Let’s play the pluralization game.  Which of the following is the “correct” plural of octopus?  Which are acceptable variants?  Which are “wrong”?

  • octopuses
  • octopi
  • octopodes

The English-speaking world is divided into about five categories of people here:

  1. People who think octopuses is best because it comes naturally, like children think foots is an acceptable plural of foot, instead of feet.
  2. People who like octopuses because it sounds like the Bond film Octopussy.
  3. People who use octopi because it seems like the way we deal with words in -us, like alumnus/alumni.
  4. People who use octopodes because they think it has a better historical basis (to be explained below).
  5. People who think octopodes sounds silly and pedantic, and octopi is wrong, so they settle on octopuses.  (Many dictionary and style guides go this route.)

Generally speaking, most people probably progress through these categories over the course of their lives.  At first, they chose octopuses, because English words ending in -s form a natural English plural with -es. Then, at some point, someone corrected their use of words in -us to employ the common Latin plural of -i, so they learned octopi either by being corrected from octopuses or else by assuming that octopus would follow that rule.

People who care a little more about words will eventually run into further trouble, though.  They learn that Latin is more complicated, and octopus is apparently from Greek, hence octopodes is supposedly more correct than octopi.  At this point, whether a person joins category (4) or (5) usually has to do with personality.  Someone who wants to be perceived as erudite even if it appears arrogant probably uses octopodes (with the confidence that he or she knows more than most people), while pragmatic people return to the common English plural octopuses.

The problem is — all these people are wrong in some way except for those in category (2).  (Categories (1) and (5) are also close, but the reasons are a little misguided.)  The only undisputed plural for octopus is octopuses.  If you think you know more than I do and want to claim validity for one of the other forms, read on.

The word octopus is not an ancient word from Greek or Latin.  Ancient Greeks called the animal a polypous, and Romans adopted the word from Greek, Latinizing it into polypus.  This word comes from the Greek roots meaning “many feet” and could refer to any sea creature, probably including octopuses, squid, and so forth.  There was a word in Ancient Greek oktapous (with an a), an adjective meaning “eight-footed,” but it did not refer to the sea creature we now call an octopus.  (Ancient Latin similarly had a poetic adjective octipes, derived from the Latin  octo- “eight” + -pes “foot.”  Note the variation between Greek -pous and its Latin variant -pus compared to the original Latin -pes.)

So where did we get our modern word octopus?  Carolus Linnaeus, the father of modern biological taxonomy, coined it in the mid-1700s on the basis of the Greek roots and by analogy to Latin polypus.  Note by this time that Latin was rarely spoken outside of church functions and traditional schools.  Linnaeus, however, had to coin a lot of new words for his classification system for all animals and plants, so Latin and Greek roots seemed the obvious place to start.

The late origin of octopus is important, because we actually don’t have a historical precedent to refer to when it comes to deciding the plural.  We can’t know what the Romans would have used, since they didn’t use that word at all.

All we can do is make assumptions about Latin grammar.  Latin has five declensions, which are five different patterns that nouns follow when deriving their grammatical forms.  Although many English speakers assume that all words ending in -us must form a plural with -i, that is only true of words in the Latin second declension.  Words in -us, however, could occur in the second, third, and fourth declensions.  The fourth declension, for example, formed its plural merely by making the last u into a long vowel.  Hence, the Latin word prospectus had a plural of  prospectus, with a long u.  In English, the proper plural is not prospecti (which would be pseudo-Latin) and thus can only be prospectuses, since English does not generally adopt the Latin fourth declension plural.  (Try saying prospectuuuuus sometime when you want to mean the plural, and see how far that gets you, even among very educated people.)

The third declension is more tricky.  Generally speaking, there are a lot of patterns lumped together into the third declension, so you can’t predict what the plural will be based on the singular in Latin.  The word octopus, if it actually existed in Ancient Latin, would arguably fall into that category.

In that case, it would be treated as a word borrowed from Greek, and it would adopt a form close to the Greek plural, hence octopodes.  Contrary to what some grammarians assert, octopus is not a Greek word (which would be octopous or oktopous) but rather a Latinized Greek word.  Educated Romans knew how to handle Greek words, much as English speakers know how to deal with Latin plurals, so octopus would become octopodes in Latin.

So octopodes is the correct plural, right?  Most pedantic grammarians like to point to that form.  It’s true that it relates well to the biological order that the octopus is classified within, namely Octopoda.  And it agrees with the English word octopod, which joined the English language in the early 1800s, clearly showing that people knew the Greek roots of the word.

But, unfortunately for all those pedants out there who love to trot out octopodes and love explaining its origin to display their erudition, it’s still not that simple.

