In a setting with multiple sapient fantasy races (ie, elves, orcs, humans, dwarves), assuming these races could interbreed and this activity became common, would they eventually reach a point of hybridization where there is really only one race that traces its ancestry to the all of the others? If so, would that race resemble a potential common ancestor of its predecessors, or would it be something entirely new?

  • 5
    $\begingroup$ Human populations are much less different between them that orcs, dwarfs and elves and we interbreed with a passion. Yet somehow we are far from reaching a uniform brown color. (Look up "assortative mating". Birds of a feather flock together etc.) And the final result of genetic mixing will obviously be different from the common ancestor -- after all, evolution did not stop during all those hundreds of millennia. $\endgroup$
    – AlexP
    May 13, 2019 at 14:52
  • 2
    $\begingroup$ Given how, in the real world, minor variations of skin color have justified hostility, hatred and bloodshed, it is highly unlikely that races with much more distinct differences would cohabitate peacefully enough for long enough for a full genetic blending to occur. Could it happen? Yes, with the results looking like a mixing of each race's traits based on their genetic dominance over similar traits from other races. Would it happen? probably not. $\endgroup$ May 13, 2019 at 14:57
  • 5
    $\begingroup$ ", assuming these races could interbreed and this activity became common" - These are some pretty huge assumptions to start from. $\endgroup$ May 13, 2019 at 15:03
  • 1
    $\begingroup$ I remember an experiment involving separating two populations of fruit flies in containers on either side of the room and breeding them for different traits. After a few generations when the populations were reunited they would/could no longer interbreed with each other. $\endgroup$
    – Muuski
    May 13, 2019 at 16:44
  • $\begingroup$ Even if they can interbreed, there is no guarantee that their offspring will be fertile. The offspring of horses and donkeys (mules and hinnys) are, with a few rare exceptions, infertile. $\endgroup$ May 14, 2019 at 7:56

6 Answers 6


Not completely.

Humans are fully interfertile, and apparently were so with other human species (genus homo) in past history. Yet, despite hundreds of millennia of interbreeding, we have never become homogeneous. We all (as modern homo sapiens) carry genes from Neanderthal, Denisovan, the recently discovered fourth species, and possibly from Floresiensis and the newly discovered species from the Philippines. We've been interbreeding on at least three continents between different "races" for as long as there have been modern humans (quite possibly longer) -- and yet we still have at least three major races and a bunch of offshoots.

I read a study once (in Scientific American, decades ago, so no link or exact reference here) that had found that all that's required for neighborhood segregation to continue is a pretty small preference for having neighbors like yourself. Add to that the known biological phenomenon of genetic attraction (the attraction for those genetically most similar to oneself, offset only by the effect of being raised together as family), and it's no big surprise that even after tens of thousands of years with only one human species on Earth and routine trade and travel (far predating latecomers like Marco Polo), there are still at least three major races.

Given this, it seems unlikely that multiple actual species would become homogeneous, even over short geological time. There might well be a true-breeding race of "half-elves" who started as hybrids, but there will still be both elves and humans. And orcs, and dwarves, and whatever else. The tendency, in fact due to the combination of biology and psychology, absent a strong selection factor, will be for the number of recognized "races" to increase over time, rather than decrease.

  • $\begingroup$ But every homo sapiens today has genes from these once distinct races, wouldn’t that contradict your claim? Or am I mistaken and not everyone today has traces from other homos? $\endgroup$
    – Soan
    May 13, 2019 at 16:41
  • $\begingroup$ We all carry Neanderthal, Denisovan, and the fourth species' genes. Yet Neanderthals and Cro-Magnons were distinct species right up until Neanderthal extinction. We did not homogenize. $\endgroup$
    – Zeiss Ikon
    May 13, 2019 at 16:59
  • 2
    $\begingroup$ H. sapiens and H. neandethalensis were separate species in western capitalist paleoanthropology. In eastern socialist paleoanthropology they were subspecies, H. s. sapiens and H. s. neanderthalensis, because eastern socialist paleoanthropologists justly suspected that they interbreeded -- there were some intriguing intermediate fossils. Today, given what we know, their status of distinct species is again questioned. And not all of us carry Neanderthal genes -- sub-Saharan Africans don't (or, at least, do at a much lesser degree than Eurasians). $\endgroup$
    – AlexP
    May 13, 2019 at 18:35

It depends.

