Phylogeny of elves finds that santa’s workers are actually dwarves


This study reconstructs the evolutionary tree of elves using 26 life history, morphological, behavioral, and magical characters. Notably, we include christmas elves, J.K. Rowling’s elves, and the elves of The Lord of The Rings and The Hobbit. Our findings suggest that christmas elves should not be classified as elves but are actually more closely related to dwarves.

Abstract (Elvish)

abstract in elvish










Elves were among the first beings to inhabit terrestrial ecosystems. Living species are present in a variety of habitats (Figure 1), including: toy work-shops, the forests of middle earth, tree bakeries, and domestic dwellings.

Figure 1: Elves thrive in a variety of habitats. We wonder how they adapted to such disparate ecosystems.

Being a highly charismatic group, scientists have shown great interest in reconstructing their evolutionary history. However, abundant hybridization,  very very very long generation times, mutations due to magic, and corruption by dark forces are all factors that have confounded previous attempts to reconstruct their phylogeny.

This study provides the first phylogeny of elves using characters of life-history, morphology, behavior, and magical ability.


We chose 15 mythical creatures for our study. This included 9 elves and 6 outgroup species. Due to patent restrictions, no tree-dwelling cookie producing elves were included in our sample.

Initial attempts to obtain genomic data from elven tissue was unsuccessful. Due to funding restrictions we could not obtain a lysis buffer that would digest an energy field of pure starlight. This was greatly disheartening considering how difficult it was for us to get permits to destructively sample living elf populations.

In the software Mesquite, we assembled a matrix of 26 categorical characters (Table 1 – see the bottom of the page) from various literature sources. This character matrix was used to construct a strict consensus of the 100 most parsimonious trees (SPR rearrangement algorithm). The tree was rooted such that gnomes, fairies, dwarves, trolls, and orcs were all contained in the outgroup.


Our resulting tree (Figure 2) has excellent resolution, despite having a large amount of incomplete data for the Ljosalfar elves and Dokkalfar elves of Nordic lore. Large amounts of incomplete data are expected for extinct species, such as these.

Figure 2: Strict consensus most parsimonious tree of elves

The three Tolkien elf species are most closely related to the Ljosalfar (light elves) of Nordic mythology, from which Tolkien conceived his elves. The Drow elves of the Dungeons and Dragons universe were found more closely related to the Nordic Dokkalfar (dark elves) despite their purported derivation from the Tolkien elves. However, we believe this could be error due to convergent evolution of the arcane arts. The blood elves of the World of Warcraft were the sister group to all other “true” elves.

Two surprising results were (1) we obtained the Christmas Elves (American mythology) and (2) the House Elves (Rowling mythology) with the outgroup. House elves are indeed more closely related to the common ancestor of true-gnomes and fairies and not any of the “true” elves.


Previous studies have theorized that selective breeding and “perversion” of elves by the dark lord Monsanto Morgoth was the origin of orcs. Our phylogeny clearly shows that orcs are more closely related to trolls and not elves.

Although more extensive taxonomic revisions are necessary we propose a number of new classifications.

  1. House elves (Rowling) should be considered house gnomes, and fairies are a highly evolved form of gnome that have achieved flight. We use new gnomeclature to refer to the Gnome+Fairy clade as Pixies.
  2. Christmas elves are in fact a type of dwarf, whose skills in stone craft have been adopted for toy making. Ancient Christmas elves probably made primitive stone toys.
  3. J.K. Rowlings gnomes, although having affinities for gardens, may in fact be distant relatives of orcs and trolls. We hesitantly suggest this, however, since this species has a large number of highly derived traits and secondary losses, which could potentially confound the analysis.

Conflict of interest statement

The authors of this work report no affiliation with: the Lich King, Sauron the terrible, Freya, or Led Zeppelin. The authors do not have any rings of power.


This study was funded by the NISP (National Institute of Sprites and Pixies).

