" <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote in message news:12fri2961usl6qoiqotcc442va44b4q1i3(a)4ax.com... > On Thu, 12 Oct 2006 03:47:25 GMT, <lucasea(a)sbcglobal.net> wrote: > >> >>"John Larkin" <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote in >>message >>news:fedri21s4toqia7j4a8o2ejko440n73gp9(a)4ax.com... >>> On Thu, 12 Oct 2006 03:18:02 GMT, <lucasea(a)sbcglobal.net> wrote: >>> >>> >>>>>>F = m1^1.0000000032 * m2^0.999999982 * (g-0.000000000016), >>> >>> Ah, gravitational force is independent of distance, as Occam's Razor >>> predicts. That certainly is a useful simplification. >> >>Oops, my mistake. Please answer the same questions as if I had typed it >>right-- >> >>Please explain why Newton chose F = m1 * m2 * g / d^2, when F = >>m1^1.0000000032 * m2^0.999999999982 * (g-0.00000000000016) / >>d^2.000000000013 would have fit the available data just as well. And if >>you're going to put it off to Kepler's law, which is the basis for >>Newton's >>law, then why did Kepler get something so simple, like Period = k * >>radius^1.5? Why didn't he say Period = k * radius^1.50000000000038? >>After >>all, if needless complexification is the norm in science, then he >>shouldn't >>have come up with anything so simple. >> >>Or are you being deliberately obtuse to cling to an untenable position? >> >>Eric Lucas >> > > Obtuse? I'm not the one who swore there was no "2" in Newton's law of > gravitation, as you did in the part you snipped. My answer remains, 2 > is fundamentally correct, and Newton did not determine it > experimentally. 2 is still not a precise number. To this day we can not say it is exactly 2 with an infinite amount of zeros after the decimal place. > Occam's Razor is folklore. Far from true. > It is unproven, Science falsifies rather than proves. Do you have an example where the simplistic solution (all things being equal) is worse than a more complex one? >proves nothing, cannot > separate good theories from incorrect ones Yes it can. > and, if trusted, shuts off > thought. Far from true. It is trusted and has led to _lots_ of new thought. Pretty much all of twentieth century science uses it. > In mathematics, only formal proof validates a theorem; in > physics, only experiment. Being simple isn't enough. No, it isn't enough. You misunderstand what Ockhams Razor is used for. If two theories make an experimental prediction, which matches the experimental results, the accepted manner of differentiating is by the simplest one being "better." This is what is talked about when scientists refer to a theory being "elegant." Remember the aim of experimentation is to falsify a theory. Ockhams Razor assists with this. |