Category Archives: language and linguistics


“It’s not the scenicest day,” I said to Aina, looking out the train window at a cloudy sky as we headed to Niagara for some wine and walking.*

Or perhaps I should spell that scenic-est, so you know I wasn’t saying it like “see nicest,” even though what is scenicest is nicest to see.

“Is that a word?” you may be thinking – or perhaps typing in an email to me. Well, I used it and you understood it, so yes. But is it a well attested word? No. You can find a couple hundred hits for it on Google, but it’s a safe bet most of them are – as I was – self-consciously using it as an awkward construction rather as Lewis Carroll used curiouser.

Why wouldn’t I just say most scenic? Because I like playing with words. Now it’s your turn: Tell me why scenicest shouldn’t be allowed. It’s a two-syllable word, after all, and it’s quite common to append –er and –est to one- and two-syllable words. The selection of those for which more and most are reserved is almost random-seeming. At the very least, the distinction is not black and white. For some words, it is a matter of personal taste which to use: beautifuller and beautifullest were formerly common enough, but now it seems we see the two-word version as the more beautiful.

I do think that what we see is part of the problem here. For assorted historical reasons (mostly to do with palatalization before front vowels in Latin and Romance languages), c “softens” before e and i. But the sound /k/ does not have an actual allophonic alternation with /s/ in modern English. We just retain the rule about c because of our borrowings from French and Latin. This makes a problem when we have something that sounds fine but runs into a spelling issue. Take chic. Lovely word, stylish, smart. Borrowed from French. By borrowed I mean adopted – actually I mean stolen. Anyway, it’s treated like an English word: it’s one syllable, so instead of saying most chic we often just add the –est and make it chicest.

Which looks horrible on the page. And chic-est looks at least as bad. And you can’t add or swap in a k because chikest would look completely wrong and incomprehensible and would conduce to yet another inaccurate pronunciation, and chickest is chick plus est. Somehow the chicest word to say is one of the unchicest (let’s say least chic) words to write.

Well, what do we expect? It should be supercalifragilisticexpialidocious?

Am I the only one who feels certain that supercalifragilisticexpialidocious should be two words? Normally, morphologically, we can add only other suffixes after a suffix, not a whole new root, let alone a prefix plus a root plus a suffix. And yet that’s what appears to come after the the ic in supercalifragilistic. Another bit of evidence to marshal for its being two words is that the spelling would seem to require a pronunciation like “–listi sexpi–,” which is clearly wrong.

Which takes us back to our problem of the orthographic scenery. Now, –ic words often used to be spelled with a k, as in musick and magick. So could we borrow on that and make it scenickest? Hmm. It looks a bit of a snickerfest. It may also tempt a person to shift the accent onto the second syllable because of the “heavy” consonant ck.

Or we could just keep using it and writing it and people will get used to seeing it and saying it. That’s how a lot of things in English have come to be as they are.

We ought not to be distracted by looks, anyway. A cloudy day may be warm and lovely. Indeed, when the sun is out and it looks most scenic, you are at greater risk of getting burned.


*It was not a reference to the fact that we would not be taking in a play at the Shaw Festival, even though scenic referred to the stage a century before it referred to the natural environment – it comes from a Greek word for a stage.

A whole nother thing

Originally published on The Editors’ Weekly, the national blog of Editors Canada

As editors, we pay attention to the written form of our language. Its relation to the spoken form is a whole other thing.

The spelling is odd, we know. But even our hyphenation doesn’t really break according to pronunciation. Consider the word breaking. Where do you hyphenate it? Break-ing. But where does the syllable break happen in pronunciation? Before the k. Don’t believe me? Shout it, emphasizing each syllable equally. You’ll shout “Bray! King!” rather than “Break! Ing!” In speech, we automatically shift a consonant at the end of one syllable to the beginning of the next if there isn’t a consonant there already, regardless of how the word is formed. But in writing we reflect the bits the word is made of, because that’s how we think of it.

Except when we don’t. And then it’s a whole nother thing.