About the author

Dominic Evangelista is an NSF post-doctoral fellow studying the phylogeny of cockroaches (not elves). Follow him on twitter @Roach_Brain or ask him a question about cockroaches!

Table 1: Character matrix used in the phylogenetic reconstruction

Table 1: Character matrix used in the phylogenetic reconstruction

Works Cited:




41 comments on “Phylogeny of elves finds that santa’s workers are actually dwarvesAdd yours →

Comments are closed. You can not add new comments.

  1. Your classification of the Christmas Elf as the nearest relative of the Dwarf is, in retrospect, less surprising than the use of the word “Elf” would lead one to believe. It has long been verified that the primary activity of the Christmas “Elf” is the mass crafting of intricate and durable playthings. This is highly reminiscent of the Dwarf compulsion to hand-craft durable but non-exquisite weapons and armor. It would not be surprising were further work to reveal that the two populations are more closely related than eveb your analysis would indicate, the apparently large dimorphism between the two being induced by the remarkably hospitable conditions under the North Pole when compared to the near-constant warfare and rigor Middle Earth.

    1. Maybe some consideration could be given to the relationship between dwarves and elves from a mythology not represented in the current phylogeny, i.e. that of The Ring of the Nibelung (Der Ring des Nibelungen) by Richard Wagner. Here, the Nibelungs, a race of dwarf-like miners and jewellers, are interchangeably referred to as ‘Zwerg’ (Dwarf) or ‘Albe’ (Elf). Indeed, the eponymous Nibelung Alberich, the forger of the Ring, has the term for “elf” in his name. The Wanderer (Wotan) refers to the Nibelungs as ‘Nachtalben’ (night elves), but to the Gods as ‘Lichtalben’ (light elves), hinting at a common ancestor to both groups, which would make the Gods and Nibelungs part of a monophyletic group.

  2. Sir,

    Teleri elves did possess a simple for of telepathy. This is most famously recounted in the historical documents surrounding Elu Thingol of Doriath. When confronted with the man Beren, he spoke to elves around him mentally before breaking the silence with haughty, and later regretted, words. This may alter your data.

  3. Didn’t Treebeard say at one point that Trolls were debased from Ents, as Orcs were from Elves? The Silmarillion clearly states that Morgorth created Orcs by kidnapping, torturing, and otherwise altering them until the result: Orcs. I think the Trolls were the same thing done with Ents he captured. Consider:

    Trolls (in Tolkien) were extremely tough, stone-like creatures with immense strength. Ents were extemely tough, stone-like creatures with immense strength. The main differences were:

    A. Appearance (Ents look like trees; Trolls look like… well, ugly boogers, but actually look more humanoid than Ents).

    B. Ability: Both can crush or crumble stone in their hands (though Ents seem to do it better).

    C. Resistance: Ents have a phenomenal resistance to piercing weapons, and good resistance to blunt force. Their main weakness is fire (being somewhat like trees, that is logical). Trolls have great resistance, but not as much as Ents: Pippin stabbed a troll in its vitals. Morgorth seems to have sacrificed a goodly amount of the Ent resistance when he “mutated” them into Trolls… possibly because he shifted them toward humanoid appearance (one wonders: did he breed mutated Ents with Orcs to get Trolls?). He also reduced their intelligence: Ents are obviously quite intelligent creatures, but Trolls are known for low intellect. This would actually be a plus to Morgorth, because his will could more easily dominate the Trolls. And though Trolls don’t seem particularly vulnerable to fire, at least some breeds *are* vulnerable to sunlight… to the point that they “return to the stone they were made from,” becoming petrified when exposed to light.

    Both Ents and Trolls are very earth-related creatures: Ents to trees, soil (and manipulating rock), and Trolls to rock.

