Hysteria over Hyphens.

The Economist‘s Johnson column (on language) has a good roundup of the vexed issue of hyphens, which starts with the classic quote “If you take hyphens seriously, you will surely go mad.” I got the link from this post by Lucy Ferriss, who is quite wrong about spelled-out numbers followed by a reference to a measurement, as in “a twenty-five-year-old car.” She thinks it’s OK to “skip one or two, e.g. a twenty-five year-old car.” No. It is not. You use every one of those hyphens or you fail my copyediting test.


  1. Crash Davis says:

    If you have sex with twenty-five-year-olds you’re not going to prison. Twenty-five year-olds, or twenty five-year-olds, however…

  2. What can she be thinking?

  3. The URL for the post at Economist is
    https: //www.economist.com/news/books-and-arts/21723088-hyphens-can-be-tricky-they-need-not-drive-you-crazy-hysteria-over-hyphens
    Now, that’s the right amount of hyphens for sure.

  4. Do you find usage has changed? I would never leave out any of the hyphens in “a twenty-five-year-old car,” but in recent years I’ve wanted to avoid constructions that call for several hyphens close together; they make me think of typewritten dissertations. Once “a mid-nineteenth-century poet” looked entirely normal to me, but now nothing seems quite right, and I try to get away with “a mid-nineteenth century poet” or “a mid–nineteenth century poet.”

  5. I’ve noticed that “mid” often doesn’t take the hyphen at all. I am not sure if that is a new thing or if there were always two competing traditions.

  6. hysteria-over-hyphens-language-hat-post

  7. Athel Cornish-Bowden says:

    Maybe I’ve gone mad (in Lucy Ferris’s sense) as well as senile (see yesterday’s posts), but you’re right, of course, and “twenty-five-year-old car” needs its three hyphens. In this case it’s not obvious how to phrase it better — maybe a “car from twenty-five years ago”. In my own writing I’d struggle to find a way to avoid having so many hyphens.

    I’m currently involved with preparation of a review on the origin of life. There are more than 20 authors, with almost as many addresses, which makes it difficult to agree. Fortunately most of them are barely contributing, but even with six or seven it’s not easy. Anyway the phrase “origins of life” occurs many times, sometimes as an adjective (“origins-of-life”) and sometimes as a noun (“origins of life”). In the first draft the presence of absence of hyphens was more or less random, and I went through making them all consistent. Unfortunately two of the authors didn’t know or didn’t understand the rule and said it was inconsistent. However, I managed to shut them up with a reference to the Chicago Manual (which does, however, express it more weakly than I would have liked).

    Incidentally why “origins” rather than “origin”? I prefer the singular but there was a clear majority for the plural. However, it’s a scientific question (how many origins do you think there were?) rather than a grammatical one.

  8. Origin of life implies single creator, I think.

    Plural is more consistent with a view that life originated as a result of natural processes rather than act of creation.

  9. How about “a 25-year-old car”? Digits can’t be that bad, can they, seeing as the number is at least 10 and is not at the start of a sentence.

  10. Five-and-score-year-old-car.

  11. Mark Twain: “the on-the-other-side-of-the-river-lying mountains” (calquing German).

  12. David Marjanović says:

    Thank you.

    Another common issue is complete confusion about hyphens and dashes. I’ve reviewed dozens of manuscripts and read plenty of published works where incorrect hyphens tripped me up-like that-see?

    What can she be thinking?

    There seems to be a widespread myth in English that you can’t have more than one hyphen in a word.

    Incidentally why “origins” rather than “origin”? I prefer the singular but there was a clear majority for the plural.

    That’s thoroughly bizarre unless they’re talking about separate origins of life on different planets. It is plain obvious that all known life has a single origin – protracted as this origin must have been.

    calquing German

    Funnily enough, Mark Twain added the hyphens to let the reader wrap their mind around German box-in-box syntax. They’re not there in the original: die auf der anderen Seite des Flusses liegenden Berge.

  13. In my field of linguistic research, there are major differences in the usage of hyphens vs. en dashes vs. open compounds between (U.S.-based) Oceanic Linguistics and (Australia-based) Pacific Linguistics. OL upgrades hyphens to en dashes when a prefix (most commonly proto-) is attached to an open compound, as in Proto–North Sarawak, while PL treats Proto as an independent word, as in Proto North Sarawak. To get an idea of the confusion that prevails on this issue, try googling “proto central eastern malayo-polynesian languages”! Of all the variants that turn up, perhaps the only missing one is hyphens throughout: proto-central-eastern-malayo-polynesian languages.

