“Dog” is a weird word

Where animal names come from

Dog and a Birch Log, Govert Dircksz Camphuysen

England, AD 1050

The hunters gather at dawn, the raw chill biting at any exposed skin. Mist hangs heavy among the ancient oaks, covering the ground like a shroud. But the mist does nothing to muffle the sharp crackling of frozen leaves underfoot.

Suddenly, the peal of a horn comes from somewhere to the left. It’s a signal. The men ready themselves.

It’s not long before they hear it. A crashing in the undergrowth, too loud for a deer. It bursts into the clearing: a massive wild boar, its black bristles standing on end, framing two wicked tusks of dull ivory. Small eyes glare out with hatred, promising that this will be no easy fight.

But the hunters have weapons of their own: a pack of lean, powerful hounds. One man gives the word, and they spring into action. In a whirlwind, the hounds surround the boar, snapping and growling. But the boar is ready for them. It lowers its tusks to meet the charge of the lead hound. The hound twists in mid-air to avoid the boar’s tusks and narrowly escapes, unscathed.

Another hound lunges at the boar’s flank and barely avoids the same fate. The boar is just too fast. The air fills with a horrible medley: the grunts of the boar, the yelps of the hounds, and the shouts of the hunters as they encircle their prey.

For a while, the battle is deadlocked. The boar cannot break through the cordon of hounds and hunters, but the hounds can’t get close enough to take the boar down. Then, suddenly, it happens: two hounds feign a headlong attack, drawing the boar’s attention. As it commits, another pair darts in from the sides.

“Hie on, hounds!” cries one of the hunters. He’s seen this before. This is how you take down a boar: with cunning, courage, and a team of stout-hearted hounds.

But, farther back, near the sputtering campfire, where the attendants await the hunters’ return, a different scene is unfolding. A scrawny, mud-caked thing with wary eyes snatches at a piece of discarded gristle.

A boy shoos it away with a stick, “Begone, dog!”

This is no hound. It’s as far from the sturdy warrior animals who took down the boar as a mangy barn cat is from a lion.

No, this is no hound. It’s just a dog.

But somehow, this word dog — a word that wasn’t even dignified enough to write down once in the entire history of English writing before the eleventh century¹ — would come to be the normal English word for man’s best friend, almost displacing the older form hound.

Stranger still, we don’t know for sure where the word dog came from.

We do have some theories, though. But I’ll need the help of some pigs, some frogs, and even some earwigs to explain.

---

You're reading The Dead Language Society. I'm Colin Gorrie, linguist, ancient language teacher, and your guide through the history of the English language and its relatives.

Subscribe for a free issue every Wednesday, or upgrade to support my mission of bringing historical linguistics out of the ivory tower and receive two extra Saturday deep-dives per month.

You’ll also be able to join the very first Dead Language Society summer book club(!), where we’ll be reading Beowulf together. We’ll compare translations against the original Old English and explore the historical and cultural context of the poem. Along the way, I’ll explain all the deep lore that most readers miss.

Our first meeting will take place on Thursday, July 3, at 3:30pm Eastern. Paid subscribers will receive an email invite in the coming week!

---

Ain’t nothin’ but a…

A hunting dog with a rabbit, Anton Enzinger (c1683–1768)

If we look at the languages most closely related to English, we would expect the word for ‘dog’ to be hound. German has Hund, Swedish has hund, Dutch has hond — all clear relatives of hound. Even Old English has the word hund, as we can see in this line from the Exeter Book riddles:

Hwílum ic beorce swá hund

‘Sometimes I bark like a dog’ (Riddle 24)

Because we have so many examples of words like hound in the Germanic languages, we can reconstruct the word *hundaz in Proto-Germanic,² the unwritten ancestor language of all the Germanic languages.

In fact, relatives of this hound word are even found in Latin (canis), Sanskrit (श्वन् śván), and Ancient Greek (κύων kúōn). This wide distribution means we can reconstruct it back even farther, all the way to Proto-Indo-European, the ultimate ancestor not only of English, but also most of the languages of Europe and many of Asia as well.³

So hound is a very old word. And yet, by around AD 1500, it had been largely replaced by the word dog. Hound does survive in some compounds (hellhound), fixed expressions (release the hounds!), and as a more specific term for hunting dogs, as in the Hound of the Baskervilles (which wasn’t a shih-tzu). But the generic term for a domestic canine today is decidedly dog.