    Another possibility is that Morgorth did a crossing between Ents he’d corrupted and *Dwarves* he’d corrupted. This actually strengthens the affinity to rock (Aule shaped Dwarves from stone, and they work stone and minerals almost exclusively). Yavanna, Aule’s wife, created the Ents to be “shepherds of the trees” and guardians for them, after Aule created Dwarves. In terms of the demigods (the Valar) who created these two species, the demigods are from the same general sphere: Yavanna over trees, plants, and animals; Aule over earth, stone, and minerals. *And*, if Morgorth crossed corrupted Ents with corrupted Dwarves, it would give him the more humanoid form that Trolls have (vs. Ents), and the affinity to stone, without sacrificing too much toughness (Dwarves are renowned for their toughness and endurance).

    So, I think the idea that Trolls descended from an elven prototype ancestor (in Tolkien, anyway) is a bit of a stretch, and doesn’t take into account the possibilities that Morgorth derived them from Ents, and possibly cross-bread the result with Dwarves (or maybe Orcs, though that’s less likely) to get Trolls.

    One other thought: there *does* seem to be quite a bit of variance among Troll types: some turn to stone in daylight and have very low intelligence, while others resist sunlight and may or may not have greater intelligence. This *might* be accounted for if Morgorth crossbred corrupted Ents with Dwarves AND Orcs, and then divided the results into groups and limited their breeding to within a group, to strengthen certain characteristics that were already prominent in the original results of his manipulation.

    1. Oops. The “them” in the first paragraph should have been “Elves” instead. “…Morgorth created Orcs by kidnapping, torturing, and otherwise altering Elves until the result: Orcs.”

    2. What your describing is a great example of something we don’t have when we do phylogenies of real world organisms. In dealing with Tolkien’s beings we have extensive documents about their creation (although I have a thought about this near the bottom of this comment). This means we have prior knowledge about what the relationships “should” be. In the real world we have no such prior knowledge (most times) and have to take phylogenies at face value since we have no way of getting into a time machine and watching species evolve over millenia. Many scientists have great faith in the relationships shown in our phylogenies, and discussing how they are made and how they can be made better is a HUGE field. However, the community has decided that any given phylogeny should just be considered a hypothesis about what our available data say about the relationships among these organisms. Again, because we don’t have time machines we can’t verify the relationships.

      Back to elves…in this case we DO have a time machine, because we can go back and reread the Silmarillion and the Lost Tales. As such, we may decide that the information in those books represents the “true” relationships. If we do that then we can indeed disprove part of the hypothesis that is the phylogenetic tree.

      Finally, my thoughts about Tolkiens documents and your comment. 1. Just because Treebeard purports that this is the origin of the trolls doesn’t necessarily mean that he is correct. The inhabitants of earth (even the highly informed ones) have many misconcenptions about the origins of things on our own planet. It is possible that treebeard is subject to the same error (although I would agree that his age gives a great deal of knowledge to call upon). 2. As every Tolkienite knows he was not always consistent across books (e.g. balrogs, glorfindel). And I believe there are some conflicting statements about the origins of orcs. Please correct me if I’m wrong.

      1. How does your tree change if you allow the possibilty of lateral transfers? You might end up with a branch of generic “Evil things”, from which the Orcs and Trolls come via lateral transfer of Elvish and Entish traits respectively.

        1. It assumes lateral transfers do not happen. This is one limitation of most phylogenetic tree making methods. Phylogenetic networks are alternative methods of inferring evolutionary histories. We do not use them for most organisms bc lateral transfers are rare overall.

      2. Treebeard may have made a mistake, that’s true—but Ents are among the oldest living things, and set great store in their oral tradition. Many societies with oral traditions tend to be very exacting in how oral history is passed down, making it a faux paux to alter even the smallest word, lest the information become corrupted over time. Given that Ents are also very long-lived (and there’s hints that they’re immortal unless killed—or they go so “treeish” that they never wake up again), it lends further credence to Treebeard’s account.