Take a word like another. It’s made of two pieces: an and other. Put them together, and the n is automatically said at the beginning of the second syllable, so it sounds like a nother. You’d think we’d still keep it in mind that it’s really an other. You would not inevitably be right.

A whole nother isn’t the only place we’ve done this redivision. Centuries ago, a newt was an ewt and a nickname was an ekename. And speaking of nicknames, Ned and Nan come from mine Edward and mine Ann (we used to alternate my and mine as we still alternate a and an), and in Shakespeare you’ll see nuncle in place of uncle.

It also goes the other way. A poisonous snake, in English, was næddre, which would normally have become nadder, but instead of a nadder we have an adder. Likewise, a naperon gave us an apron. We pronounce them no differently, unless we put another word in between, but we think of them differently. We hear the n said at the start of the next syllable, but since the n in an always does that, we reanalyze it in a way that seems – for one reason or another – more appropriate.

Do you wish you could have someone to make rulings on these kinds of resplittings (also known as rebracketing and false splitting)? Try calling an umpire – as long as you don’t mind that your umpire would once have been a noumpere.

This doesn’t mean that we have to accept a whole nother, of course. A whole other is considered formally correct, although that implies a two-word an other. Since we’ve glued the two parts together, putting whole in the middle is arguably more like what we do in abso-bloody-lutely… except we wouldn’t write a-whole-nother.

Perhaps we should reconsider what we do and don’t think of as inviolable word boundaries. We may dislike alot quite a lot, for instance, but if we can make a word like another, can you think of a truly defensible reason for it not to be another one such? Or is that a… completely different issue?

The old “ye olde”

Originally published on BoldFace, the blog of the Toronto branch of Editors Canada

If you want to make something look, y’know, old, and classy and stuff, what’s better than adding an e to the end of it? Think how much extra you pay to stay in a Crowne Plaza hotel than you would in a simple Crown Plaza. Cochrane, Alberta, has a log-and-glass event space called Cochrane RancheHouse. And of course there are all these plain old olde things.

And that’s where we cranky up the antiquity another notch, with the word you have to blow the dust off every time you use it: ye. As in ye olde candy shoppe. And As ye sow, so shall ye reap. And perhaps, at the RancheHouse, ye haw.*

Only those aren’t the same word. And the ye in ye olde isn’t ye at all. The y isn’t y.

English used to have the letters ð and þ, which stood for sounds we now spell as th as in this and thin. They mostly fell out of use during the medieval period, but a few words often kept them, such as þe (the) and þat (that). They would be reduced with the aid of superscripts, like þe and þt. But when we got printing presses, the moveable type that came with them was forged on the continent by speakers of languages that didn’t use those letters.

What was the closest letter? You might think it would be p or b, but the way þ was written in cursive was more open topped and looked like a rakish y with an ascending first line. So y became the substitution, and the was often rendered as ye (often with the e right on top of the y). This became so well established it was done that way even in hand-carved inscriptions such as tombstones, where the carver could have used a proper þ – if he had known to do so.

So. Ye as in ye olde is really just the. But how about hear ye and so shall ye reap? This is part of what causes the confusion. The ye in ye olde might be their problem, but the ye in hear ye is you.

Literally. It’s the old nominative form of you. Just as we have I and me, and she and her, we had ye and you. It happens to have fallen out of standard use over the years, gone from normal discourse by the time Shakespeare died, and gone from formal discourse before Churchill was born, but persisting in regional dialects. It’s as if the formal standard had come to be “Me gave it to her, but her didn’t want it.”

Well, ye can still keep it if ye want to be olde style. And, now, what is up with all those e’s? Well, Old English had a lot of inflectional endings that wore down over time. They included such suffixes as –an, –en, and –um. These ended up reduced to an unstressed vowel during the Middle English period. The spelling of English was in flux at the time, and scribes and, later, typesetters could make decisions about what letter to use to represent this minimal vowel. At times they used y or i, but in the end e, the easiest one, prevailed. And over time it stopped being pronounced, too. So we got all those silent e’s that make e the most common letter in English usage.