What’s particularly strange about this replacement is that (a) we don’t have any earlier forms we can derive it from, and (b) it came almost out of nowhere.

Other than a few instances of dog-like words in Anglo-Saxon land charters, where it’s used to describe property boundaries or in place names, for example: doggene ford ‘dogs’ ford,’ doggene berwe ‘dog’s hill,’ doggiþorn ‘dog thorn,’ the only place we see a dog-like word in Old English is in a gloss from around AD 1050 to a Latin text, the Liber Peristephanon.

But this appearance gives us an important clue as to how dog defeated hound, and why.

Glossing is where you take a text in another language and write word-by-word equivalents in your own language. Reading a gloss is halfway between reading a translation and reading the original language. Here’s an example of a Modern English gloss of an Old English text:

Hwílum    ic beorce swá hund 
sometimes I  bark   as  dog

The Anglo-Saxons were very interested in reading Latin, largely (but not exclusively) because Latin was the language of Christianity for them. As a result, they made lots of glosses, explaining Latin texts word-by-word in terms of their Old English equivalents.

This gloss from Liber Peristephanon is where we find the first attestation of the ancestor to dog outside of place names:

                                          docgena 
si iam tuorum perspicis languere virtutem canum 

‘If now you see the courage of your dogs falter’ 

(Boulogne gloss to Liber Peristephanon)

In this gloss, docgena glosses the Latin canum ‘of dogs.’ But what’s interesting is that this text isn’t literally talking about dogs. It’s using the Latin word canis as an insulting term for people: the term is used by a victim of persecution to refer to his tormentors.

And this derogatory use continues in early Middle English (12th–13th centuries): most of the time, when a dog word is used as opposed to a hound word, it’s used as an insult:

David i þe sauter cleopeð hine dogge… þe fule cur dogge.
‘David in the psalter calls him (the devil) ‘dog’... the foul cur dog.’ (Ancrene Wisse, c. 1230)

Over the following centuries, dog gradually broadened in its use until, by the 16th century, dog was the standard word, as it is today.

The insulting origins of dog may hold the key to its ultimate origins. But to understand why, we’ll need to look at its form: in particular, that -g at the end of the word.

---

Good doggy

Landscape with Pointer and Setter, Eugene Petit (1839–1886)

The form which we saw in the gloss above, docgena ‘of dogs,’ implies a base form docga. The -cg- combination would have been pronounced as a doubled ‘g’ sound.⁴ And it’s precisely this doubled ‘g’ sound that makes the word docga so strange: according to all we know about the distant prehistory of Old English, this is not a sequence that should have ever developed.

In other words, knowing what we know about all the ancestors of Old English, going back all the way back to Proto-Germanic, there’s no sequence of sounds that should have created this doubled ‘g’ in Old English.

And yet there it is.

More than that, docga isn’t even the only one of these double ‘g’ words. There are others, in which the double ‘g’ sound is spelled either gg or cg indiscriminately: frogga ‘frog,’ picga ‘pig,’ hogga ‘hog,’ stacga ‘stag,’ and the element wicga ‘insect, wiggler’ in earwicga ‘earwig.’⁵

Not only are all of them animals, but most of them also have murky origins. And they all have something else in common too: the ending -a. This ending⁶ is not uncommon in Old English in general, but one place you very frequently see it is in diminutives and nicknames: cute or familiar variants of names. For example, Torhthelm gives Totta and Beornfriþ gives Beoffa.⁷

Some of you may have heard of a place where one particular Totta once lived: Tottenham, which comes from Tottan hām ‘Totta’s home.’

At any rate, note that Totta and Beoffa also have doubled consonants before the -a ending! This doubling or lengthening of sounds in emotionally charged words is sooooo common across languages that it has a technical term: expressive lengthening. And nicknames certainly count as emotionally charged, since they so often express familiarity or affection.

If these -gga names are formed by expressive lengthening, it also gets around the problem I mentioned earlier: that there’s no sequence of sounds that could have naturally developed into -gg- in Old English. Expressive lengthening just lengthens what’s already there, so it provides a way for a pre-existing -g- to become -gg-

So perhaps these -gga names for animals are also nicknames for the animals that the Anglo-Saxons lived their lives alongside, like thousand-year old versions of doggy or puppy.

If that’s the case, we’d expect to see some evidence of what the nicknames are based on.