        As to the origin statements, I was taking directly from the Silmarillion’s account of the birth of the Eldar… which was written by the Eldar who went to Valinor, with at least some of the information gleaned from the Valar themselves (who *did* the creating. While it’s possible that the Valar withheld information (in fact, likely), and possible that the Elves themselves included errors or misunderstood something (they knew where they first awoke, and that some of those first Elves disappeared and that there was a shadow of fear upon them (which later they learned was Morgorth at work, and his minions)… I think they must have gotten the information from the Valar, that their kidnapped bretheren were tortured and twisted into Orcs. IIRC, Gandalf (who was the Maiar Olorin, and was also around for the beginning) backed this up at some point.

        However, I’m a Tolkien *lover*, not a Tolkien *scholar*—so I may be remembering parts of this wrong, or attributions… and I haven’t read much of the Lost Tales. My info is based on the Silmarillion and parts of LotR. 😀

        As for Tolkien errors… quite possible that they were genuine errors, though he may have deliberately created slightly different versions to portray how information can be altered over time or by different groups, etc.

        By the way, this has been enormous fun for me: a lively fun discussion on subjects not political, social, or real world. It’s like a vacation. 😀

        As to the real life science end of this, I’m not very versed on this area in science, so that’s been fun, too. Thanks.

  4. Gnomes sister taxa to orcs and Goblins, hog wash I say – see edited volumes by Rein Portvliet on the characteristics of gnomes.

    1. Ahh but these are the Gnomes as described in the Harry Potter universe. These gnomes are much more ill tempered and seem to be quite stupid…which may have influenced why I found them in the “Evil beings” clade

      I had trouble finding a source that gave a lot of information about gnomes. This is why I used David the Gnome as there was information about him and he seemed to be a good Gnome representative.

  5. Given that the Vanyar are the light-elves that remained in Valinor, it’s definitely wrong to classify them as dark-loving. Noldor – debatable. On the one hand, they journeyed to Aman along with the Vanyar and so both groups count as Calaquendi (light elves). On the otherhand, the Noldor were the ones that delved into the earth and carved gems to refract light. Perhaps leave that one as ambiguous.

    Teleri, fair enough to call them dark-loving – most of them refused the journey to Aman and are classed as Moriquendi (dark elves).

    1. I refer to them all as dark loving because in the age before the sun was made the world was in darkness, which was the state that the elves preferred the world to be in. Although they love specific lights (starlight, moonlight, the light of the trees, the light of the silmarils) their preferred world was a world of night.

      I have only read the silmarillion once so perhaps I’m missing or crossing up some things. Please correct me if I am wrong.

  6. The difference between faeries and pixies is largely good intent vs evil intent rather than a difference in morphology. Pixies will hide cups or forks until just after you have set the dishwasher going or hide your reading glasses. Faeries will put your car keys where you can find them or place a coin down the side of a sofa when you desperately need it.

  7. I believe you have left out one of the most recent ELFs, not known in the plural as elves, and always in caps. That would be an Organic Transit ELF, These are a vehicl e known as Electric, Light, Fun (ELF) and have many traits a s elves. These ELFs are native to Durham, NC, USA. Hoever, they have migrated not only across the country to many states, but now reside in many other countries as well. And their European cousin is an Evovelo,

  8. All three of Teleri, Noldor, and Vanyar possess in theory the ability of osanwe, communication or interchange of thought, which sounds very close to the concept of telepathy. As J. R. R. Tolkien’s essay “Osanwe-kenta” (read it for free here tells us, osanwe became less used over the years, due to a number of factors including the development of tengwesta (language) and the fact that Incarnates such as Elves (unlike the Ainur – Valar and Maiar – whose spirits, fear, are not united with their bodies, hroar) require strengthening factors (such as affinity, urgency, or authority) for the transmission of thought to be effective. Nonetheless, we see osanwe applied a number of times throughout the legendarium by a number of Elves, including but not limited to Galadriel (one of the Noldor), Celeborn (one of the Sindar, who are of Telerin descent), and Finrod Felagund (also of the Noldor).
    Tl;dr: Tolkien Elves are telepathic. It’s in the obscure depths of canon.

