I was reading Catherine Shoard’s Guardian puff piece on Mark Rylance (“the best actor of his generation,” “the world’s greatest actor”) when I was pulled up short by this passage:

He plays Terry, a banker who once ran a Congo-based squad of assassins (including Penn and Ray Winstone). As projects go, it feels a bit skew whiff for a pacifist so committed that he winces at the very mention of American Sniper.

I was so unfamiliar with “skew whiff” that I assumed it must be a typo, but a moment’s googling showed me my error: the Cambridge Advanced Learner’s Dictionary defines it as “sloping instead of straight, or wrongly positioned: You’ve got your hat on skew-whiff.” And it’s not some recent bit of slang; the OED (in an entry from 1933) takes it back almost three centuries:

dial. and colloq.

Askew, awry (lit. and fig.).

1754 Scots Mag. July 337/2 Behind, with a coach-horse short dock, cut your hair; Stick a flower before, scew-whiff, with an air.
1839 W. Holloway Gen. Dict. Provincialisms (new ed.) 154/1 Skew-whift, adj. (Askew, from Skef, Belg. oblique; and perhaps Whiffed, blown.) Awry.
1895 J. T. Clegg Stories, Sketches, & Rhymes in Rochdale Dial. 228 Her judgment’s getten thrut skew-wift.
1899 Shetland News 20 May 7/2, I hed ta geng skewquieff.
1935 A. P. Herbert What a Word! iv. 101 Go on cackling..until the orator has to stop and ask you why you cackle. Then tell him. He won’t get Frankenstein skew-whiff again.
1946 D. L. Sayers Unpop. Opinions 59 When Neptune shouldered Britain out of the sea, he did not make a neat engineering job of it. Characteristically, Britain came up skew-wiff, with one edge thick and hard and the other soft and thin, like a slice of wedding-cake.
1959 I. Opie & P. Opie Lore & Lang. Schoolchildren iii. 47 If a boy’s cap is on skew-whiff: ‘Are you wearing that cap or just walking underneath it?’
1974 J. Cleary Peter’s Pence iii. 82 Our plans seem to have gone a bit skew-wiff, don’t they? That’s the trouble with the Irish.
1977 Lancashire Life Feb. 53/4 Thi tie’s put on skew-wiff.

Is this a word every UKanian knows? And are any of my non-UK readers familiar with it?

Rylance is indeed an amazing actor, by the way, and my wife and I are thoroughly enjoying the BBC Wolf Hall adaptation.


  1. Jeffry House says

    Never heard of it. (Ontario, Canada)

  2. It looks vaguely familiar, but if I had encountered “skew whiff” before, I either forgot or didn’t bother to look it up.

  3. Stefan Lewicki says

    I’m a retired Englishman and have known this word for as long as I can remember. All sorts of things can be skew-whiff here: it means they’re crooked, or not on straight or level. Can apply to items of clothing, DIY efforts, lots of things. Astonished as you are not to have come across it, I’m amazed you don’t have the word on your side of the pond…

  4. Yes, very common in the UK, although less so in the abstract sense in which Shoard uses it. Normally applied to a physical object.

  5. Yep, “skew-whiff” is a common phrase here in the UK, though my daughter would be more likely to use “wonky” instead…

  6. I’m English, from London. Yes, I’d say it, in the sense of a thing being uneven. But I’d spell it (and say it) as one word. Skewiff or something. Not pronouncing two W’s.

  7. David Eddyshaw says

    I confirm that it’s normal Cispondic, if a bit old-fashioned. I don’t think I’ve ever actually seen it written down before, though.

  8. I’ve never heard of it either.

    I’m a retired Englishman

    What nationality do Englishmen assume after they retire? Norman? Portuguese? Thai? Moroccan?

  9. Jonathan Wright says

    I’m southern English, early 60s, RP speaker. I may even use the expression from time to time. I’m certainly very familiar with it but I think it might be going out of fashion

  10. David Eddyshaw says

    “What nationality do Englishmen assume after they retire?”