If there are reasons for races to separate then it is likely they will remain separated

  1. Physical - for example Dwarves being adapted to underground living, elves to forest living etc) .
  2. Social - if the races are xenophobic or at war with each other then interbreeding is less likely.
  3. Geographical - if the races are spread out over a large enough area that not much mixing happens you would see some "blurring" at the boundaries but no grand homogenization.

On the other hand if the factors above are neutralized then you would expect a hybridization process to take place.

If they had no physical reason to live apart, if the different species got on well and had no taboos against interbreeding, and if they all shared a small enough range then eventually you would end up with one species.

  • 2
    $\begingroup$ Another social factor to take into account: members of one race are unlikely to find members of another attractive enough to mate with them and will tend to favor their own. $\endgroup$
    – Muuski
    May 13, 2019 at 16:46

Given enough time almost certainly yes, but you're talking about many many generations. What comes out of the process may resemble an ancestor or be something completely new depending on a couple of factors:

  • Unique mutation fixation, if the distinct races have mutations that their shared ancestor never did and one or more of those mutations become "fixed" in the hybrid population then the result may not resemble anything that came before.

  • Founder effect, the make up of the first few generations of hybrids is going to have a disproportionate effect on the outcome compared to the races that come late to the interbreeding party. For example if Dwarves and Humans have been interbreeding for generations forming a genetically stable subspecies that can breed true within its own numbers and a few refugee Orcs and Elves are added to the mix only when both species have almost been wiped out the resulting race will technically be a four species hybrid but most of its genetics are human/dwarf with very little owing to its few Orc and Elf contributors.

  • Dominate genetics, it may be that a certain race, or races, has genotypes that are dominant in compared to similar genes in other races so for example orcish tusks may be a dominant dental formation complex leading to all hybrids being tusked, or elven hair genes may be dominant making all hybrid children blonde.

  • Underlying genetic differences, lack of, Tolkien's Elves and Orcs are, genetically, one species, but a lot of malevolent magic has been pumped in to twisting Elves into Orcs, how this magic factors in to any off-spring has the potential to change outcomes radically.

  • Transcendent traits, some traits that each race has may be due to non-genetic factors which may or may not fade as the generations mount up.


In the real world, if two creatures can interbreed, then they share a fairly recent ancestor, in evolutionary terms. There's essentially no way of getting around common ancestry in the real world, though.

However, in a fantasy world, where entirely separate kinds of life can exist, interbreed, and be brought about by magic, none of these rules might hold.

If some gods or mages wave their respective staves, rods and wands, in a non-euphemistic sense, and magically cause to spring into being from whole cloth the Elves and the Sentient Trees, and they can interbreed, then the whole thing goes out the window.

In that case, the question requires us to make some assumptions:

  • There are two purebred populations A and B
  • They begin to interbreed at some rate, producing AB hybrids.
  • If an AB successfully breeds, the result will be another AB.
  • Neither side dies out for the duration of our experiment.
  • The two populations already have a single common ancestor.
  • We have infinite time.
  • Both populations are of finite size, and cannot grow to infinite size.

Given these assumptions, the answer, I think, is a very firm "if the rate of population increase is slower than the rate of interbreeding, then eventually yes, otherwise maybe".

It simplifies things to consider only half the problem: does population A ever come to all be descended from B?

Consider the first interbreeding A+B pair. An AB hybrid is introduced into the A population.

If the population ever reaches a point where nobody is a descendant of that AB, their lineage dies out, so we consider the next breeding pair.

If the population ever reaches a point where everybody is a descendent of that pairing, then the answer is "yes".

Those are the only two outcomes. At this point it should be obvious that there's no "no, they don't" outcome, there's only a "yes" and a "try again next time two purebreds interbreed". It might take forever, but they will eventually become all purebreds.