    1. Your comment and comments by others are great examples of the importance of organismal expertise. Many people that do phylogenies of real animals do not have expertise in the animals that are actually in those phylogenies. As such, their data, or their interpretation of the tree, may be lacking because of their lack of expertise. I am fairly knowledgable in the lore of LOTR but am definitely not an expert. As such, my data was flawed. Given the wonderful suggestions by all the people who have commented I could redo the analysis. Maybe for next christmas 🙂

  9. Great job. I would also suggest to get DNA samples to reconstruct phylogenetic relationship though sequence.

  10. It would be interesting to add Finnish kotitonttu ( ‘house elf’) or at least Nordic tomte to this comparison. From Finnish perspective it was not surprising that Christmas elf was more related to dwarfs than to elves. From our perspective it is neither, but its morphology is similar to dwarfs. In Finland it is commonly believed that Christmas elf is derivative of tonttu or tomte. In old believes ‘kotitonttu’ protected house and was kept satisfied by giving it food etc. I believe that Nordic ‘tomte’ was similar, but I am not sure.
    Here is illustration about tonttu:

  11. There appears to be a slight bug in the matrix, as “intelligence” for fairy is listed as “yes”, which doesn’t appear to be one of the categories. I wonder whether a correction of this to “average” (?) affects the phylogeny.

    1. Thanks for catching that! The version of the martrix you see in the blog is not the one I used for the analysis. I had to break each category down into a binary state (i.e. 0 or 1) or a categorical state (e.g. 1, 2, 3…). During this stage I found that error and fixed it…it’s just an error in the blog. The analysis is unaffected…**phew** **wipes sweat off brow**

  12. Given how poorly elvish ears are adapted to the arctic climate, they are obviously not natives.

  13. Probably best to leave out the warcraft species as they have lore that says that Dark Trolls were actually the precursor form, of which the *five* elf species evolved after exposure to the well of eternity. You also incorrectly label blood elves as high elves. Blood elves didn’t even take that as name a moniker for their nobility such as the the night elves did.

    Five warcraft elven species.
    Kaldorei, the night elves (those of aristocratic birth took the name Quel’dorei for themselves)
    Shal’dorei, the nightborne
    Quel’dorei, the high elves
    Sin’dorei, the blood elves
    San’layn, the darkfallen

    There is also some debate as to whether Fell blood elves are a separate species based on either physical or magical mutation.

    While the Undead consider themselves a group apart because of their inability to breed, they still retain primary racial characteristics of their birth species.

  14. As a long time admirer and friend of Gnomes – I resent the implication that gnomes are “evil creatures” . Rowling’s portrayal of gnomes is inaccurate and based on long held cultural biases towards the community. There is a reason they are referred to as the ‘Forgotten Folk’.
    Yes, many Gnomrs are subterran and prefer parital or total darkness such as the Svirfneblin; while this quality may make them seem elusive, confusing and even threatening they are mostly very peaceful and tend toward good and neutral alignments. Most account of Gnomes posit them as highly intelligent creatures, often endowed with illusory magic, witty and charismatic. Some a roguish types and tricksters exist but many more seek knowledge and go into scholarly professions including design and engineering.
    You may wish to consult D&D5e – paying particular attention to the Gnomes of Faerun; Pathfinder – d20pfsrd and the Gnomes of Galorian; you may also wish to consult some of the Shadowrun sources which give account of Gnomes as a more neotenic relative of Dwarves.
    I appreciate that this is the first study of this kind to seriously evaluate the evolutionary relationships between these races but the limited use of widely available material is disheartening. A follow up is in order.

  15. Consideration of the Dark Elves of Norse mythology as revealed through the Marvel Cinematic Universe (Thor: The Dark World, 2017) would have greatly improved this data set lending to their placement as the appropriate stem group for Elvin phylogeny.

    1. I would have to say this group is an outlier data set specific to the Marvel universe, as it goes counter to just about all other elven origin stories, thus making them immune to placement as a phylum within the larger whole, and therefore must be left out, or given their own separate taxonomy.