  11. David Eddyshaw says

    Apologies. The politically correct term is “Anglo-Emeritan.”

  12. *tries to work up Emer(ald Isle) joke, gives up, gets back to work*

  13. Normal for me (Edinburgh)

    Not how I would have spelt it (mostly because I make a distinction between w and wh, and it’s definitely w if anything), but I’m not sure how I *would* have spelt it…

  14. My mother always pronounced it Skee-whiff (Southern UK). I’d never seen it written before.

  15. in South African vernacular English there is a word ‘skeef’ borrowed from the Afrikaans, which means much the same thing.

    if I ever heard skew-whiff I probably thought they were actually saying skeef.. don’t recall ever seeing it in print though.

  16. I’ve definitely read it a few times – maybe in Scottish books. There’s also the lovely word “skewbald”:

  17. Jonathan D says

    Australian, and like wn, I’d expect skewiff or something like that.

  18. Chris McG says

    Definitely a word I say and hear in the UK. The spelling looks weird, but I can’t think of what would look right (a bit like ‘hoick’, I can never find I way to spell that without it looking off to me either)

  19. Perfectly cromulent in my circles in New Zealand.

  20. Australian, and know it, but as a UKism that I would probably never use myself (like “barmy” or “Chrimbo”).

    Also, I agree with wn about the spelling.

  21. Used by Jim Prideaux in Tinker, Tailor, Soldier, Spy. A Czech character who identifies as an Englishman.

    ‘Skew-whiff. Like me. Watch,’ and turned purposefully to the larger window. A strip of aluminium beading ran along the bottom, put there to catch the condensation. Laying the marble in it, Jim watched it roll to the end and fall on the floor.

    ‘Skew-whiff,’ he repeated. ‘Kipping in the stem. Can’t have that, can we? Hey, hey, where’d you get to, you little brute?’

    -John Le Carre.

    Rendered as skewy in the US edition, but said skew-whiff in the Audio book we have, from a UK reader.

    I have not heard it in any other context, (Am/E.)

  22. squadron leader squiffy von bladet says

    Perfectly ordinary banter! (UK-born and bred; currently mid-40s and emmigrated.)

  23. Very familiar expression to me (an Emeritan, an ancient Emeritan). Michael Quinion says (World Wide Words) that it’s 18th C Scots, and draws attention to a similar N Am expression “skewgee”. Anyone over there heard of skewgee?

  24. I thought Chrimbo must be Australian, like arvo and things – definitely never heard it here!

  25. Definitely well-known in my circles in the south-east of the UK. I think it’s on TV a lot? Helps drill it in!

  26. I’m very familiar with the expression, but I take it to mean strictly “at an asymmetrical angle”. I don’t see it as synonymous with “wonky”, “pear-shaped”, etc, meaning a failure or something not quite right.

    Something I would say when I am trying to get the fit of my glasses adjusted, except that no one here would understand it.

  27. This word meant nothing to me.

    (Also, regardless of whether the OED entry dates from 1933, it must have been revised at some point since 1977.)

  28. I’ve only ever heard “Chrimbo” in a UK context (magazines and so on) — the Australian equivalent I’d expect is “Chrissie.” I’m sure there’s a lot of regional variation, though.

  29. David Marjanović says

    Oh, a lurking ideophone, I thought. But:

    in South African vernacular English there is a word ‘skeef’ borrowed from the Afrikaans, which means much the same thing.

    *lightbulb moment* German schief “transverse”. So maybe it’s related after all.

  30. David Marjanović says

    *headdesk* Not “transverse” – “oblique”!

  31. Trond Engen says

    Also No. skeiv “oblique, tilted”, now also “queer”, but a cognate would have been more like *shoaf. There’s a dialect word, skjage “be unsteady, stand swaying” (of furniture as well as drunks), that might seem to belong, but the dictionaries tell me it’s actually stjage, related to ‘stagger’.

  32. squadron leader squiffy von bladet says

    Dutch “scheef”. Let’s just announce it was borrowed by some sailors?

