Line? Cool. Resolved. Move on.

My AP Government and Politics teacher (awesome dude) likes to reply to my scientific jargon with "The shortest path between two points is a straight line drawn in opposite directions!" I then say "Only in Eucledian Space!" Sometimes I add "In Minkowski Space-Time, a straight line connecting two points makes up the *longest* possible path."

SO. Any of you know any other spaces in which a straight line is actually the longest path between two points - or just simply not the shortest one?

## Comments

Sort by:

TopNewestFirst of all, a "straight line" in any space is presumed to be a geodesic, which is, by mathematical definition, an extremium between two points. Notice I said "extremium", it doesn't necessarily mean the shortest distance, sometimes it can be the greatest distance [even if a straight line path is followed--per Calvin Lin's comments]. I can fly on a great circle route from Los Angeles to New York over America, or I can choose to go the other way and get there just the same. The same thing can happen on pseudospheres as well as regular spheres. When extended indefinitely, geodesics can often eventually intersect themselves, and so you'd have a "short way", and "the long way", even though both paths are on the same extremium "straight line".

See graphic for examples of geodesics on a pseudosphere.

For those interested in differential geometry, geodesics on a surface have the property that at all points of the geodesic curve, the curvature vector of the geodesic curve is the same as the normal vector of the surface at that point, differing only in magnitude. In other words, a plane defined by the tangent of the curve and the normal of the surface at any point contains the circle defining the radius of curvature of the geodesic at that point. – Michael Mendrin · 2 years, 3 months ago

Log in to reply

The definition of a geodesic, is that it is

locallythe shortest distance between 2 points.It also need not be true that "geodesics eventually intersect themselves", even on closed, compact, connected, manifolds. For example, take the torus \( [0,1 ] \times [0, 1 ] \), then any geodesic is of the form \( y \equiv kx \pmod{1} \), and clearly it doesn't intersect itself for irrational \(k \). Since there are uncountably many irrationals vs countably many rationals, it is unlikely for a geodesic to intersect itself.

The torus is also a good example of "infinitely many geodesics between (any) 2 points", which can appear contradictory at the start (so why isn't it contradictory?) – Calvin Lin Staff · 2 years, 3 months ago

Log in to reply

The one place where the wording in my previous post can be tightened up is this sentence, "Notice I said "extremium", it doesn't necessarily mean the shortest distance, sometimes it can be the greatest distance." The weasel word "sometimes" apparently isn't enough to save it from criticism, and so more words have to be added to it, such as, "Notice I said "extremium", it doesn't necessarily mean the shortest distance, sometimes it can be the greatest distance

even if a straight line path is followed". Thiscanbe true in some cases, as with perfect spheres, where a straight line is a great circle.We agree on one point, which bears repeating. While "shortest distance between two points" has meaning both in Euclidean geometry and local differential geometry of curved spaces, it doesn't have useful mathematical meaning to speak of "greatest distance between two points" without qualification.

I'd like to point out that the mathematics of extremium, as in variational calc ulus, plays a significant role in the mathematical apparatuses of physics. In many instances, a thing or a system seems to behave as if it wants to minimize something, such as "taking the shortest path", which is often used to describe which path a ray of light takes. But light does not specifically seek the shortest path, it often ends up taking a roundabout way. Why? Because it is merely following the local conditions which defines a geodesic, which just so happens to be the shortest path in many--

but not all--cases. Which is the point I think John Muradeli is trying to make. Physics is often guided by paradigms such as, "shortest path from A to B", or "least action", or "shortest spacetime interval", etc., but it's important to keep in mind that they are just paradigms, and do not actually indicate "motivation" of any particular system, i.e., a system isn't necessarily trying to do any of those, it is merely following a certain set of local conditions, similar to those defining a geodesic in differential geometry. – Michael Mendrin · 2 years, 3 months agoLog in to reply