And when you’re a scribe paid by the letter, or a typesetter who needs to make the text fit the line, and these e’s are silent and seem to show up in random places, why not toss in extra ones here and there? And so a word that in Old English was eald and came with the changes of the time to be auld and aud and awd and old – and many other forms – could not avoid being olde occasionally.

Between then and now, some advocates of tidying up English spelling have had some minor success, and one of the things they prevailed in was removing most of the unetymological e’s from words. So all those unnecessary oldes became good old olds again. But when we want something to look old (and perhaps therefore classy), we herd towards that little mark of antiquity, the easy e. We see old (but, except for in Scots, not auld or awd) and shoppe (but not schopp) and crowne (but not croun or crowune). And we see ranche, which really was spelled with that e at times in the 1800s in spite of coming from Spanish rancho.

And what about the missing space in RancheHouse? Aw, that’s just branding. You know, as they do to cattle at ranch houses.


*Calgarians, please do not write to me telling me that it should be yahoo, not yeehaw. I know. I was making a funny.

In case you’re wondering, it’s a callow mistake

A recent cartoon from Randall Munroe’s xkcd, which is the most intelligent cartoon in existence, has been brought to my attention. You can see it at but, since Munroe gives permission, I’ll reproduce it here for ease of reference:

This interpretation is not just pedantry and not just a turn-off. It’s a callow mistake.

It’s callow because it transgresses standard expectations of interpersonal decency in conversation – if the other person is using a common turn of phrase and you understand what they mean, don’t be a dick about it (or, as I put it in a haiku for the ACES Grammar Day haiku contest, Which do you prefer: / keeping your friends’ grammar right / or keeping your friends?). But it’s also callow because it imposes an inappropriate misreading, the sort of simple-minded overbroad application of a rule that is characteristic of an immature understanding of grammar. The clause in question, you see, only looks like a conditional.

I’m put in mind of an anecdote I recall from the actor Simon Callow. When he was first trying to make it in theatre, he worked in the box office of a theatre. Later on, when he was becoming established as an actor, he was in the cast of a play that happened to be performing in the same theatre. One evening before a performance, one of the box office staff saw him and, thinking he still worked in the box office, asked him to come help with some box office function.

Thinking this if you want to hang out is a conditional is like thinking Simon Callow still works at the box office. It’s the same clause, but it’s been elevated. It’s a sentence adverbial.

In case the term isn’t familiar, a sentence adverbial is a word or phrase that, within the ambit of a verb phrase, could serve to modify the action of the verb, but that is instead applied to the entire sentence to frame it within a discursive context such as the attitude of the speaker or writer towards the utterance. “Frankly, Scarlett, I don’t give a damn” uses Frankly to mean ‘I speak to you frankly and say…’; “Among other things, this book explores the concept of silliness” uses Among other things to position the following statement within a larger possible set of observations (‘There are various things this book does; one of them is that…’); “Going forward, we’ll do it this way” uses Going forward to mean ‘I am making a prescription that applies to future instances when I say…’. They do not mean, respectively, that Rhett doesn’t give a damn frankly but he may give one covertly; that the book only explores silliness when the book is among other things; or that we will do it this way only when we are progressing ahead. And In case the term isn’t familiar doesn’t mean the preceding applies only in the case where the term is unfamiliar. It means I’m saying it in anticipation of the possibility of unfamiliarity.

Because sentence adverbials use words and phrases that can be used to different effect at lower levels, they are like candy to immature minds who are eager to pounce on other people’s “errors” to show their superior knowledge. But, as with so many rigid “rules” propounded by people who claim to care about grammar but really care mainly about demonstrating superiority, the “pedantic” interpretation is founded on a simple-minded misunderstanding. We have no difficulty understanding the sentences as they are intended – the pedants don’t even have the excuse that the box-office employee (who evidently didn’t read the programmes) had. The most they can argue is that the sentences are ambiguous. That can be something worth fixing, but it’s not a grammatical error. And they’re not always ambiguous, either. We usually understand them with no risk of confusion.

