Are theorems inventions or discoveries?
--
Best wishes, Tom.
Veni Vidi Bibi
tom_c...@msn.com
Most likely both, same as the "Is it math or science?" thread.
0+0=01 is a theorem in base-2 arithmetic with the meanings of
"0" and "1" swapped.
---
Jim
You invent a conjecture and then discover whether it is a theorem.
--
Mark Folsom, P.E.
Consulting Mechanical Engineer
http://www.redshift.com/~folsom
Some mathematicians have made a big deal over this one.
The general concensus is they are discovered.
For example, When Edison ______ the light bulb, it is considered that he
invented it because no light buld existed before.
Mathematical truths, however, were always true, even if they werend
discoverd yet.
Columbus discovered America, he didn't invent it
Shelly
They are discoveries. Once a set of axioms is set up, they imply all
the theorems that can be derived from them. The theorems are sitting
there waiting to be discovered.
However, a theory, as in physics, is a speculation which has yet to be
proven and is an invention I would think. But what if it turns out to
be absolutely and forever true: then wouldn't it be a discovery?
Charles
Well, that's where the whole invent/discovery issue gets really
tangled. Columbus did not discover America. He invented the story
that he discovered America.
Actually Eric the Red and the rest of the Vikings wanted to keep it a
secret.
--
***************************************************************
The true test of genius is the ability to see the follies of
one's own times. The ability to change one's own times is the
true test of a leader. The ability to do both is the true test
of a visionary who will never be elected. --Marilyn vos Savant
***************************************************************
And you certainly can not prove the Theorem of Pythagorous so its
"truth" was discovered by a lot of persistence and empirical evidence
(the 3:4:5 as an initial observation).
Charlie
> > > Columbus discovered America, he didn't invent it
> >
> > Well, that's where the whole invent/discovery issue gets really
> > tangled. Columbus did not discover America. He invented the story
> > that he discovered America.
>
> Actually Eric the Red and the rest of the Vikings wanted to keep it a
> secret.
They didn't keep any secrets, though. They went back to
Scandinavia with the invention "We discovered new lands
far away, too far from home though."
---
Jim
: Are theorems inventions or discoveries?
Disoveries. An invention is something we create, something that did not
need to be just the way we made it, something that did not exist prior to
our creating it. It's existence is somehow contingent on our existence and
creativity.
But a theorem is like a mountain we've never been to before. It was there,
we find it, and we (or someone) climbs (proves) it. FLT will be
(or has been) discovered by other cvilizations at the other end of the
universe. Toasters and sweat socks are a bit more contingent.
-Dave Davis
And where they landed in the new world somebody else called it "Canada"
which I understand is Anglicized from the Spanish "Ca nada" which
loosely means "Nothing Here." Maybe that is why the Vikings didn't
stay? <grins>
> >
> > ---
> > Jim
>
> And where they landed in the new world somebody else called it
"Canada"
> which I understand is Anglicized from the Spanish "Ca nada" which
> loosely means "Nothing Here." Maybe that is why the Vikings didn't
> stay? <grins>
They also called Greenland "green" for some reason. They
may have just been looking for a nice place that also
had a good name.
---
Jim
Any invention that attempts to fulfill a practical function in a way
contrary to the laws of nature fails to work and is therefore generally (it
might do something other than it was intended for) useless. There are thus
workable solutions to problems and unworkable ones. And, the viability of
various practical problem solutions is predictable (given good enough
science and analysis) before they are built or even fully designed. The
"truth" of the workability of an invention exists before any embodiment of
the invention exists.
>But a theorem is like a mountain we've never been to before. It was there,
>we find it, and we (or someone) climbs (proves) it. FLT will be
>(or has been) discovered by other cvilizations at the other end of the
>universe. Toasters and sweat socks are a bit more contingent.
>
There easily could be other civilizations that have not "discovered" FLT.
There could also be other civilzations that lack any knowledge of the
calculus. Also, assuming that they have anything to put in one, they could
build a toaster and if they built it as we do, it or something like it would
likely work for them as well, while utter failures as toasters would
likewise fail for them.
For something to be proven as a theorem, someone must first make a
conjecture or ask a well defined question and then someone must seek the
answer to the question or seek proof or disproof of the conjecture. The
question or conjecture is a kind of conceptual invention. No invention, no
theorem. Similarly, an inventor makes a conjecture that some well defined
means will achieve some desired end. Someone with gumption and wherewithal
will (or more often, will not) then contrive such means and whether they do
in fact achieve that end. Of course, if it doesn't work, it's still an
invention and it may even be described in a patent that can be used to
collect dust and decorate one's resume.