  33. Huh. Like mollymooly, I don’t recognise the word, although I have definitely read Tinker, Tailor, Soldier, Spy.

  34. A friend introduced me to Crimbo, as he spells it. He’s English, with an Australian mother and a Hungarian father. He is also a syntactician of English.

  35. Ian Press says

    Native speaker of English, from Lancashire but long exiled to London and wasted some time commuting to Scotland. A word very familiar to me, but I don’t think I’ve ever uttered or written it myself. Possibly a bit dated.

  36. UK. Normal speech if a bit dated. Would expect to see it as one word not two, although I am not sure I’ve seen it written before. Other expression related to ‘on the skew’ (skewiff’s more formal brother) might be ‘on the bonk’ said in my hearing (as a young man) on a construction site about a misaligned wall. I never heard ‘on the bonk’ ever again, except when I’ve said it myself, which makes we wonder if the block-layer invented it the spot.
    If I had to write skewiff I would be tempted by skewhiff with an ‘h’. But let’s not get all ‘cool whip’ about it.

  37. Paul (other Paul) says

    But I once saw Rylance murder Twelfth Night by playing Viola (yes, a man playing a woman impersonating a man) as a Japanese kabuki actor with the make-up and mincing little steps, while the rest of the case played it straight. It has (quite unfairly) jaundiced my appreciation of him ever since.

  38. Western isles Canuck here. I too have heard and used this expression all my life it seems. I know it pronounced ‘skew-whiff’ so agree that it’s a combo of ‘askew’ and ‘whiffed’, i. e. ‘knock aside by a puff of wind’. It also partners nicely with ‘wonky’.

  39. Trond Engen says

    Seems like English ‘skew’ is syncretic in origin. Etymonline:

    skewbald (adj.)
    1650s, “having white and brown (or some other color) patches, spotted in an irregular manner” (used especially of horses), from skued “skewbald” (mid-15c.), of unknown origin, + bald “having white patches” (see bald). First element said to be unconnected with skew (v.) (but Klein’s sources say it is); OED suggests perhaps from Old French escu “shield,” but also notes a close resemblance in form and sense with Icelandic skjottr, “the history of which is equally obscure.” Watkins says it is Scandinavian and akin to Old Norse sky “cloud” on the resemblance of the markings to cloud cover.
    When the white is mixed with black it is called ‘pie-bald,’ with bay the name of ‘skew-bald’ is given to it. [“Youatt’s ‘The Horse,’ ” 1866]
    As a noun meaning “skewbald horse” from 1863.
    skewness (n.)
    1877, from skew + -ness.
    skew (v.)
    late 15c., “to turn aside” (intransitive), from Old North French eskiuer “shy away from, avoid,” Old French eschiver (see eschew). Transitive sense of “turn (something) aside” is from 1570s. Meaning “depict unfairly” first recorded 1872, on notion of being “give oblique direction to,” hence “to distort, to make slant.” Statistical sense dates from 1929. Related: Skewed; skewing. The adjectival meaning “slanting, turned to one side” is recorded from c. 1600, from the verb; noun meaning “slant, deviation” first attested 1680s.
    askew (adv.)
    1570s, of uncertain etymology; perhaps literally “on skew” (see skew), or from the Old Norse form, a ska. Earlier askoye is attested in the same sense (early 15c.).

  40. Trond Engen says

    Old North French eskiuer “shy away from, avoid”

    Thinking of it, this looks very much like a loan of a Scandinavian cognate of ‘shy’, Danish sky “avoid”. But I think the Danish verb may be an early loan from some north German variety (or else it would end in -g), so maybe it’s an older Germanic loan.

  41. David Marjanović says

    German scheuen would fit.

  42. John Wells says

    An unremarkable part of my vocabulary since childhood. (England – born and grew up in the north, mostly schooled in the south.)

  43. Trond Engen says

    …or else it would end in -ge, I meant.

    Yes, that German verb, but Danish seems to have borrowed a northern form early enough to change *ew > *ju > *y:.