To add another analogy: When I was in New Zealand, I rented two different cars on separate occasions. In New Zealand they drive on the left, and so some of the driver’s controls are also the reverse of what I’m used to. With the first car, I managed to get used to the turn signal being on the opposite side from what I expected. Then when I rented the next car it was a model with the turn signal on the North American side. So I had to get used to it again and not keep turning on the wipers when I wanted to turn left or right. But in all of that, I did not say that the controls were wrong and I was right and stick to my preferred sides. I did not insist on turning on the wipers when I wanted to turn off the highway – or on signaling a turn when I wanted to wipe the windshield – because those were the correct sides for the controls to be on. I did what worked. When my expectations did not correspond with the results, I corrected my perspective. Which is what those who care about understanding language must do if they do not wish to be wrong.

So this pedantry is both a turn-off and a callow error.

Which, of course, Randall Munroe knows. He also knows that linguistics isn’t his area of expertise, and I’m not going to hold it against him for missing the analysis. He’s not the only one.

More honoured in the breach or the observance?

Originally published on The Editors’ Weekly, the official blog of Editors Canada

It is tempting to say that getting classical quotations right is more honoured in the breach than the observance. But if we did, we’d be guilty too. In the original, Hamlet is telling Horatio about the tradition of drinking sprees in the Danish court; he says it makes Danes look bad to other nations. So when he says

But to my mind, though I am native here
And to the manner born, it is a custom
More honour’d in the breach than the observance

he doesn’t mean they don’t do it; he means they shouldn’t do it. Honour’d here means ‘honourable’, not ‘complied with’.

Sometimes our errors come from shifts in culture. In a time when fires were the main source of heat, for instance, a fire that burnt bright but didn’t give off much heat was not much use. So when Polonius advises Ophelia to watch out for the ardor of young men (such as Hamlet), he uses this metaphor:

I do know,
When the blood burns, how prodigal the soul
Lends the tongue vows: these blazes, daughter,
Giving more light than heat, extinct in both,
Even in their promise, as it is a-making,
You must not take for fire.

These days, we use light as a positive metaphor in conversation, and heat more often as a negative one, so people often say a topic generates “more heat than light” – quite the reversal from the original.

We may look on such misinterpretations and say “Now is the winter of our discontent” with cultural knowledge. But we would be stopping short; here’s the whole opening sentence of Richard III:

Now is the winter of our discontent
Made glorious summer by this sun of York;
And all the clouds that lour’d upon our house
In the deep bosom of the ocean buried.

So he’s saying that their unhappy period is now made happy by the new king. (Admittedly, the man who is saying this is not happy about the state of affairs.)

Sometimes we just get a word wrong, perhaps because another word seems to go better with it (and another author, perhaps). We play our cats to sleep and say “Music has charms to soothe the savage beast” – and think it’s from Shakespeare – when the original is from William Congreve’s play The Mourning Bride, and it’s “Musick has charms to soothe a savage breast.”

Now, why not use a popular variation when appropriate, right? But we’re editors, and part of our job is to keep writers from looking bad, which means we have to take a do-or-die approach to quotations. Or, um, well… Tennyson’s original in “The Charge of the Light Brigade” is as follows:

Theirs not to make reply,
Theirs not to reason why,
Theirs but to do and die

Do and die? Perhaps we would do better to quote Yoda: “Do or do not. There is no try.” Just as long as we get the wording right.

Assimilation by the mutants

This article was originally published on BoldFace, the blog of Editors Toronto.

Every so often, someone asks, “If it’s one foot and two feet, and one tooth and two teeth, why isn’t it one book and two beek? If we have louse and lice, and mouse and mice, why not house and hice? If more than one goose is geese, why isn’t more than one moose meese?”

The answer is that the feet, teeth, lice, mice, and geese have been assimilated by mutants. And there’s more, so much more. It involves men and women; it involves our food. If you tell the tale, you too have been assimilated; if you try to heal, you find that the mutants have taken over even there. You cannot escape the strength of the mutants—nor their filth. The only thing you can take consolation in is that it was much worse a thousand years ago.