Conjecture and proof operate in a similar way in mathematics as in the
process of practical invention.
> Jim Hunter wrote:
> >
Platonists discover. solipsists invent.
To explain: The base postulates of a numeric or geometric system are
invented by those who create the system. There is no requirement that
these postulates have any relationship to reality or "truth." Euclid
postulated that parallel lines in a plane will never meet, which is
arbitrary in the sense that it only aplies to Cartesian coordinate systems
where there is no mass and time is frozen. (I can postulate that parallel
lines always meet, and you can postulate that they sometimes meet. All
of this is arbitrary hence we are inventing things.)
Theorems, since they are proved in the context of a given system defined
by it's arbitrary postulates, are derivations about that system.
Pythagoras' theorem is simply a statement about right triangles in
Euclid's system, but it is not a discovery in the literal sense since you
can derive this while sitting at your computer without ever seeing a right
triangle. More to the point, a computer with no concept of nature or
truth or right triangles could arrive at this theorem after you've
programmed in Euclid's postulates - the computer hasn't "discovered"
anything.
Discovery comes in to play when physicists measured existing right
triangles and found that Pythagoras was correct. Physicists later
discovered that light from far-away stars was bending as it headed towards
us through warped space time, and thereby found that Pythagoras was not
always correct. Both discovery and invention are infusions of new
information, while theorems are useful restatements of existing knowledge.
Regards,
Reuben Logsdon
I guess my errant and truant teenage daughter is a "solipsist" because
she invents a lot of excuses. (I get all tongle tanged when I try to
phoneticize "solipsist" Guess I will need to consult yonder lexicon.)
Hmm, makes one wonder if they were a bit tipsey from too much grog?
It's still unresolved though, whether
platonists discovered solipsists, or
solipsists invented platonists.
Just to add a little coal to the fire...
Clearly a conjecture is an invention (if you disagree, please provide basis
for disagreement).
Once a conjecture has been proved it becomes a theorem.
Can something be invented before it is discovered? How?
--
FLT doesn't appear to have been "derived" in any such way. The four color
map theorem also didn't just fall out from some set of postulates. I also
doubt very much that when Pythagoras' theorem was first accepted as truth
that it was done while sitting at a computer. The computer can also
generate horrendous numbers of absolutely trivial results. In any case,
what one can do now has nothing to do with whether the first knowledge of
something constituted a discovery. Does the fact that I can fly to Hawaii
or look at its image on a map mean that it was never discovered? You can
also sail over there in your sloop and just happen upon it by chance without
thinking about it.
>Discovery comes in to play when physicists measured existing right
>triangles and found that Pythagoras was correct. Physicists later
>discovered that light from far-away stars was bending as it headed towards
>us through warped space time, and thereby found that Pythagoras was not
>always correct. Both discovery and invention are infusions of new
>information, while theorems are useful restatements of existing knowledge.
>
I have patents on inventions that are simple applications of the laws of
physics, and the behavior of those inventions was predictable based on known
information. A great deal of invention now is applied science (even applied
mathematics); it could hardly be otherwise.
Please derive for me a new non-trivial theorem in mathematics, noting each
step of the process as you go, so that we can all see that there is no
invention or discovery involved.
As a platonist, I have resolved the matter to my own satisfaction!
As it is often phrased in sci.math, that
doesn't REALLY resolve the matter.
Can't a conjecture be a 'conjectured' discovery e.g. I conjecture that if I
keep sailing West then rather than fall of the end of the world I will reach
India.
Martin McGranaghan
nen...@globalnet.co.uk
If I were a solipsist, it would.
By going through the process of proof (or disproof) you *discover* whether
the conjecture is (or never will be) a theorem.
Not so fast (or so clear). A mathematical conjecture is a discovery (if
correct) and will be an understood discovery if proven.
On the other hand, an incorrect conjecture is a fantasy, and can easily
be indentified by it's remarkable likeness to the last time I tried to
get original taking a test in Abstract Algebra!
Shelly
> Well, that's where the whole invent/discovery issue gets really
> tangled. Columbus did not discover America. He invented the story
> that he discovered America.
Youche! But you get the idea
Shelly
Let me try again...