So, the answer's "yes" so long as the number of AB members of the the A population that the As are willing to breed with, grows faster than the number of purebred As.

The problem then becomes mathematical. When 100% of an AB's potential partners are likely to be As, the rate of growth of the ABs will likely be close to 100% per generation - they will all partner with As.

100% is a lot, but numerically, with only a single AB, this is the lowest-growth point for them. If the purebred A population grows by 2 in the same generation, then the proportion of ABs actually falls.

If ABs reach 50% of the whole population, then 50% of male ABs will partner with female ABs, and 50% with female As (and vice versa for female). So their population will grow only by 50%, but this will be numerically huge: the population will go from 50% AB to 75% AB in a generation.

When it reaches 99%, though... there are probably reasons those remaining 1% of purebred As remain purebred. Perhaps they are part of a culture which is extraordinarily insular and inbred. But over an evolutionary timescale, the inbred microsocieties will either die out, or interbreed. Even if they outlast the greater society, that just means (from our assumptions) that they instead will start to breed directly with the Bs.

So it just becomes a matter of time.

An illustrative exercise is a very simple Solitaire game:


Split a deck of cards into a red deck and a black deck.


1) Replace the top card of the black deck one from the red deck: if the card taken off was a black card, place it in a discard pile. If it was red, place it back in the red deck.

2) Shuffle the black deck, and deal it out in 12 pairs:

  • If a pair is two of the same color, leave them unchanged.
  • If the pair is a red and a black, replace the black with a card from the red deck, placing the black card in a "discard pile".

3) Take your pile of dealt pairs, and repeat from 1).

That's it!

This version of the game represents the best possible case, where ABs (represented by the red cards) are always successful. Even this simple case will take some time to play, but it's clear that there's only one possible outcome: all the cards will eventually turn red.

To get the reproductive success lower, play again but this time, replace as:

  • If a pair is two of the same color, leave them unchanged as before.
  • If the pair is a red and a black regular card, flip a coin: -- heads: replace the red with a card from the red deck, placing the black card on the discard pile. -- tails: replace the red with a black card from the discard deck, and place the red back in the red deck.

This represents equal reproductive chance. But again, red will eventually win, as there is no "lose" state for red. Red will have generations where every coin flip comes up tails, and the red population will decline or even be wiped out. So the game will be a whole lot longer, but just mathematically, there is no outcome other than that all black cards are replaced by red.

You can play AGAIN, this time rolling any number of dice, and only replacing the black with a red if they all come up with sixes. This represents a case where there's a HUGE reproductive disadvantage to the ABs.

And yet, eventually, possibly some time after the heat death of the universe depending how many dice you choose to roll... the black deck must eventually turn all red. Because there is no other possible end-state to this game.

What if you want to reduce the rate of interbreeding? Change step 1 to:

1) If you know the black deck holds only black cards, replace the top card of the black deck one from the red deck: place the removed card on the discard pile.

This represents an interbreeding rare so low, that nobody interbreeds when there're any hybrids. (there might be many generations of interbreeding in between the death of the last hybrid, and the first interbreeding, but we skip over this).

Even so, with that one "seed" red, either the whole deck will turn red (perhaps because you always roll only 6s whenever that one came up), or you will try again with another "seed" red. There's no other outcome.

But you are playing with fixed population sizes. Now play the same game, but add infinite black cards to the discard pile and infinite red cards to the red pile.

Every time you loop, add one black card to the black deck from the discard pile.

Clearly, now you are far more likely to add a purebred black card into your deck than you are to roll sixes (or tails, or whatever your criteria are for reproductive success) for the red card.

It's still possible for red to win. In fact, in our initial case, where we were assuming red was always reproductively successful, red will still win: gen 1 it will go from 1 to 2, gen 2 it will go from 2 to 4, and so on, not always doubling (since sometimes reds will form pairs), but always increasing by more than 1.

But if you change "add one black card" to "add 50% of black cards" to represent a 50% population growth per generation, then it becomes impossible for red to win: even if we go for our "best case" above, there will always be more black added than red can hope to create on one generation.