  44. “skew-whiff” is absolutely commonplace for me, a 65 year old Australian living in Melbourne and also for my 88 year old mother, but I am not so sure about my 33 year old son. To me it means “awry”, either figuratively or literally. I would spell it thtat way too.

  45. Very familiar to me, though maybe I’d associate it more with my parents’ (wartime) generation. The DSL entry is livelier than the OED, and suggests ‘squiffy’ comes from here too:

    SKEW, v.2, adj., adv., n.2 Also skeu, sku(e), scew, scue; skeugh, skeuch (ne.Sc.); sceow, sk(y)(e)ow, skyou, skowe; scoo (Gall. 1904 E.D.D.) and erron. schew- (ne.Sc.); and reduplic. nonce forms skaoowaoo, squouwow, skeewow, skee whee. Sc. forms and usages.
    [skju; ne.Sc. skjʌu, †skjux]
    1. v. ¶1. tr. To shun, avoid, take shelter from (rain or the like) (Rxb. 1825 Jam.).
    2. intr. To go in an oblique direction, to move sideways, to go off the straight, to sway from side to side in an affected manner, to swagger (Bnff. 1866 Gregor D. Bnff. 165, skyow; Bnff., Abd. 1970). Ppl.adj. skewed, see 1880 quot.
    Abd. 1813 D. Anderson Poems 112: 
Contemplating ilk foppish brat That’s got a sword and cocket hat To see them skew and skip about.
    Bnff. 1866 Gregor D. Bnff. 165: 
The dyke’s beginning t’ skyow.
    wm.Sc. 1880 Jam.: 
A half-drunk person, when walking zig-zag, is said to be skew’d.
    Sh. 1897 Shetland News (16 Oct.): 
Try ta mend dy weeked wies, an’ niver skew.
    3. tr. and intr. To twist, distort, turn sideways, screw round: (1) in gen. and fig. (Sc. 1880 Jam.; ne. and em.Sc.(a), Lth., wm. and sm.Sc. 1970). Now chiefly dial. in Eng. Ppl.adj. skeuched, twisted to one side (Abd., Kcd. 1825 Jam.), badly made, shapeless (Gall. 1904 E.D.D., scood), skyowt, off the plumb (Gregor), down at heel, of shoes, skewed, demented, off one’s head (Per. 1808 Jam.).
    Bnff. 1866 Gregor D. Bnff. 165: 
He hid on aul’ skyowt bashles o’ beets.
    Sc. 1869 A. Leighton Sc. Words 24: 
His theologic veesion may be skew’d.
    Kcd. 1933 Scots Mag. (Feb.) 333: 
Rob skeughed his face round, What, money for school?
    (2) of the feet, legs or gait: to splay, turn outwards or awry; “to walk with a waddling gait” (Bnff. 1866 Gregor D. Bnff. 165). Ppl.adj., vbl.n., skyowan, -in (Id.). Hence skyeowed, splay-footed (Abd. 1970); deriv. skewtsie, “rubbing the legs together as one walks” (Ork. 1958).
    Abd. 1880 W. Robbie Yonderton 8: 
The latter turned out a good deal at the toes, or, to speak more plainly, “skyeowed.”
    (3) intr., of the eyes, glance, etc.: to squint naturally or on purpose, to look askance in a suspicious or disapproving manner. Also in Eng. dial. Ppl.adj. skewed, squint(-eyed) (wm.Sc. 1880 Jam.).
    wm.Sc. 1907 N. Munro Daft Days ii.: 
The letter-carrier’s eyes may — may skew a little.
    Lnk. 1922 T. S. Cairncross Scot at Hame 13: 
I’ve skewed and skeighed and skirled and skelped.
    Kcd. 1934 L. G. Gibbon Grey Granite 33: 
Ma Cleghorn would skeugh at you over her specs.
    †4. intr. To quarrel, to fall out, disagree (Abd. 1911 Abd. Jnl. N. & Q. IV. 50; ne.Sc. 1970).
    Mry. 1921 T.S.D.C.: 
“Far’s Sammie?” “A dinna ken; me an’ him’s skeowt i’ noo.”
    ne.Sc. 1924 Swatches o’ Hamespun 11, 66: 
The wife an’ me’s like to skyow. . . . Sceowt, hae ye?
    II. adj. Off the straight, oblique, wry, slanted (Gall. 1904 E.D.D., scoo; Ork. 1920 J. Firth Reminisc. 154, skaoowaoo, Ork. 1970, skeewow; Inv., ne.Sc., em.Sc.(a), Lnk. 1970); of the feet: splay (ne.Sc. 1966).
    Sc. 1911 S.D.D.: 
Skew-mouth. A crooked mouth. Skew-mouth plane, a kind of joiner’s plane.
    Abd. 1920 A. Robb MS. iv.: 
His feet was gey and sair laid oot — skyow as a body wad say.
    Kcd. 1933 Scots Mag. (Jan.) 248: 
Reddish hair and a high, skeugh nose.
    Combs.: (1) skyow-fittet, splay-footed (ne.Sc., Ags., Dmb., Lnk. 1970); (2) skew-whiff, -wheef, -quieff, awry, ajee, at a rakish angle. Gen.Sc. Also in dial. or colloq. Eng.
    Abd. 1875 W. Alexander My Ain Folk 220: 
Ane o’ yon chiels, yon skyeow-fitted breet.
    Abd. 1912 G. Greig Mains’s Wooin’ 50: 
Whether your man’s to be bow-hocht, or skyow-fittet.
    Bnff. 1937 E. S. Rae Light in Window 15: 
Chilpit-leukin’, booet-owre, schew-fittit, bowe-hoched, plavers o’ poets.
    Sc. 1754 Scots Mag. (July) 337: 
Stick a flower before, scew-whiff, with an air.
    Sh. 1899 Shetland News (20 May): 
I hed ta geng skewquieff.
    Sc. 1931 F. Niven Paisley Shawl 245: 
My time-table is all skew-whiff.
    Sh. 1949 J. Gray Lowrie 94: 
Twa o’ her wheels i’ da stank o’ da rod, an’ shu wis aa staandin skeow-wheef.