Sex was discovered, while the definition of sex was invented.
I hate to see how long this thread gets. (in a math group, no less)
Shelly
:)
Can't a conjecture be a 'conjectured' discovery e.g. I conjecture that
if I
keep sailing West then rather than fall of the end of the world I will
reach
India.
I think I understand what you are saying, although your wording is rather
confusing: a conjecture can be invented and this could lead to the discovery
of a theorem.
I have a problem with this: Fermat (almost certainly) invented his last
'theorem'. Wiles theorem has the same wording as Fermat's Last Theorem,
i.e. x^n + y^n # z^n, n>2. It is, perhaps, not the theorem which has been
discovered but the proof of the theorem.
I propose a point of view on that. I do not say that it *is* my point
of view... it is just to "add a little coal to the fire"...
When we invent something, really, our invention is a discovery. For
example, I invent a program in C++ who does something. The program is
n bytes in size. Do I really *invent* that, or can we think I just
have chosen a sequence from the 8^n possible .cpp files of that size?
They potentially exist before my writing of the program.
Well, let's apply that concept on math inventions/discoveries. Wiles
..ed the proof for FLT, and we can say that he invented it, but
simoultaneously the proof already existed in the set of all possible
ways to prove FLT (being them ranging from 1 to infinite).
When I invent a picture, for example Munch's The Cry, I "invent" it,
right?
But it is just a combination of a huge set of very small pixel, each
having a possible (and enormous) range of possible colors: then it
potentially already esisted, but was not a material thing.
Math facts are never a "material" thing: but instead, they become from
unknown to man to known. It is the corresponding passage to the
creation of the picture. We can both say that a theorem proof is
invented and discovered.
Tell me whadduya think of this idea (altough it is a bit bypassing the
point of the topic...)
Bye - Stefano
There are exceptions to the granting of patents for inventions that seem
impossible, and Patent Offices have shifted the burden of proof to the
inventors.
>
> Conjecture and proof operate in a similar way in mathematics as in the
> process of practical invention.
>
If the above-mentiond exceptions are noted.
Malome.
>Hi All,
>
>Just to add a little coal to the fire...
>
>Clearly a conjecture is an invention (if you disagree, please provide basis
>for disagreement).
>
>Once a conjecture has been proved it becomes a theorem.
>
>Can something be invented before it is discovered? How?
Tom,
Your subject line is a bit inflammatory so I thought I'd take
it on, since I was one of those who gave a response earlier. I said a
theorem is always a discovery and not an invention and I think, from
what you say above, that if we disagree, it's only semantics. First I
have to "conjure" up a possible theorem if it's to be a new statement
at all. After that I go about proving or disproving it. If it's
correct, it's a theorem, else it was just a guess which proved wrong.
In any case, it is not an invention since the axioms on which it is
based were already there which implied the theorem which was also
already there, just not found until I stated it.
As to your question, the answer is "Of course". No one discovered a
working light bulb before Edison invented one. In a more mathematical
sense, no one discovered the axioms of Euclidean Geometry until good
old Euclid wrote them down on a piece of paper; invented them if you
will. Other geometries were similarly invented with different axioms.
Charles
>Hi Martin,
snip
>I have a problem with this: Fermat (almost certainly) invented his last
>'theorem'. Wiles theorem has the same wording as Fermat's Last Theorem,
>i.e. x^n + y^n # z^n, n>2. It is, perhaps, not the theorem which has been
>discovered but the proof of the theorem.
I have a problem here also. Don't theorems have to be true, for sure?
Fermat's "Theorem" has never been proved except by a proof that even
Fermat may have made a mistake in. (To get it almost into the margin
of a page would be a near miracle considering all the mathematicians
would have failed to find it!) Would theory be a better word for it?
Thus we can still correctly say Fermat invented the thing rather than
discovering it.
Charles
<<They also called Greenland "green" for some reason. They may have just
been looking for a nice place that also had a good name.>>
The south coast of Greenland in the summer is indeed very green.
>>No one discovered aworking light bulb before Edison invented one.<<
I would shorten this comment to: 'No one discovered a working light bulb'.
As The Light Bulb was invented not discovered, no one can ever claim to have
discovered The Light Bulb. Of course, you can discover a light bulb at the
back of your cupboard; but discovering The Light Bulb and discovering a
light bulb are different things.
> Hi All,
>
> Just to add a little coal to the fire...