    BTW, one of my favorite obscure dissertation titles is “The Proto–North Sarawak Vowel-Deletion Hypothesis”!

  14. J.W. Brewer says:

    I saw someone advocating that style (w/ en dashes) recently, and it struck me as particularly bizarre because it encouraged misparsing. I think one of the examples was “post–Cold War,” which seemed to me to suggest [[post cold] war] rather than [post [cold war]]. Similarly, I assume the internal structure of compound Joel mentions is [proto [north Sarawak]] not [[proto north] Sarawak]?

  15. The idea is that hyphen is close scope, space is wider scope, and en-dash is widest scope. It would be simpler to remember if the en-dash were wrapped in spaces: Proto – North Sarawak, but that looks too much like an em dash: Proto — North Sarawak.

  16. I think one of the examples was “post–Cold War,” which seemed to me to suggest [[post cold] war] rather than [post [cold war]].

    That’s odd; as a copyeditor who routinely inserts that en dash, it seems obvious to me that it’s the en dash that suggests [post [cold war]] and the hyphen that parses as [[post cold] war]. I mean, look at “post-Cold War”: how can that not read as “a War that’s post-Cold”?

  17. In LaTeX, the en-dash is set lower than the others, with the intention that it only be used between ranges of numbers. As a result, I never use en-dashes for grammatical purposes at all.

  18. David Marjanović says:

    Do you use em-dashes instead? Those seem to be restricted to the English language.

  19. J.W. Brewer says:

    My vague impression is that most of the people who view en-dashes and hyphens contrastively, i.e. as signalling different semantic content when put in the exact same place, are in the tiny minority of the population that has worked in the printing/typesetting/copyediting industries. To the extent they think any material proportion of their readers in the general population are grokking the distinctions they think they are signalling according to the internal conventions of their subculture, I suspect they are naive. That neither en nor em dash exist on an old-timey typewriter keyboard (or in the basic ASCII set) except for the ad hoc — (two consecutive hyphens) approximation is probably a contributing factor to making these distinctions part of specialized/subcultural jargon.

  20. J.W. Brewer says:

    Hmm, I guess when I typed in two consecutive hyphens it automatically displayed as an em-dash? That’s not how it worked in the old days … (get off my lawn etc etc …).

  21. @David Marjanović: Yeah, I only use em-dashes, except to denote inclusive ranges (which can occasionally be done with words instead of number, like “Fri.–Sun.”). I use “grammar dash” as a virtual synonym for “em-dash,” although I’m not sure if I made that up myself, or if I picked it up somewhere.

  22. Marja Erwin says:

    I use hyphens. I disable any kind of autocorrect or autoformat, since it goes haywire. I sometimes use double hyphens, but not n or m dashes.

    I take it post-Cold means Holocene?

    And proto-North refers to the fringe Borean hypothesis?

  23. Regina Lusca says:

    I only use em-dashes on Twitter, because they save the two spaces I’d need with an en-dash.

  24. I think many people don’t really know that en-dashes are a thing but nevertheless interpret them differently from hyphens when reading; somewhat like how native speakers can apply grammar rules they are unaware of.

    Judging by Wikipedia, Canadian electoral districts scope order is hyphen — space — unspaced-em-dash.

    The busiest section of RER B is probably the Châtelet–Les Halles — St-Michel–Notre-Dame section (demographically and punctuationally).

    [edited as the thin space wasn’t showing thin enough]

  25. > Do you use em-dashes instead? Those seem to be restricted to the English language.

    How so? I’m quite sure they’re not.

  26. David Marjanović says:

    They’re definitely never used in German in any case. The Gedankenstrich – “thought stroke”, used like parentheses but for emphasis instead of deemphasis, or for dramatic pauses – is an en dash between two spaces as I’m demonstrating.

  27. Trond Engen says:

    A Norwegian tankestrek has the same function as a German one. But we use the longer dash — probably tje one you call m-dash — instead of double spaces. I have no idea where it’s hidden on my keyboard. In the days of Ascii internet I learned to use a double hyphen, and after Microsoft word started turning double dashes into long dashes it’s been a moot point.

