I'll take a shot and providing one potential answer (other historians here can probably come up with their own unique answers).

It's hard to answer "what will they do in the future" since, it's the future and who knows what context their situation will be. But, I can provide perspective on how historians today decipher factual accounts from fictional accounts.

I deal with a particular aspect of history that frequently gets fiction mixed into it, even during the period in which it occurred - the Golden Age of Piracy of the late seventeenth and early eighteenth century. Since serious scholarship on pirate history began in the 20th century, some accounts of pirate history have been revealed to be outright fiction or unfounded local legends. For instance, the famous Charles Johnson's [99% certain it was Nathaniel Mist - Johnson was his pseudonym] General History of Pyracy a period publication from 1724-1728, has entire sections of fiction since it's a period example of "factual fiction". In those works, fiction is mixed into the fact when convenient for plot, for a point the author wanted to make, or to make the book more appealing to reading audiences (among other reasons). The most notorious example of complete fiction in that book is the chapter in the second volume on a Captain Mission. The whole chapter has been shown to be a fictional invention.

Now, how do we today decipher if 300-year old accounts are fictional? We research a lot of different aspects of the account. For one, the author is researched if possible, to see if they existed, what they did for a living, if they would have any involvement in the situation they described, and in general just validate the authorship (see if someone else was possibly writing it). The writing style can also be examined to see not only if it was written by another known writer of the period, but also if the style of the narrative fits with something a witness to historical events would tell (rather than a conveniently written narrative for entertaining an audience).

Finally, we use other period documents and records to verify the writing in question. Are there other period records that describe the event? How do they compare to the one in question? Is it obvious that the questioned document utilizes other sources or does the writing include a unique, yet consistent, perspective about the historical event the questioned document pertains to? For pirate history, an historical account can be compared to a number of pirate trials, depositions or testimonies from captured pirates or victims, government communications that detail what information they gathered about pirate activities, Navy records about the pursuit of pirates (and the information they gathered in the process), period newspaper accounts, and more. Examining the context of anything is valuable. Looking for how accurate description of minor background information about something like a place visited can be useful. While a narrative might be mistaken on a few occasions about something, too many or too large of errors can be suspicious.

That's how those who study pirate history go about the subject of telling fact from fiction (or at least it's what I could think of for now). As for our historians in 500-1000 years? I have no clue what Professor Farnsworth will have to deal with at that point. Will our age of digital data-keeping come back to haunt us? I don't know. I guess you'll just have to have faith that something will be left by us that expresses what life was like for us. I know there are those out there who have faith - I always look at the people who run the "longplayer" for that. Those guys have faith that their 1,000-year song will be able to be played for 1,000 years. They have faith they can train and maintain enough people to play that song for that long, faith in having stable enough governments to allow people to stay around to continue performing it, and even faith that the building they are in will shelter the performance of the song for the period. It's a lot of faith in humanity, but I guess someone has to have it.

Historians put documents in context. In your example, a future historian studying a fictional novel would ask himself some questions. What type of document is this? Who wrote this? When? What was his purpose? In your example, if the book were in complete condition then it would not be difficult to find who the author was, when he wrote the work, and that his purpose was to create fictional entertainment.

It becomes more difficult in the case of someone trying to pass off fictional accounts as fact, or otherwise intentionally or unintentionally misrepresent or distort events. This is why historians always employ supporting evidence, and try to utilize a wide range of evidence. Historians do not just study a single document in a vacuum. They would employ as much supporting evidence as possible to construct their understanding. If any information on the author was available, they would carefully research it to better understand his purpose in writing the document. They would study the document to find any inconsistencies with known facts.

There is not a binary distinction between "factual" and "fictional" Fictional accounts can have bits of truth in them, and factual accounts can be embellished, distorted, or contain outright misinformation. For instance, consider a novel written and set in the 1990s. A future historian might be able learn quite a bit about 1990s life and culture from it! Of course, separating the truth from the fiction is a daunting and difficult task. The example novel may replace Bill Clinton with another, fictional president whose daughter is kidnapped by terrorists for narrative purposes. In that case, the future historian would need other documents to understand that that part of the story was fiction. Even in the case of a purely fantastical story, the document could tell you a lot about the culture of the period.

As an undergrad I wrote my senior thesis on medieval Saint's Lives. They're filled with miracles, allegories, and misunderstandings, but for a large period of early medieval history these are some of the best documents that we have, and they tell us a lot of about early medieval culture and events. It's the historian's job to try to get as much information as possible from all historical evidence by combining disparate pieces together. That example is slightly different from fiction in the modern sense, because people in the period believed the stories to be true, but hopefully it gets my point across.

It's important to note that fictional stories can be of tremendous worth to historians. Fiction can be read to reveal certain assumptions about how we view the world around us. For instance, Arthurian romance literature is an excellent source that can tell us a great deal about the political, social and cultural realities of high/late medieval Europe. The Death of King Arthur allows us to examine the changing relationship between king and aristocracy at a time when royal power was becoming increasingly centralized. Sir Gawain and the Green Knight can be viewed as a celebration and critique of knightly aggression. While these fictional stories are exciting reads that often include fantastical events, they are still of great value to historians.

Another important point to stress is that factual accounts are almost always problematic. Primary sources can conflict with one another. If we were to attempt to describe the sack of Rome by the Gothic king Alaric in 410 we might read Jerome's Letter to Principia or Augustine's City of God. Both texts describe the event and contain factual information (although they did not witness the event themselves). However, they are highly emotional and opinionated accounts. In them we read about “bloodthirsty” savages that conquer a city whose walls have already been weakened by lust and avarice. If we look to Orosius' Histories Against the Pagans we receive a slightly conflicting interpretation of the same event. Like Jerome and Augustine, Orosius presents the sacking of the city as divine punishment. However, Orosius downplays the severity of the siege and presents the ultimate outcome as a victory which unites both Romans and barbarians. The point I'm trying to make here is that these authors shaped and arranged the factual information about the sack of Rome in such a way as to accentuate their argument. That doesn't mean we have to discard these texts as “unusable.” It just means that when reading factual accounts we have to be aware of an author's biases.

Sorry if my response is a bit excessive or divergent. I think your question is interesting and approaches the subject of how historians engage critically with the historical evidence. What I'm emphasizing is that a historian's job is not simply to discern what is fact and fiction in historical texts but to understand how an author collected and shaped the facts.

The Death of King Arthur. Trans. Simon Armitage. New York: W.W. Norton and Company, 2012.

Sir Gawain and the Green Knight, Pearl, Sir Orfeo. Trans. J.R.R. Tolkien. New York: Ballantine Books, 1975.

Jerome's Letter 127 (To Principia) and Augustine's City of God (Book I. Chaps. 1-10) can be found at: http://www.newadvent.org/fathers/

Orosius, . "Histories Against the Pagans." From Roman to Merovingian Gaul: A Reader. Ed. Alexander Murray. Ontario: Broadview Press, 2000.