    III. adv. At a slant, in an oblique direction, askew, this way and that, sideways; in a distorted manner, with an affected, waddling gait (Bnff. 1866 Gregor D. Bnff. 165).
    Sc. 1883 J. Kennedy Poems (1920) 112: 
Awa’ the erring quoit gaet skeugh.
    Sh. 1923 T. Manson Lerwick 268: 
The houses have been built on no plan except that of being set “squou-wow” — anyhow.
    Kcd. 1933 L. G. Gibbon Cloud Howe 248: 
A lorry came down the road and went skeugh.
    IV. n. 1. A shelter, protection. See I. 1.
    Slk. 1826 Wilson Noctes Amb. (1863) I. 211: 
But house nor hame aneath the heaven Except the skeugh of greenwood tree.
    2. A twist, turn, sideways movement (Ork., ne.Sc., Slg., Lth., wm.Sc., Kcb., Rxb. 1970). Comb. skee-whee, id.
    Bnff. 1866 Gregor D. Bnff. 165: 
He ga’s fit a skyow, an’ caed himsel’ oot o’ the queet.
    Ork. 1880 Dennison Sketch-Bk. 130: 
His heid ap i’ a skue.
    Cld. 1880 Jam.: 
Gie the stane a skew this way.
    Cai. 1929 John o’ Groat Jnl. (22 Nov.): 
If ‘e aeroplane thing took a skee-whee.
    3. A squint, sidelong glance, a look askance (Mry., Abd. 1970).
    Kcd. 1933 L. G. Gibbon Cloud Howe (1937) 220: 
With his skeugh and his puzzled eyes.
    4. In mining: a piece of rock lying in a slanting position and tapering upwards which overhangs a working-place and is liable to fall, a hitch (Sc. 1886 J. Barrowman Mining Terms 61; Ayr. 1949). Also in Eng. mining usage.
    5. A quarrel, a row (Mry., Abd. 1970).
    Abd. 1905 Banffshire Jnl. (18 April) 7: 
Tho’ sair I leuch at this bit rowie, I kent it wis a passin’ skyowie.
    6. A kind of wooden vane or cowl in a chimney which revolves according to the direction of the wind and prevents smoking (Kcd. 1825 Jam.). See wind-skew s.v. Wind.
    [O.Sc. skew, protection, 1475.]