  28. The names of these dashes can be written as “em-dash”/”en-dash” or “M-dash”/”N-dash,” but “m-dash”/”n-dash” is incorrect. The names come from the fact that the dashes are conventionally the same width as a capital M or N. These widths are used a (typeface-dependent) units of spacing in typography, but they are conventionally spelled “em” or “en” (rather than “M” or “N”) to avoid confusion with the actual letters themselves.

  29. David Marjanović says:

    I have no idea where it’s hidden on my keyboard.

    MS Office:
    – Ctrl+minus on the decimal block
    — Ctrl+Alt+minus (or AltGr+minus if you have an AltGr key)

    – Alt+0150 on the decimal block
    — also has such a code, but I forgot it.

    Alt+0150 is what I’m using here.

  30. — also has such a code, but I forgot it.

    — Alt+0151 on the decimal block

  31. AltGr+Q m and AltGr+Q n on the Moby Latin and Whacking Latin keyboards.

  32. In Portuguese we definitely use the longer travessão, albeit with spaces, for this kind of parenthetical. Our typography tends to be ripped off the French so I bet they do it too? We also use em-dashes to mark dialogue lines (yes I know U+2015 HORIZONTAL BAR is supposed to be used for that rather than em-dashes proper; but I’m pretty sure non-typographers think of them as the same travessão). I’d never even seen the en-dash used for that before I started reading English; and I believe its Portuguese name, meia-risca, is much less well-known. I find it uglier, but eventually got used to it; I always prefer it for text that may be read in monospace (code, README files, email etc.), since unspaced em-dashes don’t really work in mono.

    It’s surprising to me that to this day Windows, touted for its user-friendliness, still relies on Alt+digit sequences as the only way to input special characters. On Linux I just do Compose+two hyphens for en-dashes, Compose+three hyphens for em-dashes, (Compose,a,e) for æ, (Compose,=,y) for ¥, and so on. Wouldn’t live without a Compose key.

  33. It’s surprising to me that to this day Windows, touted for its user-friendliness, still relies on Alt+digit sequences as the only way to input special characters.

    It isn’t the only way, it’s the only way that works on any physical keyboard and keyboard driver. The Moby Latin (for physical U.S. keyboards) and Moby Whacking (for physical UK and Ireland keyboards) drivers allow typing about a thousand different characters with AltGr+key or an AltGr+key followed by key sequence. I’ve never gotten around to making variants for other physical keyboards — lack of demand so far.

  34. @John Cowan: Yes, “only way” is an error on my part. I mean more like the default, built-in, cross-locale, multilingual, non-extra-software-needed way, like Compose is for Linux. (Making (Compose, ], h) output a ‹˺› (end high tone) is my personal customization, and making ‘rabu’ print (ღ˘⌣˘ღ) is part of my Japanese input method; but (Compose, -, -, -) will make an em-dash on any Linux machine).

  35. Trond Engen says:

    I should perhaps say that I knew how to produce a long dash on the digital pad. It’s on my laptop computer without a digital pad that I’m resorting to double hyphens and MS Word Autocorrect.

    I don’t think I’ve ever composed anything more advanced than a Usenet post on a Linux system, The “Compose” key does look handy. If not for anything else, so because the sequences seem easier to remember than numerical strings.

  36. the default, built-in, cross-locale, multilingual, non-extra-software-needed way

    That works because Compose is higher-level than the keyboard driver, so you specify compose sequences in terms of keyboardable characters, not physical key positions. This means that a Compose sequence won’t work if the underlying keyboard does not have the character available for typing.

    In Windows there is no such user-programmable layer between the driver and the rest of the system, so everything has to be defined with respect to physical key positions. Furthermore, support for more than two keystrokes is very limited, and I don’t attempt to use it.

    Still, I get a lot of mnemonic mileage out of what I’ve got, as you can see. README file.

  37. Lars (the original one) says:

    Danish typographical conventions were probably ripped off from German, but FWIW there’s only hyphen and em-dash, and with one obscure exception em-dash is for grammar and has spaces. hyphen is for composing words and ranges and never does.

  38. Trond Engen says:

    And Norwegian conventions are Danish.

  39. Hat, what do you use for the Russian City dash (“The town of M——” etc.)? What about the Victorian Curse dash (“D—— it to H——!”)?

  40. I don’t think I’ve had to do either, but if I do I’ll try to remember to copy yours — they look great!

  41. I think you are supposed to use the q― d―, officially known as HORIZONTAL BAR for some obscure reason.

Speak Your Mind