So... Really, a list of fabric management systems (4D spacetime is one) with defined shortest paths would be great. But, now I have questions. So, geodesic. First time hearing this word. This stuff:

So, why are we complicating things? Why take a geodesic when you can... do this

Is it because we define those restrictions? I'm not as advanced in non-traditional geometry, but I can give an example - say, if you're passing by a black hole, it may be a good idea to take a curved path from A to B - if between A and B we can draw a line that passes through the event horizon. I'd reason this to be like finding the shortest path between two points located on opposite faces of a rectangular prism - assuming you must travel on the surface. The shortest path won't be a direct straight line over the prism, because when we unfold the shape into its 2D form, we see that when we draw a straight line and fold it back together we see a curved straight line (erm... bent). So is this it? Is this why when working with space under the influence of gravity, you use differential geometry and define new spaces? Lol I'm probably totally off xD

One more ex. So when we have undisturbed space, straight line path is the shortest path (RIGHT?). Now it may APPEAR as if we're taking a shortest path by traveling directly to point B (or, in pic, D), but this may not be so:

We rather take a curve around the spatial disturbance, and thus travel lesser overall distance. Is this it? Am I hitting it? – John Muradeli · 2 years, 3 months ago

Log in to reply

If you are on a surface trying to go from point A to point B, the "real off-the-surface shortest straight line distance from A to B" may have no meaning at all, because there can be a multiplicity of other curved spaces in which that surface is defined, so that the "real off-the-surface shortest straight distance from A to B" can depend on that other curved space, and so can have any range of values, including zero.

It should be noted that wormholes are not examples of "going from point A to point B off-the-surface". Wormhole travel is still very much on the surface.

Again, a technical point should be made here that "flat" Euclidean spaces do have some local properties that are unique among spaces. However, in physics, it's a chicken or the egg problem--is flat Euclidean space a fundamental, or is it emergent, i.e., dependent on a host of conditions that must be present before it can even exist? Topology addresses this question. Conservation laws in physics also critically depend on such things, so that, for instance, we cannot necessarily assume that energy is always conserved. Conditions have to first exist before we can have a law of conservation of energy.

Additional Note--When I was a teenager in the early 1960s, I bought a pocketbook in a local liquor store (that's right, a liquor store, where they sell booze!) of an hoary old 19th century text on non-Euclidean geometry. I think I still have it. It was the reason why I got interested in differential geometry. Today, in America, you can't find any math books at all any more in any "local stores". Indeed, even public libraries today carry far fewer math and physics books than they used to. Times have changed. – Michael Mendrin · 2 years, 3 months ago

Log in to reply

"because there can be a multiplicity of other curved spaces in which that surface is defined, so that the "real shortest straight distance from A to B" can depend on that other curved space, and so can have any range of values, including zero."But why bother, say, drawing a sphere in a vacuum, undisturbed, 3D space, around two points for which you're trying to find the shortest distance between, and analyzing the system from spherical perspective when you can just take the straight line? I mean, it seems to me that all you're doing is ascribing unnecessary definitions to physical properties, when more straightforward models exist. And, I think I came up with the best definition of shortest distance: "Assuming an object travels at a constant speed, for what path will the object reach point B from point A in shortest amount of time?" Now I expect you to say "Well, it depends, because quantum space stuff and your object can teleport to B without having to traverse the space in-between." Is that it? Do you define various spaces to account for possible external influences?