Remember — octopus was coined after Latin was essentially a dead language.  It was still used in many scientific writings, and it was spoken in increasingly obscure contexts, but Linnaeus was able to make use of novel Latin and Greek words in part because they could be differentiated from the “common name” of the animal or plant, which by that time would have generally been known in the local vernacular language.  How can we decide the “proper” form of a word coined in a dead language?

Obviously many pedants think we follow the standard rules for Latinized Greek words, which in this case leads to octopodes.

Unfortunately, these “rules” were followed very inconsistently even in Classical Latin, and octopus itself poses a particular problem.

Recall where Linnaeus got the word — he produced it by an analogy to the Greek/Latin word for the animal, namely polypous/polypus.  (Perhaps he was aware of the incredibly rare Greek adjective oktapous, but that form seems irrelevant, since he adopted a different spelling.)  But there’s one problem here — the accepted plural for polypus in Ancient Latin, Medieval Latin, and even the Latin of Linnaeus’s time is not polypodes but polypi!

That’s right — polypus is a specific example of a place where the Romans broke the “rules.”  Even Linnaeus himself uses polypi as a plural.  But the Romans weren’t using this form out of ignorance.  Even the Greeks didn’t know the proper forms of polypous; some treated it as third declension (hence polypodes), while others treated it as second declension, which would lead to a Latin form polypi.

So, Linnaeus comes along and he takes polypus, which is a generic word that could refer to any number of legs, and he turns it into a specific word for eight, namely octopus.  But then he uses the Greek plural octopodes rather than octopi, even though octopi would make more sense based on the word polypi from which octopi was derived.

All you pedants out there take note — octopodes is absolutely not an undisputed plural of octopus according to the “rules” of classical languages.  In fact, it violates the precedent of the very word it is derived from, namely polypus/polypi.  We can argue that the Ancient Romans were “wrong” with their own language, but this is not the only irregular plural by any means.  Variations in Greek plurals were incredibly common, and the “rules” distilled in our modern Latin grammars are only guidelines that were violated quite often.  In this particular case, the guidelines were even violated in Greek itself, since polypous did not clearly belong to a specific declension.

The only people who decided that octopodes was more correct than octopi must have been from the 1700s or after, a time after spoken Latin ceased to be a living evolving language for the most part.  Thus, there is actually no undisputed classical precedent for what the “proper” plural must be in Latin (or Greek, for that matter).  All we have is Linnaeus’s and his followers’ inconsistent use of polypi with octopodes, which also led to octopi occurring in English almost as soon as octopus became an English word, possibly because of its similarity to the Latin polypi.  From the standpoint of classical languages, the people using octopi in the 1700s had just as sound a basis for their choice as the octopodes users.

Where does this leave us today?

Well, octopus has one undisputed plural, namely octopuses, which is the native plural using -es.  As far as my recommendations go, this is the only plural you should ever use, because, like my earlier example of prospectus, there actually is no undisputed way to form any English plural in this case that would be analogous to a Latin or Greek form.

For those who use octopodes, stop being so smug.  You have the “rules” of Latinized Greek words on your side, but as I’ve pointed out, it is not at all clear that those rules should be obeyed in this particular case considering the derivation from polypi.  Claiming that octopodes is the only correct plural would be like claiming that foots is the only correct plural of foot because it follows the English “rules.”  Obviously there are words in Latin, just as in English, that don’t follow the simple “rules.”  Since octopus is a recent coinage, the question of what “rules” it should follow is not clear-cut.

For those who use octopi, be aware that you are too on shaky ground.  You are going against the “rules,” and your form goes against the historical precedent set by the guy who coined the word himself, who clearly preferred the standard Greek form of the Latin plural.  But now you can defend your form against those octopodes pedants, who also have a questionable foundation for their preference.

In the end, just use octopuses.  If someone questions you, say you use it because it’s funny and sounds close to Octopussy.  If they start turning into a pedant and endorsing octopi or (heavens!) octopodes, you now have the knowledge to show them they can use what they want.  But they have absolutely no basis for calling your octopuses wrong, since there are plenty of historical reasons why their forms are also potentially wrong.  In almost all cases, the native English plural is never an incorrect choice.  Use your -es with pride!

NOTES:

This is in part a response to a video that has been circulating recently of a Merriam-Webster editor whose historical information is a bit shaky and whose conclusions about proper usage (use any of them!?) seems ill-conceived.

Also, for those who would like more guidance on plural issues in general when dealing with the many declensions of Ancient Greek and Latin, this is a great reference.  (Don’t be afraid of the title of the page; it really is a grammar article.)