What are these mutants? Mutated forms of words, subject to i-mutation. A form of assimilation also called umlaut. You recognize that term, umlaut? It is sometimes used to refer to the two dots over ü and ö and ä (and a few other letters if you’re dealing with the names of heavy metal bands). But originally—and still—it refers to what those dots signify: a vowel pushing up and forward in the direction of the i sound (not as in Modern English “long i” but as in what i stands for nearly everywhere else, the sounds it makes in machine and prison).

Why does the vowel push up? Is it an uprising, a prison break? No: it is an assimilation. Welcome to the machine. Here is how it works: a word has a vowel that is low or back in the mouth (or both), and then—just on the other side of a consonant—it gets a suffix with a vowel that is high in the front of the mouth (i, generally, though ü could do it too). And that i in the suffix, that secret agent for the mutants, exerts a mysterious force on the vowel that was already there in the root. The root vowel wants to be more like that i. It moves towards it.

It’s not really such a mysterious force, actually. It’s just economy of effort. (“Laziness” is what your grandma probably called it.) If your tongue is going to have to be up there anyway, why not get there a little sooner? If you think about it, you’ll realize we do this all the time with all sorts of letters. For example, we move the n in think back to the same place as the k.

So where is the suffix-causing mutation in all these words: feet, teeth, geese, and so on? It’s long gone now. We lost many suffixes over time. But a long time ago, they were there. And the vowels moved towards them. And then moved some more. There was fót (foot), plural fóti, which moved the o forwards to assimilate and become fœti. Then it lost the i—and the agent of assimilation disappeared! And, over time, the œ unrounded and became a long e. And then the Great Vowel Shift occurred and long e moved up and became just like the old long i. The imitation of the lost i was complete. The mutation took over. It did not affect the singular, but there is danger in numbers: where there are two or more, the mutation takes over. Footfeetgoosegeesetoothteethman, menbook,beek.

Only not beek. But in the past it was! Yes, in Old English, a millennium ago, book pluralized the same way: bóc (the form of the word at that time) became béc. But between then and now, it regularized to books. Many other words that had this mutation also regularized. The mutation is curable, you see. It has to be taught at each new generation, in fact: all those parents saying, “It’s not foots. It’s feet.” (There was never hice, though; it wasn’t in the noun class that the mutant agents infiltrated. And moose was taken in the 1600s from a North American language, so it missed the mutant plague altogether.)

But wait! There’s more. Plural nouns are not the only things subject to i-mutation. Think about strength, length, and filth: they’re formed from strong, long, and foul (yes, foul, which in Old English was fúl). How does –th cause this assimilation mutation? (For clarity, I’m using th in place of the old runic-derived character þ that was actually used.) It was once, a long time ago, –ithu. So strongithu became strengithu, which became strength when the provocative agent i disappeared (and so did u).

There’s even more. Some verbs formed from nouns or adjectives, and the verb ending had—you guessed it—an in it that disappeared once it had done its job. So fóda (food) plus the –ian suffix became fédan (feed). The same thing gave us tell from talefill from fulldeem from doom, and even heal—originally hælan—from hál, the source of modern hale and whole. Causative verbs could also be formed from the past tenses of other verbs: for instance, drincan (drink), past tense dranc, got this mutating agent on it to make it drencan (drench). It’s even where we got lay from lie…but the difference between the past-tense læg and the causative lecgan disappeared over time and they merged as lay.

In fact, the larger part of these Old English i-mutants was neutralized by mergers. Remember that Great Vowel Shift I mentioned? In many cases, vowels and diphthongs that were different in Old English ended up sounding—and even being spelled—the same in Modern English. Our old noun léoht (light) and mutant verb líehtan, and thurst (thirst) and thyrstan, and weorc (work) and wyrcan—and oh, so many more—have come back together. Others (like beek) were lost due to regularization, and still others were lost because we just don’t use those words anymore—no frófor and fréfran, meaning “comfort,” noun and verb, respectively. But for the most part, the mutants that were formed by phonology were neutralized by phonology; the power that created them destroyed them. And no new mutants are being created…well, not of this kind, anyway.