>
> Clearly a conjecture is an invention (if you disagree, please provide basis
> for disagreement).
>
> Once a conjecture has been proved it becomes a theorem.
>
> Can something be invented before it is discovered? How?
Gee, I wonder...but then, nobody has discovered pink, singing unicorns, even
though I can conjecture that there is a herd living in the steps of Mongolia!
:)
Matthias Weiss.
- Errata Corrige -
Excuse me, I am a bit distract... the number of possible combinations
(considering also non-C++ combinations of bytes) is 256^n (including
null characters). Thanx to Lee for letting me note that... anyway the
point does not change...
Lee, even I love precision, I just made a little mistake :-)
Bye - Stefano
I would not be so certain about that. However unlikely it seems to be,
Fermat may have seen some method of proving said theorem that has not been
tried since. I agree it is unlikely, but it is possible. My formal math
education stopped at Partial Diff. Eq., 8 years ago so I'm no math god, but
from what I remember, as theorems become more complex, multiple proofs
become more likely. We now have ^one^ proof of Fermat's last theorem,
perhaps someone will find another...
I do not know how much of an intuitive mathematician Fermat was, but if he
was one of the more incredible math/music geniuses of that era, I would
actually think it was likely... I think it was Mozart who could write
symphony orchestras and operas without ever having to hear any of the
instruments or voices. A similarly gifted intuitive mathematician might be
capable of equivalent extraordinary feats of mathematical reasoning.
Matthew H. Burch
ap...@pipeline.com
}
} Gerald Campbell wrote in message <6s9akr$6or$1...@news-1.news.gte.net>...
} >
} >
} >But I have heard that almost certainly Fermat lied about his "proof",
} >because mathematicians have built the math tools needed for it only in
} >this century... probably Fermat just tried to prove it with some Ns
} >and then extended the statement for each N..
}
} I would not be so certain about that. However unlikely it seems to be,
} Fermat may have seen some method of proving said theorem that has not been
} tried since.
I think there's fairly general agreement that this was the conjecture [that
Fermat had a flash of insight that some technique he had been playing with
would apply to his 'Last Conjecture"], and also fairly general agreement
that he [Fermat] was mistaken. They've been over all of his notebooks and
such and there are definitely notes about techniques that -might- have been
appropriate for this theorem, but the fact is that they aren't up to the
task.
So: if the 'hidden trick' is actually true, then Fermat managed to come up
with an idea that he didn't write down and didn't use for -anything- else.
More likely, I think, is just to admit that he was in error in guessing
that one or another of his 'tricks' would actually serve to prove the "last
conjecture".
[contrast that, for example, with Heywood's approach to the 4-color-problem
[now "theorem", of course]. I could easily see someone believing with a
flash of insight that Heywood's approach would surely prove the theorem and
then happening not to return to it to do the deep and difficult analysis to
discover that it doesn't actually work, and so leaving us with an enigmatic
"wonderful proof" of the 4-color-conjecture. Of course, in the end
Heywood's approach was the one that worked, only it took a humongous
computer analysis to beat down all of the loose ends]]
} I do not know how much of an intuitive mathematician Fermat was, but if he
} was one of the more incredible math/music geniuses of that era, I would
} actually think it was likely... I think it was Mozart who could write
} symphony orchestras and operas without ever having to hear any of the
} instruments or voices. A similarly gifted intuitive mathematician might be
} capable of equivalent extraordinary feats of mathematical reasoning.
this is all true, and Fermat certainly qualified as a gifted intuitive
mathematician. But as I say: if he did have this flash of mathematical
reasoning, it used approaches and techniques that he never mentioned in any
of his other writings and wasn't of use in -any- other of his
investigations, since all of that stuff has been analyzed and doesn't do
the job. We'll never know for sure, of course, but the current general
opinion is that Fermat "guessed wrong" on this one...
/Bernie\
--
Bernie Cosell Fantasy Farm Fibers
ber...@fantasyfarm.com Pearisburg, VA
--> Too many people, too few sheep <--
:( sad face
If a mathematical theorem is true, can't it always be proved? I don't
see how to do it with the Pythagorean Theorem, but isn't there a way?
Charles
Justin Moston <J.Mo...@amtp.cam.ac.uk> wrote in article
<6sh3rq$kgs$3...@pegasus.csx.cam.ac.uk>...