  46. (I took the liberty of bolding the relevant bits, since it’s a long entry.)

  47. To users of the expression, has it ever been used, or can it be used as a eupemism for gay, like ‘bent’?

    zhoen: perhaps I missed something, but I thought Jim Prideaux was British born and bred.

  48. Y,
    Quite correct, he spoke Czech and most of his education was outside of England, so I mis-remembered.

  49. Poodlemom says

    Southeastern American here.

    I’ve never heard the expression skew-whiff, but I have heard of skewbald (a horse color) and askew, meaning crooked. If I were to use a colloquial expression with a similar meaning, I would say “out of whack.” Whatever in whack is, I’ve never known.

  50. David Eddyshaw says


    “To users of the expression: has it ever been used, or can it be used as a euphemism for gay, like ‘bent’?”

    No. Well, I can’t say about “has it ever been used”, but I’ve certainly never come across such a use.

    Not sure if I’d call “bent” a euphemism; more of a deliberate slur.

    Actually I’d say that in the UK “bent” is a good bit more likely to mean “corrupt” than “homosexual.” In particular, “bent copper” is pretty much a standing expression for a corrupt policeman.

  51. Australian, Victorian – hear semi-frequently, use less frequently

  52. Trond Engen says

    Euphemisms and slurs are not mutually exclusive. It’s inherent to the euphemism that there’s something really distasteful about the referent.

  53. Trond Engen says

    I see that some of the Scottish forms above end in [x]. That would seem to suggest that they’re inherited rather than borrowed from French or Scandinavian.

    I wonder if Da. sky and Sw./No. skygga might be Kluge doublets < *skewhW-.

  54. David Marjanović says

    German eu comes at least sometimes from MHG /yː/ (which is preserved in Low German and also in Not Too Low Alemannic), so the Danish form doesn’t need to be a loan at all; its sk means it would have to be really old.

    The Verner alternant of */xʷ/ was */ɣʷ/, which turned into */w/ in a lot of contexts, so you may well be right about Kluge doublets.

  55. Bob Gillham says

    Full disclosure: I’m 61, used to watch ALL the British WW2 films so may have picked it up from RAF slang but as far as I know I’ve known and used this all my life. I was born and raised in the Welsh Marches (Oswestry) so NOT just a SE thing…

  56. I don’t remember ever seeing skew whiff before, or noticing it anyway, but in the west of Ireland skew-ways is not uncommon, and I remember hearing it since childhood. Different forms are attested: scew-ways, skow-ways, skeow-ways; the last of these is closest to how I know it. The two reference books I looked at point to Irish sceabha ‘slant, skew’, as in ar sceabha ‘askew’.

  57. Is the stress in skew-ways on the final syllable, as I gather it is in skew-whiff?

  58. Hat: No – I’ve only ever heard the stress on skew in skew-ways.

  59. @David: sk- in Danish loans from Low German is not probative of age, there was enough awareness of correspondences that sch- could be ‘corrected’ to sk- when borrowed.

    By coincidence, my example is the homonymic noun which in Danish denotes a meat broth gelatin:
    L. iūs (/i̯u:s/ or /ii̯u:s/?) > F. jus /ʒy/ > LG. /ʃy/ (probably) > Da. sky /sky?/.

