The free will that humans enjoy is similar to that exercised by animals as simple as flies, a scientist has said.
The idea may simply require "free will" to be redefined, but tests show that animal behaviour is neither completely constrained nor completely free.
The paper, in Proceedings of the Royal Society B, suggests animals always have a range of options available to them.
"Choices" actually fit a complex probability but, at least in humans, are perceived as conscious decisions.
The idea tackles one of history's great philosophical debates, and Bjoern Brembs of the Berlin Free University brings the latest thinking from neurobiology to bear on the question.
What has been long established is that "deterministic behaviour" - the idea that an animal poked in just such a way will react with the same response every time - is not a complete description of behaviour.
"Even the simple animals are not the predictable automatons that they are often portrayed to be," Dr Brembs told BBC News.
However, the absence of determinism does not suggest completely random behaviour either.
Experiments including Dr Brembs' own 2007 work with flies has shown that although animal behaviour can be unpredictable, responses do seem to come from a fixed list of options.
"Free will is not that lofty metaphysical thing that it was until the 1970s or so," Dr Brembs said.
"More and more people are realising that it's a biological property, a trait; the brain possesses the freedom to generate behaviours and options on its own."
The exact mechanism by which brains - from those of flies up to humans - do that generation remains a matter for experiments to more fully prove.
Dr Brembs and others have used mathematical models to simulate brain activity on a computer, finding that what worked best was a combination of deterministic behaviour and what is known as stochastic behaviour - which may look random but actually, in time, follows a defined set of probabilities.
This "stochasticity" shows up in, for example, earthquakes. While they cannot be accurately predicted, a given fault will over time show earthquake timings that neatly fit a curve.
As with animal behaviour, there is an underlying order and probability to a process that may appear random.
"It is a probability, and that's as far as we can take it if we try to abstract it from thinking," he explained.
"In thinking, we have all the options, and theoretically all the options have the same probability attached to them. However, this is not how it's going to turn out."
While we imagine having the option of choosing to walk off the edge of a cliff, Dr Brembs said, it is an option that would only very, very rarely be chosen.
Dr Brembs said brains are likely to include mechanisms that can ramp up or down the probabilistic element of the behaviour, depending on the situation at hand, and that in sum, the whole system was an evolved survival strategy.
"The variability that is inherent in the behaviour is something that is a prerequisite for survival in a competitive environment."
That is, a predator should not be able to always guess its prey's actions, but the actions should not be so random as to include options even more dangerous than the predator.
The idea holds a certain currency in the world of neurobiology, but it is clear that what is needed is more experimental results that match the mathematical models.
Christof Koch, a biologist from the California Institute of Technology and frequent author on topics of free will and biology, said that the work hits at the heart of "one of the oldest problems in philosophy".
In writing about Dr Brembs' research, he suggested that "the strong, Cartesian version of free will—the belief that if you were placed in exactly the same circumstances again, you could have acted otherwise—is difficult to reconcile with natural laws".
"There is no way the conscious mind, the refuge of the soul, could influence the brain without leaving tell-tale signs. Physics does not permit such ghostly interactions."
Professor Koch told BBC News that "it is entirely possible that 'indeterminism' - under certain conditions - is a useful trait - but not usually: when you drive your car at high speeds down the freeway, you want to be highly deterministic."
Dr Brembs stressed that the current debate did not address the even stickier debate about consciousness, its origins, or whether animals share it.
"I would not expect fruit files or worms to comtemplate their options; I would think this is something that is clearly more built in than it is with us. But I would then say that most of the decisions we're making are also built in," he explained.
"[Free will as described in the paper] is a very low-level, necessary prerequisite, but it's not even close to being sufficient for addressing things like morality and responsibility.
"But without this very basic capability of choosing between options, we wouldn't have to think about all the other things that come on top: consciousness, upbringing and what have you."