Remember that there are two groups here, A and B - we can cherrypick whichever of the two most rapidly becomes all-hybrid,

TL;DR: Essentially, we're setting up an experiment with a population of As and ABs, where we guarantee that there is always a minimum level of ABs in the population by guaranteeing a minimum level of inbreeding with Bs.

Unless we grow the population faster than the ABs grow, there's just no way you can not end up with all ABs, eventually.

However, it might take a looong while, especially if social or biological effects make ABs less reproductively successful than As.

  • $\begingroup$ I'm not sure I understood your example perfectly but I think you ignored something. The genes that makes an individual A or B can be diluted and lost. Parent A and parent B have a child : AB (50%A, 50%B). AB has a child with another A, this child is 75%A, 25%B. This new individual has a child with another A, so the new child is now 87.5%A and 12.5% B. Continue like this for a few generation and B genes could almost completely disappear from the treeline. So as with humans, I don't think you'll ever have a good homogeneous population but mostly a very varied and continuous spectrum. $\endgroup$
    – Jemox
    May 13, 2019 at 22:41
  • $\begingroup$ @Echox: Given the title, I agree that other interpretations are possible, but I interpreted the question not as asking "would they be visually homogenous" but "would they all share a common ancestor". With this question, the family tree of a member of the A population either includes a B, or not: an indilutable boolean true/false, passed to descendants with 100% reliability. Could be 40 generations back, so only a trillionth of their heritage is B (and we only have 3 billion base pairs in our genome), but they're still descended from a B. $\endgroup$ May 14, 2019 at 13:47

Blue eyes, brown eyes. Make the different races like dominant and recessive traits.

Option 1: Your races (dwarf, elf, orc, human) as a whole are dominant/recessive. No matter how mixed up they are you will still have all of the races, though some may be more common and some more rare - about 75% dominant and 25% recessive.

Option 2: Certain traits from each race are dominant/recessive. They can get all mixed up but they won't really homogenize. You might have orc-ish humans or dwarfy-elves, but the basic traits all stick around even if they don't all stay together. This way you can have certain areas were some race-traits are more common or more rare depending on who's been mixing with who.

Option 3: Left handed people and color blind people represent a fixed percentage of the population. No matter who or where, about 10% of people are left handed. There's an evolutionary reason for this - left handed people have an advantage at winning sports vs right handed people (but rank about the same at everything else) and color blind people are better at seeing through camouflage (useful for hunting, but has obvious drawbacks). The genetic mechanics of this aren't well known yet as far as I know. You could tie races to something like this. Maybe 10% of people in human populations are always elves or orcs or whatever you like.



Two reasons:

1 Preferential mating

Many people prefer to marry within their group, whether it's religious, social, racial or any other subset of society. Whatever the origin and definition of the group it will slowly diverge. The separation is never 100% so there will always be mixed race people, but races as such will not disappear, unless you enforce mixing forever or start eliminating minorities.

2 Discrete traits

Many racial traits, like skin colour, pointed ears or allergy to sunlight are defined by small number of genes. Because of the way genes work this means that if you have a mixed race family (A+B) their children will usually look like a melt of their parents races (AB), this makes intuitive sense. But, children of two mixed race parents can show any extreme racial features, two mixed race parents may have one child that looks almost like pure race A, and the second child that looks mixed like the parents. Bear in mind that inside they are all definitely still AB not a pure race, but by statistical fluke they may look like A.

3 Result

Now because of the way society works (combine points 1 and 2) those pure looking mixed race kids are very likely to marry with the other people that look like them, maybe a pure A, maybe another mix like AC. Again, because of the way genes work, each of those mixed race kids that look like pure race A has about half of genes from race A, and half from another race (B, C, or D whichever was their other grandparent from). If you get a group of A-looking kids (AB, A, AC, AD, A, A) together in the next generation their kids will be almost pure A. The races will keep coming back.


The only way to homogenize a society is if the other races die out (naturally or otherwise) or are outbred by a large margin. Some of their genes will obviously remain, like we still carry some neanderthal genes, but neanderthals are gone.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .