Writing a story where people have evolved to harness their brains to have control over different elements depending on personality. Is 1000 years a long enough time for this to plausibly happen? Does there need to be a bigger time gap into the future?
Related question: Is waterwalking possible?
The short answer is 1000 years is probably not long enough.
Evolution is a very slow process. Even very minor changes can take hundreds of thousands to millions of years to develop. Evolutionary processes span many thousands of generations. For instance, this article describes how adding 50 billion neurons to the brain of Homo erectus to form the brain of modern humans required two million years! The adaptation you describe seems very advanced, and I would expect such a change to occur over an even longer period of time.
The determining factors in this scenario are going to be two things: the mechanism for the ability you describe, and the mechanism for selection.
By mechanism, I mean the physical process by which the adaptation you want takes place. In familiar paleontology, this might be realized by an animal being able to run faster than its ancestors because its skeleton is shaped differently. It the skeleton only needs to change a little to get this result, the evolutionary process is faster (but remember, fast on en evolutionary scale can be 100,000 years). If you can find a plausible explanation for the adaptation you want which requires a minimal physical change in the brain, then your evolution could occur quickly.
The mechanism for evolution is the reason this trait develops. Evolution is based on survival of the fittest, so for a trait to develop, there has to be a reason why individuals carrying the trait are more likely to survive and reproduce than those that don't. There is an urban legend of sorts that evolution will lead to humans losing their pinky finger because it serves no purpose. I don't think that this is true. Even if the pinky does serve no purpose, nobody is more likely to die or not reproduce because they have one. Conversely, you must find a reason why people with the adaptation you describe, are more likely to survive and reproduce than those that don't. Even more importantly, evolution doesn't happen in one step. It is the result of many thousands of generation, each with small changes. You have to determine what small changes lead to the result you want, and why each change is a biological advantage for the individual possessing that change.
Evolution occurs quickly when disadvantaged individuals die off before reproducing. This means that humans are less likely to evolve quickly from here on out because we take care of those members of our species who are dying. With average life expectancy well beyond the child bearing years in most parts of the world, there is not really any such thing as biologically disadvantaged anymore, at least, not on a large scale.
Finally, note that evolution happens fastest in small, isolated groups. The human race is basically the opposite of that. The idea is that it is easier to evolve a few individuals at a time than billions. Humans are so homogeneous that given our current culture, we would have to evolve all at once, which is much slower.
All in all, I think that for any creature whose life span is measured in years, little to no significant evolutionary changes could be seen in the species as a whole in just 1000 years.
Perhaps 10,000,000 years would be sufficient.
Natural and artificial evolution have been covered by Alex S and Henry Taylor, so I will present option 3: controlled breeding.
Humans have been doing this for at least 5000 years with domestic plants and animals, selecting for qualities we find desirable and breeding the stock to get these qualities an heritable characteristics. Modern maize and wheat look nothing at all like the ancestral plants from Central American or Turkey, and the multitudes of breeds of dogs and cats (much less cattle, sheep or poultry) should convince you that this can be done very effectively, and fairly quickly as well.
The two primary bottlenecks for doing this to humans are cultural and biological.
Culturally, although it is quite possible to "breed" humans for traits the same way we breed dogs or cattle, this is frowned upon. The Eugenics movement of the early 20th century may have been the best organized attempt to do so, with proponents ranging from the founder of Planned Parenthood in the United States to a well known German politician in the 1933-45 period, but the implementation and results were so horrifying that it is pretty much swept under the rug these days. (To put it into perspective, virtually all proponents of Eugenics were thinking in terms of eliminating "inferior" breeding stock, targets being Black Americans or European Jews depending on where the eugenics movement happened to be centred). The notion of "breeding" humans seems to be looked on negatively in almost every historical culture, however, so there are probably deeper roots to resistance to the idea than the experience of the 20th century ("culling" the "inferior" stock is probably the sticking point among normal human beings).
The biological reason selective breeding will be difficult in Humans is because we have long life spans and reach sexual maturity fairly "late" compared to other animals, so breeding for traits will take a fairly long time to show results (much longer than breeding dogs, for example). A breeding program for humans, if it was sanctioned and supported, would take generations to show results, long past the lifetime of any originator and difficult to keep the goals aligned over a period of centuries. That said, assuming al the objections could be overcome, a 1000 year period of selective breeding could be enough time to get the traits you are looking for.
Ruling out natural evolution (thanks to Alex S's excellent answer), you might want to consider artificial evolution...