  60. David Marjanović says

    …Wow, that’s impressive. 🙂

  61. David Marjanović says

    But then, I speak a dialect that has borrowed OK with /ɛ/ because of a regular correspondence with Standard German…

  62. This is common enough to have a name, “contact-induced doubl(et)ing” (or in the more general case, “analogical transposition”). Dutch dijk, borrowed into English as dike, appears also in Standard German in the form Deich. Pure borrowing in the relevant (Middle High German) timeframe would give Deik today, but the borrowers understood the correspondences between Low and High Germanic well enough to generate a pseudo-High form, though not so perfect that it collided with the true cognate Teich ‘pond, pool’ (originally one created by damming a stream, I presume). The English cognate is ditch, which is also semantically separated, except in Irish English where ditch means a dike rather than the typically adjacent drain.

  63. Another term for it is etymological nativization, which we’ve actually used here in the past.

  64. Davis X. Machina says

    Quite common in rural New England and the Maritimes.

  65. David Marjanović says

    originally one created by damming a stream, I presume

    Yes; I’ve seen classifications (of what counts as a “lake” or “pond” etc.) that specified that the water can be drained from a Teich.

  66. hoick

    I don’t know this verb, but it means, according to various dictionaries, ‘lift up or hoist, often with a jerk or rapid movement’ (so the OED). This immediately made me think of hike, which I do know and use in the exact same sense, and indeed the OED conjectures that they are dialect variants. Hike definitely began as a dialect verb, got into AmE, and then borrowed back into BrE in the sense ‘walk, ramble’ < ‘walk laboriously, march’. Back to the sense of hoisting: you can hike up your pants/trousers or hitch them up; the OED doesn’t relate these forms, but they look related to me for all that.

  67. marie-lucie says

    skew (v.)
    late 15c., “to turn aside” (intransitive), from Old North French eskiuer “shy away from, avoid,” Old French eschiver (see eschew)

    It looks like the “Old North French” still exists in Standard French, written esquiver which means something like ‘to dodge’ (a blow) with a quick, agile motion. As far as I know it is not used figuratively, unlike English “eschew”. It is also used as a reflexive verb: s’esquiver means ‘to escape quickly and unobtrusively, to make oneself scarce’.

  68. I grew up in Australia, hearing and knowing the word skew-whiff. It was used to mean something was crooked, such as a picture hanging on the wall. I still use and hear this expression from time to time, though generally, it is really only heard among older generations these days.

  69. I just thought to check Green’s Dictionary of Slang, and the only entry I could find that remotely resembled this was skeef, which I see Doug K (April 30, 2015 at 7:04 pm) mentioned above (“in South African vernacular English there is a word ‘skeef’ borrowed from the Afrikaans, which means much the same thing”):

    skeef adj.
    [Afk. scheef, askew]

    1. (S.Afr.) crooked, off-beat; also as adv.

    1969 [SA] A. Fugard Boesman and Lena Act I: Why you looking at me so skeef?
    1977 [SA] C. Hope Ducktails in Gray Theatre Two (1981) 56: That old queen there, Sharon’s mom, […] Should have seen the way she was checking me skeef when I asked her for more beer.
    2001 Wavescape Sl. Gloss. [Internet] Are you checking me skeef? (Are you looking at me crooked, ie, spoiling for a fight?).

    2. (S.Afr. gay) thus in fig. use, homosexual.

    So I wonder: are the words related? And how did Green miss the apparently very common “skew-whiff”?

  70. As David remarked, German schief must be in the mix. It means “crooked/skewed/out of plumb” in many of the contexts in which these are used in English – crooked smile, crooked house, picture is hanging crooked. It’s pretty much synonymous with schräg.

    windschief is a variant for houses, sheds etc that look as if they had been bent by the wind, as trees are.

  71. David Marjanović says

    It’s pretty much synonymous with schräg.

    Yes, and both refer to straight lines. I thought crooked implies irregular bends?

  72. Not necessarily: a picture hanging crooked has no bends at all.

  73. A main difference between schräg and schief is that schief always implies deviation from the norm / ideal, while schräg may be the way something is intended to be (e.g. a roof). schräg also can mean “strange, weird, eccentric”.

  74. True, I’ve never thought about it closely.

  75. David Marjanović says

    always implies deviation from the norm / ideal

    …True. I never consciously noticed; they are interchangeable when there’s no norm.