>
> > And you certainly can not prove the Theorem of Pythagorous so its
> > "truth" was discovered by a lot of persistence and empirical evidence
> > (the 3:4:5 as an initial observation).
>
> :( sad face
>
I beg to differ. It is actually quite a easy proof and can be found as
Appendix 1 of Simon Singh's book: Fermat's Last Theorem.
Well worth a read.
Thank you. I thought that was the case. Now: can all theorems, known
as true, be proved? Fermat's is the only one which comes to mind and
perhaps we're not 100% sure it is true.
Charles
Jim Hunter wrote:
> They also called Greenland "green" for some reason. They
> may have just been looking for a nice place that also
> had a good name.
>
> ---
> Jim
They called it Greenland because they wanted people to go there. With a
name like Greenland, who could resist?
Derek Williams
Ranked #2 in the nation in mathematics
http://home.earthlink.net/~fireballdw
"Love is like pi - natural, irrational, and very important"
I've heard that too, but I wonder why the same guys named Iceland
Iceland?
Charles
hehe - Anyone know if a lodestone embedded in wood floating in wine would
read True North if it were used on Northern Greenland? - with no world map
in front of me I can't remember if Greenland extends far enough North to be
further North than the magnetic North pole.
Matthew H. Burch
ap...@pipeline.com
>Charles Riggs wrote in message <3606a561....@news.anu.ie>...
ap...@pipeline.com skrev i meddelandet
<6ti6fr$k58$1...@camel19.mindspring.com>...
All theorems known as true have already been proven.
Until there is a proof, it is not a theorem, only a conjecture!
--
Klas
Charles Riggs skrev i meddelandet <35f8ebcb....@news.anu.ie>...
>On 7 Sep 1998 21:02:32 GMT, "Tony&Maggie" <ellio...@clara.net>
>wrote:
A lodestone embedded in wood floating on wine or suspended from a string
will point to MAGNETIC NORTH which is distinctly different from TRUE
NORTH. In the western hemisphere there is zero difference between
magnetic and true north on a wiggley line (more correctly known as an
"isogonic" line) that tracks through the eastern seaboard of the United
States and passes throught the Great Lakes into Canada. Based on this
(I don't have my map with isogonic lines in front of me) but from
anywhere in Greenland MAGNETIC NORTH will be in the general direction of
TRUE WEST. Maybe the northern tip of Greenland is at a more northern
latitude than Magnetic North in which case your lodestone will point
south of west but the predominant direction will still be west.
Charlie
--
***************************************************************
The true test of genius is the ability to see the follies of
one's own times. The ability to change one's own times is the
true test of a leader. The ability to do both is the true test
of a visionary who will never be elected. --Marilyn vos Savant
***************************************************************
Matthew H. Burch
ap...@pipeline.com
hehe this coming from the guy who made a distinction between grid,
geographic, and magnetic North a few months ago... Double Ooops.
Robert Baer wrote in message <36074D...@olywa.net>...
>Klas och Lotta wrote:
>>
>> Yes it would. The north pole is ice-covered ocean.
>> --
>> Klas
>>
>> ap...@pipeline.com skrev i meddelandet
>> <6ti6fr$k58$1...@camel19.mindspring.com>...
>> >hehe - Anyone know if a lodestone embedded in wood floating in wine
would
>> >read True North if it were used on Northern Greenland? - with no world
map
>> >in front of me I can't remember if Greenland extends far enough North to
be
>> >further North than the magnetic North pole.
Please reply to this or e-mail, I am genuinely interested!
Learn something new every day :)
Matthew H. Burch
ap...@pipeline.com
>A lodestone embedded in wood floating on wine or suspended from a string
The line of constant difference between magnetic and true north is
called an "isogonic" line. In our half of the world the zero difference
line passes through Georgia and the Great Lakes.
Isogonic lines are determined from measurements in the field and not
analytically computed. So they take into account any anomolies that
will move the magnetic needle from magnetic north. Depending on what is
happening locally the lines may shift independently of the shifts in the
north magnetic pole. Aviation charts warn of severe magnetic
disturbances around Arkansas--the isogonic lines also deviate
appropriately. They are anything but straight lines.
If you have a pilot friend, ask him to show you the isogonic lines on
his navigation charts. They are found on the WACs, Sectionals, and Low
Altitude Enroute charts.
Matthew H. Burch
ap...@pipeline.com
Charles Revie wrote in message <360AEE...@zianet.com>...