At the rate that our medical knowledge has advanced over the last hundred years, it is easy to believe that major genetic enhancements will be applied to members of our genome long before your 1000 year deadline.
Also, given the population pressures that are already facing our planet, it is easy to believe that breeding rights may one day be restricted to only the most successful participants in corporate and financial battlegrounds of the future.
With these assertions in place, you only have to justify how your particular psychic element-manipulation abilities would help their owners win breeding rights. From there, its just a numbers game.
1000 years is plenty of time
I would say it's plenty of time for a species with a short lifecycle, or a situation of quite extreme environmental change. You could define (roughly) the speed of evolution by "environmental pressure divided by lifecycle length" (you could go more complicated by adding something in there relating to mutation rate).
With human scientists experimenting on them, evolution can occur very rapidly in fruitfly populations, for example (fruitflies, or Drosophila Melanogaster to give them their proper name, are used in a lot of research). Note that a "population" here is a group of fruitflies in a lab - i'm not talking about the species as a whole evolving.
Even for a species with a long lifecycle, eg humans, a powerful environmental pressure can still produce noticable results over 40-50 generations (what you might get in a 1000 year period). It's quite plausible that within the next 100 years there will be routine screening of all embryos for genetic "defects" (quoted because it may actually include selection for preferred traits, as well as eliminating congenital diseases for example). This is going to result in an "evolution" of the human race, if it's applied widely enough.
Evolution can proceed much faster than is often thought, via "punctured equilibrium".
In "normal" times, a species is in equilibrium with its environment. Its range of natural variation creates individuals better suited to the "edges" of that environment than the mainstream, either by the random permutation of existing genes or by the occurrence of mutated genes. These latter will persist if they do not cause any great damage to the survival chances of individuals inheriting them. At a later time they may become crucial to surviving in drastically altered conditions, should "edge" suddenly become the only survival opportunity.
In mathematical terms, and often also in everyday terms, equilibrium can be overturned by a catastrophe. That's something which exerts extreme pressure on individuals to adapt or die. As far as I know, the only such factor to have acted this way on the globally successful species homo sapiens are global pandemics. Also, the primary effects on our genetics are not obvious without the tools of 20th century medicine. For example, blood groups A and B confer resistance to cholera (and near-immunity in the case of AB). These groups are common in Europe, which has cholera. They are rare in native Americans, because cholera did not exist in the new world until recent times.
I do wonder whether AIDS might have completely reshaped humanity had it emerged in (say) the 15th century. Who is most likely to catch it? Those who have multiple sexual partners. Who are most likely to successfully raise their children to adulthood? The strictly monogamous couple. Absent modern medicine, it would have been a strong evolutionary pressure in favour of strict monogamy. "Human nature" might have become very different after five hundred years, if we'd never had a cure.
So to your story. If you want rapid evolution, you need something that afflicts virtually the whole human race, that causes death (or sterility) for individuals who do not have magical abilities. It's not necessarily a direct effect. If there's any genetic linkage between a gene that confers resistance to a particular plague and a gene that enables magical abilities, then repeated waves of the right plague will coincidentally select for magical ability. Then, if people start to recognise and desire that ability, they'll (possibly unconsciously) select partners who have or who share that ability, after the plagues have become history.
Ten generations (300 years) might be enough, absent modern medical knowledge.
Is 1000 years enough? I don't know. You are literally talking about a magical property with unknown mechanics and origins. It could be enough, it could not.
More likely it isn't enough, but we can play some games to make it appear to be enough. Evolution is not wasteful. It is rare to see an entirely new organ grown to manipulate elements -- and that would likely take far longer than 100 years. However, working with what you have, you may have nearly had the physical structures required, they were simply around for different purposes. We never evolved the ability to manipulate the valves on a trumpet. We evolved fingers for other purposes, and then figured out that we could use the trumpet.
Thinking this way, your ability to manipulate the elements could be a psychological thing, a learned ability. Simply no one learned how to do it until recently. That would easily be doable in a 1000 year window, and has the nice advantage of creating an odd puzzle: why did the body develop the correct anatomy to develop this ability? What behaviors in our lives evolved the right pieces? Is there a greater plan? All of these questions would remain.
This is not far fetched. Most of us cannot echo-locate like bats, but it turns out we have the hardware. Of course, nobody knew our body was capable of echolocation until we started to see blind people clicking their tongue while they walk, and echolocating with that sound. We've had the hardware to do this for milinia, we just haven't learned the skill to let us use that hardware for echolocation.
