In Greek mythology, Cerberus—or, to use a spelling that is more faithful to the Greek pronunciation, Kerberos—is the ferocious, many-headed watchdog who guards the entrance to the underworld. Ancient Greek sources disagree on exactly how many heads he has. The Theogonia, a long narrative poem in dactylic hexameter composed by the poet Hesiodos of Askre in the late eighth or early seventh century BCE, says (in lines 308–312) that Kerberos has fifty heads. The later poet Pindaros of Thebes (lived c. 518 – c. 438 BCE) says in a fragment (F249a/b SM) that he has no less than one hundred heads. In Greek vase paintings, he is usually portrayed with either two or three heads. He is often also depicted with a serpent for a tail. From the fourth century BCE onward, though, three heads seem to have become the standard.
There is a popular meme that has been circulating on the internet for a long time claiming that the name Κέρβερος (Kérberos) actually means “Spot.” This is a cute story. How deliciously ironic it would be for a literal Hellhound to have such a normal dog name like “Spot”! Unfortunately, the story probably isn’t true. Kerberos’s name certainly does not mean “Spot” in Ancient Greek, and it probably doesn’t mean this in any language at all.
A very brief introduction to Indo-European linguistics
First, let’s get something out of the way: Κέρβερος does not mean “Spot” or “Spotted” in Greek, nor is there any word in Greek with this meaning to which the name Κέρβερος might be related. In order to understand where this whole supposed etymology for Kerberos’s name comes from, it is important to understand at least a little bit about Indo-European linguistics.
Ancient Greek is one of many languages that belong to the much larger Indo-European language family. Other ancient languages that belong to this same language family include Hittite, Proto-Celtic, Latin, Proto-Germanic, and (most crucially for our purposes) Sanskrit—the classical language of the Indian subcontinent and sacred language of Hinduism. All Indo-European languages, including Ancient Greek and Sanskrit, are ultimately derived from a common ancestor, which linguists have dubbed “Proto-Indo-European.”
The Proto-Indo-European language is not directly attested. There are no speakers of the language alive today and there is no surviving writing in it, but scholars can reconstruct some of the basic morphology, syntax, phonology, and core vocabulary of the language by comparing the various languages derived from it that are directly attested. Through careful comparative study, it is also possible to acquire some information about the original speakers of this language, who are known as the Proto-Indo-Europeans.
The most widely accepted hypothesis today is that Proto-Indo-European was most likely spoken in the steppes north of the Black Sea in what is now Ukraine and southern Russia, perhaps from around 4500 BCE to around 2500 BCE. Somehow or another, speakers of this language spread out across Eurasia, bringing gradually diverging forms of their language with them. These dialects eventually diverged into separate Indo-European languages.
ABOVE: Map from Wikimedia Commons illustrating the spread of early Indo-European languages, according to the most widely accepted version of the Kurgan hypothesis
A highly speculative etymological hypothesis
The Ṛgveda is probably the oldest surviving text in the Sanskrit language. It is generally thought to date to sometime between c. 1500 and c. 1200 BCE, making it roughly the same age as the standard Akkadian version of the Epic of Gilgamesh. Because it is one of the earliest surviving texts in any Indo-European language, the Ṛgveda is an especially important source for reconstructing the Proto-Indo-European language.
In the nineteenth century, some early western philologists who were trying to reconstruct the Proto-Indo-European language noticed that, in the Ṛgveda 10.14.10–12, the Sanskrit word शबल (śabála-), which means “various in color,” is used to describe the two dogs of Yama, the god of the underworld. They interpreted the word in this context to mean “spotted” and hypothesized that śabála- and several similar-sounding Sanskrit words with similar meanings might all be derived from a common Proto-Indic root word, which they reconstructed as *śárbara-.
These nineteenth-century philologists then further hypothesized that the hypothetical Proto-Indic root *śárbara- and the Greek name Κέρβερος might both be derived from a much older, common Proto-Indo-European root word. They reconstructed this hypothetical Proto-Indo-European root as *kérbero-. They hypothesized that this root might mean “spotted” and that it might be the name of the guard dog of the underworld in Proto-Indo-European mythology.
ABOVE: Photograph from Wikimedia Commons showing a relief carving from Angkor Wat in Cambodia depicting Yama, the god of death, with eighteen arms riding on the back of a buffalo
Scholarly objections to this etymological hypothesis
Although most scholars still believe it is likely that the Proto-Indo-Europeans believed that some form of dog or dogs guard the underworld, most of them no longer believe that it is possible to reconstruct a Proto-Indo-European root for the name Κέρβερος.