– John Muradeli · 2 years, 3 months agoLog in to reply

I think I can safely say that nobody understands quantum mechanics--Richard Feynman, who won the Nobel Prize for his work in Quantum Field Theory.Addendum: In answer to your question about getting from A to B in the shortest amount of time", it's all about constraints, isn't it? For example, the shortest distance from A to B may be right across the lake, but I'll be taking the long way around if I have to walk. A lot of physics is like that. – Michael Mendrin · 2 years, 3 months ago

Log in to reply

So if we drew the triangle on the ground, each leg say 3ft, it would be a regular triangle. But you're telling me that if we make those legs really long, the angles will be all messed up? So are angles dependent upon leg length relative to the circumference of the sphere, or...? Probably. Well, and I think I now understand somewhat how non-Eucledian geom can be useful. Like, we think we're going in a straight line on a mobius strip, but in reality, not really. Or on Earth. And thus, by creating a model that obeys the same restriction (i.e. I could get to China from USA by digging through Earth, but I probably will take ... a more civilized approach) as the system we're trying to analyze, we can make accurate predictions. Rite? I think so :D Now try and solve the new noobs weekly, if you dare ^.^ – John Muradeli · 2 years, 3 months ago

Log in to reply

Technically speaking, as long as the triangle is on a non-Euclidean patch of surface, it is non-Euclidean, but in the

limiting case, it can betreatedas an Euclidean triangle. As an analogy, mechanics in physics is supposed to be relativistic, but as agood approximation in most ordinary casesclassical Newtonian mechanics can be used to describe it. Physics is full of "okay, it's good enough to use" stuff. Niels Bohr even referred to this with his "correspondence principle:", but he was talking about quantum physics.An interesting trivia is that Kip Thorne recently was a consultant in the making of "Interstellar", which has a lot of weird warped spacetime stuff.

Another trivia which may be a little mind-bending is that John Archibald Wheeler co-authored with Richard Feynman the infamous "Wheeler-Feynman Absorber Theory", which suggests that, among other things, events in the future can affect things in the present. Quantum-mechanically of course. Quantum physics is supposed to be weird, right? – Michael Mendrin · 2 years, 3 months ago

Log in to reply

And I knew this would be the case. It's just like with \(E=mc^2\) - the \(m\) is \(m_0\) divided by \(\sqrt{1-v^2/c^2}\) - but we don't care for that because our \(v\) is never so great on common examples to significantly affect our \(E\). So we're "approximately Eucledian," – John Muradeli · 2 years, 3 months ago

Log in to reply

The book assumes that you have at least undergraduate level of math ability. But the book learning curve is still phenomenally steep.

Anyway, I have to go too. – Michael Mendrin · 2 years, 3 months ago

Log in to reply

– John Muradeli · 2 years, 3 months ago

Lol damn I thought it would cost like 1$ or something - some good old books drop that low. By the way, ever read Penrose's "Road to Reality"? That book is my #1 book to read. But it will probably be the last cause MATHS QUANTUMS!!!Log in to reply

Because it's M.C. Escher's birthday today, I'm gonna throw this in. More warped space stuff. Are you sure you know what a "shortest straight line distance" means in the geometries of Escher's feverish mind?

M C Escher – Michael Mendrin · 2 years, 3 months ago

Log in to reply

– John Muradeli · 2 years, 3 months ago

You see? This is what I'm talking about. This stuff is insane! And imagine learning ALL THAT in a day. With 2045 avatars. But instead imagine trying to learn that manually and have $10,000 in your bank account. PRIORITIES! ("all that" I mean all math and all physics ever). But the link is totally awesome - thanks for that. And you should totally read that Penrose book - I heard he took like 8 years to write it! Alright I'ma let you go now. Thx \(\forall\)Log in to reply

The only such space is the space where the distance between any 2 points is 0.

Note that if we have 2 points \( x, y \) such that \( d(x, y) > 0 \), then we can transverse indefinitely between these 2 points to lengthen our journey, and hence the "straight path between these 2 points" will never be the longest distance.

Furthermore, you can show that if we have 3 points with non-zero distance between each pair of points, then it will not satisfy " \( d( a,b) \geq d (a,c) + d(c,b) \)" for all triplets of distinct points. This means that even if we add the restriction of "path cannot intersect itself", we are still doomed to trivial spaces. – Calvin Lin Staff · 2 years, 3 months ago

Log in to reply