Earth spins about an axis that runs approximately from the North Pole to the South Pole. It completes one rotation per day (and, indeed, a day is defined by the period of rotation). the degree of tilt of the axis never changes. This is because one of the fundamental laws of physics is that angular momentum must always be conserved. If the angle of tilt changed then the direction of the angular momentum would also change and this isn't possible without some sort of external influence (like an asteroid, which isn't quite something we want to happen).
What is this angle of tilt relative to? Well, it's the amount by which the axis of rotation differs from making a 90º angle with the plane of Earth's orbit around the sun. Hopefully the image below helps.
|Nabbed from Wiki. Credit: Dennis Nilsson, NASA.|
|Light and heat from the sun is hitting more of the southern hemisphere than the northern hemisphere.|
The greater the axial tilt, the more pronounced the differences between summer and winter will be (see the bit about Uranus at the very end for a very extreme case).
OK, so axial tilt causes Earth's seasons. What else can cause seasons?
Near and Far
Another possible cause of seasons is an elliptical orbit. This occurs when for part of the year a planet is noticeably closer to its sun than for the rest of the year. So when the planet is physically closer to the sun, the whole planet experiences summer (not just one hemisphere). During the more distant part of its orbit, the whole planet would experience winter.
On Earth, key seasonal dates, such as the solstices and equinoxes, are defined by the length of daylight (shortest day/night of the year and equal day and night respectively). On a planet with an elliptic orbit the key dates would be defined by significant points in the orbit. The equivalent of the summer solstice would be the periastron, the point of the planet's closest approach to the star. The winter solstice would be replaced with the apoastron, the time when the planet is furthest from the sun.
An interesting thing to note is that, thanks to Kepler's second law, a planet will move more quickly in its orbit when it's closer to its star than when it's further away. (If you follow that wiki link, there's a nice little animation which sort of explains it.) The result is that summer on such a plane will be briefer than winter. The more eccentric (non-circular) the orbit, the greater the difference between the lengths of summer and winter (and the closer the planet will be to its sun during summer). Could make for an interesting cultural interpretation of the seasons.
Some stars vary their brightness. Cepheid variables, for example, pulse with a regular period (which can be anywhere from one day to a few months). Theoretically, this could induce a seasonal variation for any surrounding planets. However, there is a problem when it comes to life evolving on such a planet. These stars are unstable and won't last very long (on an astronomical scale) in their pulsing state. This makes it a bit more difficult to justify having an inhabited planet around them. Maybe a planet with a colony that's studying the star. Anything more natural probably wouldn't last or might not have enough time to have evolved (depending on the size of the star). Of course, there's no rule saying impending doom couldn't be central to a plot.
|Uranus is sideways. (Also, the rings aren't |
really red, just coloured that way for emphasis
here.) Credit: Lawrence Sromovsky, (Univ.
Wisconsin-Madison), Keck Observatory.
Of course, Uranus is too large and with too dense an atmosphere to support life as we know it. However, it's possible that a rocky planet more suitable for human habitation could also have this kind of extreme tilt. Everywhere except on the equator there would be periods of multiple days of darkness. Even on the equator, polar summer and winter would be spent in perpetual twilight.
I suspect this sort of configuration could also cause interesting weather/climate issues, but I think that would depend a lot on the atmosphere as well. Could be problematic.