The German scholar Bernfried Schlerath dealt a major blow to the hypothesis linking Kerberos’s name with the Sanskrit word śabála- in an article titled “Der Hund bei den Indogermanen” (“The Dog among the Indogermans”), published in November 1954 in the journal Paideuma: Mitteilungen zur Kulturkunde (Bd. 6, H. 1, pp. 25-40). (This article is available on JSTOR, if you have access.)
Schlerath argues based on comparative mythological evidence that the Proto-Indo-Europeans believed that there was not one, but two dogs who guard the underworld: one of them white and the other black. He further contends that the use of the word śabála- to describe the two dogs of Yama in the Ṛgveda 10.14.10–12 does not mean that both dogs are spotted, but rather that the two dogs are of different colors (i.e., that one is black and the other white).
On top of this, Bruce Lincoln, the Caroline E. Haskell Distinguished Service Professor Emeritus of History of Religions at the University of Chicago, raises several serious linguistic objections to the “spotted” hypothesis for the etymology of Κέρβερος in his book Death, War, and Sacrifice: Studies in Ideology and Practice, published in 1991 by the University of Chicago Press, on pages 96–98.
Lincoln points out that any attempt to connect Κέρβερος with Sanskrit śabála- necessarily involves some pretty dubious linguistics on the Sanskrit end—namely collapsing an entire class of Sanskrit adjectives into a single Proto-Indic root form. He also points out that, in Proto-Indo-European, the phoneme *b is extremely rare or perhaps even nonexistent, but the supposed reconstructed root *kérbero- includes an internal *-b-, which automatically makes it highly dubious.
Finally, Lincoln points out that the putative reconstructed proto-meaning of “spotted” comes solely from Sanskrit. Some scholars have tried to link the supposed root *kérbero- with the Slavic word со́боль (sóbolʹ), meaning “sable,” and the Old Irish word corbaim, meaning “soiled” or “stained,” but these words are quite distant in meaning from “spotted.”
Lincoln goes on to try to link Kerberos’s name with that of Garmr, the dog of Hel in Norse mythology. As we will see in a moment, though, Lincoln’s efforts are no more convincing than those he dismisses and his proposed etymology has not caught on.
ABOVE: Illustration made by the German painter Johannes Gehrts in 1889 depicting how he imagined the goddess Hel, alongside her guard dog Garmr
The Dutch linguist Robert S. P. Beekes rejects the “spotted” hypothesis for the etymology of Κέρβερος in his book Etymological Dictionary of Greek, published in two volumes in 2010 by Brill, noting that the word śabála- in Sanskrit may, in fact, be of Austroasiatic, not Indo-European, derivation. Beekes further dismisses Lincoln’s hypothesis that Κέρβερος might be related to Garmr as “unfounded speculations.”
Daniel Ogden, a professor of ancient history at the University of Exeter, discusses and rejects the “spotted” hypothesis for the etymology of Κέρβερος in his book Drakōn: Dragon Myth and Serpent Cult in the Greek and Roman Worlds, published in 2013 by Oxford University Press, in a footnote on page 105. Ogden further rejects Lincoln’s hypothesis linking Κέρβερος to Garmr, noting that this explanation requires deriving the two names from completely different Indo-European roots: namely, Κέρβερος from the root *ker- and Garmr from *gher-.
Thus, the general consensus among linguists and ancient historians for the past seventy plus years or so seems to be that Kerberos’s name does not come from a Proto-Indo-European root meaning “spotted.” Sadly, this means that Hades did not, in fact, name his guard dog “Spot.”
ABOVE: Photograph from Wikimedia Commons showing Side A of a Caeretan black-figure hydria by the Eagle painter dating to c. 525 BCE, depicting the hero Herakles bearing a version of Kerberos with three heads and snakes all over his body
I imagine the reason why Cerberus was depicted with two or three heads in Greek art is because it was easier than trying to depict him with fifty or a hundred heads.
That is certainly possible, although it is also possible that the tradition of Kerberos with two or three heads may, in fact, be the older one, since Hesiodos and Pindaros might have exaggerated the number of heads for poetic effect. The way Hesiodos says “fifty” and then Pindaros comes along and says “one hundred” makes it almost seem like a sort of competition, with each poet trying to make Kerberos out to be a bigger, meaner, scarier monster than the poet who came before.
Well sourced, and clarifying on a difficult matter.
Thank you! I’m glad you found this post informative!
