I have noted before that, despite the name, not all legal fictions are in any obvious way fictional; the 'fiction' in 'legal fiction' seems to indicate not (as philosophers of law have often taken it) 'falsehood' but rather 'artificial construction for a domain'. Legal fictions may at times be falsehoods treated as true for convenience, but the label also includes things that are really such and so for the purposes of law; when so constructed, it is not a falsehood but a truth that they are such and so in legal contexts. If a whale is counted as a fish for purposes of fishing law, this is not a falsehood, but a fact; it is just a fact that depends on the interpretation and application of the law rather than the biology of whales and fish.
Legal fictions are inevitable in law because a very large portion of law is entirely about practical classification. Does this count as a fish for such-and-such statute? Are this person's actions of a sort to make him responsible in a way that matters to us for a court decision? Who counts as a citizen? Which behaviors are subject to punishment? Does this crime meet the requirements for this specific kind of sentence? Does this business firm have basic legal protections as a whole under a law that concerns persons, or is it necessary to spend the time and resources to sort out the specific legally relevant protection of each and every actual person in the firm? Indeed, almost all of the distinctive features of law as a field have to do with the fact that a legal system is a huge jumble of systems of artificial classifications for a very large variety of practical purposes.
However, law is far from the only field that uses fictions in this way. A good example of a field that extensively uses fictions is physics. Physics fundamentally differs from law in that it is much more theoretical than practical, and this does have a significant effect on how physicists use scientific fictions to do their work. But there are at least two practical purposes that lead to fictions that have many similarities to legal fictions: pedagogy and problem simplification.
Physicists use a lot of pedagogical fictions; they are notorious for them. Teachers in physics courses are always saying things like, 'Imagine an evenly spaced field of clocks' or 'Think of an electromagnetic wave in this context as a sound wave' or 'Think of the atoms as being people in a crowded room'. Part of the reason for this, I think, is that the teaching of physics mimics in a simplified way how physicists often solve problems and make discoveries, namely, by comparing the structures of different problems, finding similar structures, and then fine-tuning to take into account any differences. In advanced physics, this is often quite abstract, and often occurs at the level of mathematical equations; but not always, and you can fairly easily find physicists trying to get a first beginning on a difficult problem by first imagining a more loosely defined but simpler problem. And in pedagogical contexts, of course, the context itself limits the abstraction, so physicists in teaching things are always trying to give narrative pictures that are very loose but capture some of the relevant structure.
Physicists are not, of course, unique in this regard; virtually all teachers use these sort of artificial constructions ('For our purposes here, think of X as being Y' where this is only true 'for our purposes here', whatever they might be) to bridge the gaps between where students are and where they need to be. Pedagogical fictions give a way to direct the attention of students to the things to which they need to be attending, and a way for them to start getting used to thinking of situations in certain ways without feeling completely lost and getting discouraged. Many of these pedagogical fictions are fictional in a fairly straightforward way, stories told to help orient students; but other pedagogical fictions are actually simplified versions of the physicist's actual problem-solving toolkit, and thus serve a broader purpose in the field of physics. These fictions are in some sense more interesting, and they also tend to be the fictions that have the most similarities to legal fictions, because they really are about problem classification.
We all know the standard postulation process for physics problems -- we get problems where we assume there is no friction, or no air resistance, or no outside forces. These are sometimes known as spherical cows, based on the physics joke. A farmer wanted to know how many cows he could pasture on a weirdly shaped piece of land, so he went to the smartest person he knew, the physicist at the local college. (Physicists are always the smartest people in physics jokes.) When he told his problem to the physicist, the physicist said, "Let me see if I can come up with something helpful tonight; come back tomorrow and I'll let you know if I have."
So the farmer came back the next day, and he was met at the door by a very excited physicist. "Your problem was a great problem, but it turns out that it's easy to solve. All we have to do is assume that every cow is a perfect sphere!"
What makes the joke so lasting is that this is exactly how physicists handle many, many kinds of problems and come up with extremely good solutions for them. Cows, of course, are not spheres. But they are three-dimensional objects, and they need space around them which, because the cow has to turn around, is not all that far off from a circle. And a three-dimensional circle is a sphere. But, of course, the key is that if you assume that every cow is a perfect sphere, you turn a complicated problem about cows into a relatively simple sphere-packing problem. The mathematical apparatus for handling sphere-packing problems is very, very well-developed. So by assuming that cows are spheres, you place the problem that needs to be solved under a category of problems that is very well understood.
It ties in to something once said by a physics professor in a class I was once in (it was actually an intro astronomy class): The first step to solving any problem is to figure out what kind of problem it is. Physicists use problem-simplifying fictions, postulates as they are sometimes called, to figure out what problem they are actually solving and remove anything that could obscure the path to the solution. Assume that a body is perfectly rigid, assume that a volume is actually a point, assume that lines are perfectly straight, assume that there is no gravity, etc. The point is not the assumption, which is often not true (although occasionally it may be -- your assumption that there is no air resistance, for instance, could conceivably turn out to be the case if you later discovered that the interaction was happening in space), but the problem: the scientific fiction, properly speaking, is that a cow problem is a sphere-packing problem. And for the purposes of solving the problem that is true.
This (true for the purposes of solving the problem) is more or less the end of the story for the legal fiction -- we might change over at some point to something that works better, but if the legal fiction works, it works. In physics, of course, it is not, precisely because all practical matters in physics subserve larger theoretical aims, and you need not only to know how to co-classify problems for practical results but also, as Pierre Duhem noted, how well those artificial classifications capture the relevant natural classifications.
The same is true, of course, for all the sciences; if treating whales as fish gives us an appropriate classification for a practically sensible fishing law, that is justification enough, but while a biologist might conceivably treat a whale as a fish for a particular problem, it matters to the biologist that a whale is in fact not a fish, because that's something you need to recognize to understand whales and fish. A physicist may solve a particular problem by assuming perfectly rigid bodies, but it in fact ultimately matters to the physicist that the bodies he's dealing with are not perfectly rigid, because that's something you need to remember in trying to understand the physical world. In law, a realm full of the artificial, it often doesn't matter that your classification is artificial, even at times to the point of being entirely arbitrary; in scientific fields you don't want your classifications to be entirely artificial, much less arbitrary. Scientists spend their days solving problems, but science is not a sport of problem-solving; problem-solving is not valued solely for its own sake, but for what it contributes to our understanding of the world.
