Ask vs. Aks: Let me axe you a question

Do you know which one of these forms is the correct one? You sure about that?

Four things are inevitable: death, taxes, the eventual heat-death of the universe, and language change. All (living) languages are constantly in a state of flux, at all levels of the linguistic system. Meanings change, new structures come into being and old ones die out, words are born and die and pronunciations change. And no one, it seems, is happy about it. New linguistic forms tend to be the source of endless vitriol and argument, and language users love constructing rules that have more to do with social norms than linguistic reality. Rules that linguists create, which attempt to model the way language is used, are called “descriptive”, while rules that non-linguists create, which attempt to suggest how they believe language should be used, are called “prescriptive”. I’m not going to talk that much more about it here; if you’re interested, Language Log and Language Hippie both discuss the issue at length. The reason that I bring this up is that prescriptive rules tend to favor older forms. (An occasionally forms from other languages. That whole “don’t split an infinitive” thing? Based on Latin. English speakers have been happily splitting infinitives since the 13th century, and I imagine we’ll continue to boldly split them for centuries to come.) There is, however, one glaring exception: the whole [ask] vs. [aks] debate.

Axt zum spalten
In a way, it’s kinda like Theseus’ paradox or Abe Lincoln’s axe. If you replace all the sounds in a word one by one, it is the same word at the end of the process as it was in the beginning?
Historically, it’s [aks], the homophone of the chopping tool pictured above, that has precedence. Let’s take a look at the Oxford English Dictionary’s take on the history of the word, shall we?

The original long á gave regularly the Middle English (Kentish) ōxi ; but elsewhere was shortened before the two consonants, giving Middle English a , and, in some dialects, e . The result of these vowel changes, and of the Old English metathesis asc- , acs- , was that Middle English had the types ōx , ax , ex , ask , esk , ash , esh , ass , ess . The true representative of the orig. áscian was the s.w. and w.midl. ash , esh , also written esse (compare æsce ash n.1, wæsc(e)an wash n.), now quite lost. Acsian, axian, survived inax, down to nearly 1600 the regular literary form, and still used everywhere in midl. and southern dialects, though supplanted in standard English by ask, originally the northern form. Already in 15th cent. the latter was reduced dialectally to asse, past tense ast, still current dialectally.*

So, [aks] was the regular literary form (i.e. the one you would have been taught to say in school if you were lucky enough to have gone to school) until the 1600 or so? Ok, so, if older forms are better, than that should be the “right” one. Right? Well, let’s see what Urban Dictionary has to say on the matter, since that tends to be a  pretty good litmus test of language attitudes.

“What retards say when they don’t know how to pronounce the word ask.” — User marcotte on Urban Dictionary, top definition

Oh. Sorry, Chaucer, but I’m going to have to inform you that you were a retard who didn’t know how to pronounce the word ask. Let’s unpack what’s going on here a little bit, shall we? There’s clearly a disconnect between the linguistic facts and language attitudes.

  • Facts: these two forms have both existed for centuries, and [aks] was considered the “correct” form for much of that time.
  • Language attitude: [aks] is not only “wrong”, it reflects negatively on those people who use it, making them sound less intelligent and less educated.

This is probably (at least in America) tangled in with the fact that [aks] is a marker of African American English. Even within the African American community, the form is stigmatized. Oprah, for example, who often uses markers of African American English (especially when speaking with other African Americans) almost never uses [aks] for [ask]. So the idea that [aks] is the wrong form and that [ask] is correct is based on a social construction of how an intelligent, educated individual should speak. It has nothing to do with the linguistic qualities of the word itself. (For a really interesting discussion of how knowledge of linguistic forms is acquired by children and the relationship between that and animated films, see Lippi-Green’s chapter “Teaching children to discriminate” from English with an Accent: Language  ideology and discrimination in the United States here.)

Now, the interesting thing about these forms is that they both have phonological pressures pushing English speakers towards using them. That’s because [s] has a special place in English phonotactics. In general, you want the sounds that are the most sonorant nearer the center of a syllable. And [s] is more sonorant than [k], so it seems like [ask] should be the favored form. But, like I said, [s] is special. In “special”, for example, it comes at the very beginning of the word, before the less-sonorant [p]. And all the really long syllables in English, like “strengths”, have [s] on the end. So the special status of [s] seems to favor [aks]. The fact that each form can be modeled perfectly well based on our knowledge of the way English words are formed helps to explain why both forms continue to be actively used, even centuries after they emerged. And, who knows? We might decide that [aks] is the “correct” form again in another hundred years or so. Try and keep that in mind the next time you talk about the right and wrong ways to say something.

* “ask, v.”. OED Online. December 2012. Oxford University Press. 12 February 2013 <;.


What words are easy to say?

Ok, so in the last couple posts I’ve been throwing around terms like “easy to say” without giving a whole lot of explanation. And that’s a pity, because the study of what words are “easy” and what words are “hard” is, in my opinion, one of the greatest sub-disciplines in linguistics: phonotactics.

Imperial Russian soldier with phone
No, that's phone tactics, not phonotactics. They're completely different.
Phonotactics is like your great-aunt who always arranges the seating at family reunions becuase she remembers who fought with whom twenty years ago and knows not to sit them together. Basically, some sounds really like to be next to others. Like vowels. Vowels like to be next to everyone. In Japanese, for example, with a couple of exceptions, most syllables have to be made of a consonant plus a vowel. (In ling speak, this is known as “CV”. C for consonant, V for vowel. Yeah, unlike physicists, we like to keep things simple.) What’s even more amazing is that within six months of birth, Japanese infants prefer sounds that are CVCV to those that are CVCCV or CVCVC.

Polish, on the other hand, notoriously plays fast and loose with syllable structure. You can have consonant clusters up to five sounds long in Polish that, most weirdly, don’t follow the same sorts of rules that other languages do. Like English. English can have pretty big consonant clusters… but they’ll only get really big if the first or last sound in the word is ‘s’. (Protip: That’s why ‘s’ is such a great letter in scrabble; there’s a bunch of things you can slap it on to piggyback of someone else’s word, even outside of its morpheme status.) If you’ve ever stumbled over a Polish last name, there’s a sound linguistic reason you found it hard.

Why is this useful? Well, besides its obvious use in language teaching and being great cocktail party conversation material,  if you want to make a plausibly difficult-to-pronounce alien language, screw up your phonotactics and you’ll leave audio book readers in tears.