Hi Spencer
My doubt is not related to this article and comes from a previous one (https://talesoftimesforgotten.com/2021/11/08/ganymedes-the-adolescent-boy-zeus-abducted-and-raped/)
In the article, you mentioned Plato depicting sexual activity as a distraction, except for procreation. And Plato not indulging in sex for procreation purposes as well. Could it be that Plato was asexual? Do we have any evidence to support or reject the statement?
I do think it is possible that Plato might have been asexual, but it is impossible for us to know whether he was or wasn’t. The problem is that Plato wrote almost nothing about himself, and he certainly never wrote anything about whether he personally experienced sexual desires or attractions. For my own part, I personally suspect that Plato was not asexual and that he did experience sexual desires, but deliberately repressed them and did not act on them.
The main reason why I suspect this is because, in Plato’s Symposion, Socrates describes attraction to the bodies of beautiful boys as the first and lowest step toward love of the idea of Beauty itself. Later in the same work, Alkibiades describes how, when he was a young man, he made multiple unsuccessful attempts to seduce Socrates. He describes these attempts in a way that makes it sound like Socrates was very much tempted to have sex with Alkibiades, but he resisted the temptation through his strength of will. The Symposion therefore makes it seem like sexual attraction to adolescent boys and young men is natural and expected, but not an attraction on which a man should act.
Obviously, it is dangerous to make inferences about Plato’s own sexuality based on his portrayal of Socrates, but I think that Plato’s portrayal of Socrates in the Symposion would have the most resonance with a person who had a repressed sexual desire for adolescent boys and young men.
Thank you for clarifying that!!
Speaking of repressed sexual desires, what do you think of Bishop Spong’s theory (in “Rescuing the Bible from Fundamentalism”) that Paul of Tarsus repressed his own attraction to men (as when Paul writes of an unnamed wound/weakness that he carries)?
Personally, I’m not really convinced by Bishop Spong’s theory. I don’t think there’s really any good evidence in Paul’s case to suggest that he was sexually attracted to men.
There is a significant difference between how Plato writes about same-sex attraction and how Paul writes about it. Plato writes in the Symposion as though he thinks it is perfectly normal and expected for men to be attracted to other males. Then, in his Republic 3.403b-c, he portrays Socrates as arguing that, in the ideal state, the erastes should only be allowed to touch and kiss the eromenos innocently, as a father might touch a son, and should not under any circumstances be allowed to have intercourse with the eromenos. Finally, in his Laws 1.636b-d, he portrays his Athenian speaker as denouncing same-sex intercourse as immoral and unnatural. All this evidence put together makes it seem like Plato believed that same-sex attractions are normal and it is only immoral to act on them. Paul, on the other hand, writes in Romans 1:18-32 as though he thinks same-sex attraction on its own is disgusting and unnatural.
The fact that Paul never married or had children cannot be used as evidence that he was sexually attracted to men either. Firstly, Paul describes his religious reasons for not marrying at length in 1 Corinthians 7. If we take Paul at his word, then he may very well have been what we would consider completely straight and may have simply chosen not to act on his desires for women because of his religious convictions. Secondly, even if we assume that Paul was not attracted to women (which we can’t necessarily assume), there’s no evidence to suggest that he was attracted to men. After all, it is entirely possible that Paul may have been asexual.
It is true that Paul does say in 2 Corinthians 12:7, “. . . a thorn was given me in the flesh, a messenger of Satan to torment me, to keep me from being too elated,” but the context of this line strongly suggests that the “thorn” he’s referring to are the various persecutions he has faced for the sake of Jesus. Just a few lines later, in 2 Corinthians 12:10, he writes: “Therefore I am content with weaknesses, insults, hardships, persecutions, and calamities for the sake of Christ; for whenever I am weak, then I am strong.”
Thank you! I’m not worried about Paul’s orientation one way or another (though I’d be fascinated to know more about his psychology). It’s just refreshing though (these days) to read an argument based on actual evidence and familiarity with the source material. Again, thank you for doing what you do.
You’re breaking my heart, Spencer!! But thank you for another fab piece which I look forward to reading.
You’re welcome for the article! And I’m sorry for breaking your heart; breaking anyone’s heart is certainly not my intention!
Marvellous. Still, now that we know what Cerberus doesn’t mean, have we any idea of what it does mean?
Unfortunately, we don’t really know where the name comes from, much less what it actually means.
Robert S. P. Beekes says that the name might be of Pre-Greek origin, but he says that about basically every name in the Greek language that there is no generally accepted etymology for, so I’d take what his opinion with a grain of salt.
Joel P. Christensen, a professor of classical studies at Brandeis University, has collected various proposed etymologies for the name on his blog Sententiae Antiquae, but I’m not sure that I find any of them particularly compelling.