Usually, but not always, nicknames are formed by modifying an existing name: Nicky from Nicholas, Robby from Robert, etc. Sometimes, of course, there are cases like Bob from Robert or Peggy from Margaret, where the pathway from the name to the nickname isn’t as clear, but most often we can tell.

More evidence for this theory comes from the fact that the earliest uses of picga were to refer to piglets rather than full-grown pigs, which is exactly what we’d expect out of a diminutive nickname: to be applied first to the young of the species, which, after all, are the cutest.

It turns out that one of the -gga words does have a clear base form: frogga ‘frog’ almost certainly is a variant of the older term frosc ‘frog’ (the ‘sc’ is pronounced like ‘sh’). German still has the related word Frosch for ‘frog’. If a frosc needed a nickname, frogga would do nicely!

So if the -gga words were all originally nicknames for animals, the mystery becomes: what were the base forms of these other -gga words? What were the words that the nicknames came from?

---

‘Dog’ isn’t the only weird word

The Sultan, A Greyhound (1838), J. Thompson

The quest for the source words for these animal nicknames still continues, especially for dog. One theory⁸ that I really like points to the Old English word dox, which is found in glosses next to the Latin words furva ‘dark, dim’ and flava ‘golden, yellow.’ Reconstructing what historical colour terms meant is tricky, but since there are both dark and golden dogs, dox is a plausible source for docga either way.

Perhaps dog started as a term for a particular, despised kind of dog: given that most of the earliest uses outside of place names are derogatory, this stands to reason.⁹ Nicknames aren’t always affectionate! But, with time, the word dog lost its negative charge. It’s as if the word mutt were gradually to become the normal word for ‘dog’.

Interestingly, the replacement of hound by dog is not the only time a descendant of the original Indo-European word has been replaced in the Indo-European languages: Spanish perro ousted can in exactly the same way dog replaced hound. Likewise, Irish madra succeeded cú (which you still see in the name of the legendary hero Cú Chulainn ‘Hound of Culann’). Finally, Russian собака sobaka (partially) replaced пёс pyos as the generic word for dog, restricting пёс pyos to the meaning ‘male dog’.¹⁰

Why do words for these familiar animals, especially ‘dog,’ keep getting replaced? Here we can only speculate. But I suspect it’s precisely because our linguistic ancestors were so familiar with dogs and other domestic and household animals (even earwigs are, in an unfortunate way, household animals) that they ended up acquiring some emotional charge: even today, people have a hard time resisting making diminutive forms for the dogs in their lives: hence the terms of the internet of yesteryear pupper, doggo.

And even when these nicknames start out derogatory, the fact that the nicknames exist at all betrays some emotional involvement on the part of the namer. And derogatory names can easily be applied ironically to something beloved. After enough time has passed, even a once-unkind term can become entirely generic.

What do you think? Is the mystery of dog solved? What animal nickname do you think will turn into the accepted term a thousand years from now? Let me know in the comments.

1. Words ancestral to dog do appear in place names in a few earlier land charters (e.g., doggene ford ‘dogs’ ford’), but we don’t see uses where it clearly refers to the animal until later. More on this below.

2. The asterisk means the form is reconstructed rather than attested anywhere in writing. The -az ending in Proto-Germanic corresponds to the -us ending found in many masculine nouns in Latin, including names like Marcus Aurelius. This ending drops out in most Germanic languages, but remains in Old Norse -r and Gothic -s.

3. The reconstructed Proto-Indo-European form is *ḱwṓ. The sound reconstructed *ḱ becomes k-like sounds in some branches of the Indo-European family (such as Italic, hence the Latin canis), and s-like sounds in other branches (such as Indo-Iranian, hence the Sanskrit śván).

4. IPA [gg].

5. Technically, picga should be written *picga, as the word itself is unattested in Old English: what we have is a compound picbred ‘pig-bread.’ The word *stacga is also unattested in Old English, but it’s attested in Early Middle English (c1185) in the form staggon, which suggests an Old English *stacga/stagga.

6. For nerds: this is the ending for masculine weak nouns.

7. This argument is from Piotr Gąsiorowski in his 2006 article The etymology of Old English *docga.

8. Gąsiorowski (2006) again.

9. We see the use of ‘dog’ words as insults even today, e.g., bitch.

10. The origin of пёс pyos is disputed: it may not have come from the Indo-European *ḱwṓ. But, ironically, its replacement собака sobaka actually did! It’s a loanword into Slavic from an Indo-Iranian language, in which the initial *ḱ became an s-.