This article was copy edited by Savanna Scott Leslie.

Sentence fragments? If you like.

As I sometimes do, I guested into a friend’s online copyediting course as a grammar expert for a week recently. One of the questions I answered was about whether “If you like” is acceptable on its own in any context. The questioner felt that in a conversational context it was acceptable (“Shall we leave at noon?” “If you like.”). Another student said that it doesn’t work because there’s an if but not a then. I said the following:

There are a few important things to remember.

First is that there are many kinds of English, suited for many different situations. To insist on standard formal English in all contexts is like wearing formal wear every day all the time. To use formal English in colloquial contexts doesn’t bespeak class and elegance; it bespeaks tone-deafness and rigidity. Rules are made to serve communication, not vice versa. Get to know the kind of English that is expected and used in each context you’re writing for. The point of editing is to make sure that the text produces the desired effect on the readers. Your job as an editor is to minimize the impedance in the circuit between author and audience. This often involves fixing infractions of rules, but not always. Indeed, sometimes the way to signal the tone of the text is to break a formal rule.

Second is that even in formal standard English, there are many things that are matters of preference, not rules.

Third is that not everything you do with language is a matter of grammar. Spelling mistakes, for instance, are not grammar errors. Neither are malapropisms. They’re errors, but they’re different kinds of errors (and in fact are the kind you can make sure to fix everywhere regardless of the tone and audience).

So, for instance, if as in “I want to know if you’re in town” is not a bad habit you need to cure yourself of for once and for all. The colloquial use of it where whether is the formal standard is very well established, and for some texts using if in place of whether will be the sort of little adjustment you can make to make it seem more relaxed. Bear in mind that “If you’re in town, I want to know” is acceptable even formally (the then can be and often is left out), which means that in the same sense, “I want to know if you’re in town” is also formally acceptable to mean “Let me know in the case that you are in town.” This, I believe, is why it has come to be used in the other sense, “Let me know whether you are in town or not.”

A very common mistake made by people who are eager to be right about grammar is to infer an absolute rule from one case, or to take a rule as learned overtly and apply it too broadly, declaring many common usages to be wrong because they don’t fit it. This is like pulling out a field guide to birds, looking at the picture of a magpie, deciding that all magpies must exactly resemble the exact specific colouring of that picture, and declaring any that don’t not to be magpies (and perhaps repainting or just killing them).

The effective approach is to read widely, see what kind of usages are common in what kind of contexts, and figure out the real rules on that basis. Often the real rule is not so simple and clean-cut; some things that are perfectly standard formal English still provoke arguments among linguists as to their actual grammatical structure.

To address a specific question: “If you like” by itself is a “sentence fragment” because it uses a subordinating conjunction (“If”) without a main clause to be subordinate to. But that doesn’t mean it can’t be used; we use sentence fragments all the time (I won’t say “The more, the merrier,” but if I did, it wouldn’t have a verb!). Only in the starchiest of contexts is it necessary to avoid starting sentences with conjunctions such as But and And, and in those cases only because some people in the past decided to repaint the magpies. In conversation, it is quite normal to leave out established material, especially in responses: “Shall I join you?” “[You can join me] If you like.” In more formal texts, where it is a monologue, not a dialogue, and is expected to convey clearly the logical connection, you would just use a single full sentence: “You can join me if you like.” (I’m not going to wander into the can/may argument here, but here is a full article on things many people think are errors that aren’t:

An important step on the way to being an expert user of the language is to read authors you respect in as many different genres as possible. Learning cut-and-dried rules and trying to apply them as broadly as possible won’t make you an expert user; in fact, you risk destroying your ear for the language. You need to be able to hear and read it as your readers will. You won’t be in a position to give them lessons in how to hear it the way you’ve learned to.