  76. No, it’s just the antonym of straight. Besides the examples above, a metal coat/shirt/pants hanger is straight if it has the bends it is supposed to have, and crooked if it deviates from those.

  77. a metal coat/shirt/pants hanger is straight if it has the bends it is supposed to have, and crooked if it deviates from those

    I disagree. The “straight” has nothing to do with “the bends it is supposed to have,” it’s only about the part that’s supposed to be straight. If the whole hanger were disassembled and laid out in a straight line, would you call it “crooked”? I thought not.

  78. PlasticPaddy says

    From Wiktionary (siehe da for [1]). I thought Germans were more consistent but maybe the foreigners are to blame:
    “Dieser wertende Bedeutungsanteil lässt sich aber nicht verallgemeinern. So spricht man abwertend beispielsweise von einem »schrägen Argument« und fachsprachlich ohne Wertung von einer »schiefen Ebene«.“[1]

  79. PlasticPaddy says

    I also like Bedeutungsanteil but wish the author could have worked the adjective “wertend” in, which would have mitigated the stark simplicity of Bedeutungsanteil.

  80. Ah, a lovely test sentence for the code I’ve just been working on for $EMPLOYER to disassemble ad hoc German compounds. Here’s what it produces using the largest of my wordlists (which are used to try to reject overenthusiastic splits made by the primitive engine):

    Dieser wert·ende Bedeutungs·anteil lässt sich aber nicht verallgemeinern. So spricht man abwer·tend beispielsweise von einem »schrägen Argument« und fachsprachlich ohne Wertung von einer »schiefen Ebene«.

    The biggest problem is that it can’t tell the difference between a long enough prefix or suffix and a compound element, thus wert·ende above, since it doesn’t know that ‘worthily-end’ (v.) doesn’t make sense. (Ich wertende dich, what the executioner says to the criminal, perhaps?) Similarly with abwer·tend.

    You can’t expect it to be perfect, given that the primitive engine is based on n-gram statistics. Here’s a better result from a corpus of German newspaper sentences:

    Die Technik setzt sich aus dem europaweiten Mobil·funk·standard Gsm, der in Deutschland über das D1-Netz angeboten wird, und dem weltweit verfügbaren System von Navigations·satelliten (Gps) zusammen.

    Note that Deutschland is in the wordlist, so no attempt is made to split it.

    Another bad result from the same corpus is hier·herver·legt instead of hier·her·verlegt. And sometimes even the best result is terrible:

    Schleswig-Holstein wolle bei der Reform des Öffe·ntli·chen Dienstes in der ‘Führungs·rolle’ bleiben, sagte Regierungs·chefin Simonis auf ihrer Jahres·pressekonferenz.

    Öffe·ntli·chen should of course just be left alone, but it isn’t in the wordlist even in lower case, and this hairball is coughed up instead

  81. PlasticPaddy says

    Maybe teach your SW some nonos. For example No “nt” at beginning and no separate “lich” so öffentlich. No “ek” in middle so “Presse-konferenz”. But maybe you expect SW to create its own nonos.

  82. Reminds me of the early days of computerized proofreading of American newspapers, when they were full of off-ended lan-downers.

  83. Well, as Don Knuth, the father of algorithmic typesetting said, when you have gotten the hyphenation of some-thing and any-thing right, you find yourself with dubious no-thing (probably better in BrE than AmE) and downright incorrect ba-thing. You can’t have every-thing. Even the very best TeX algorithm for American English has a long list of exceptions: it simply doesn’t know that har-bin-ger is correct, for example, to say nothing of ono-mat-o-poe-ia.

    In any case, my program has no such rules, just a statistical model of what is likely to appear at the beginnings and ends of words and unlikely to appear in the middles, itself a result of running it over a large training set of correctly split compounds (mostly or entirely nouns). The core algorithm was developed by Don Tuggener (about whom I know nothing) for his thesis “Incremental Coreference Resolution for German” (2016). He was just trying to isolate heads (always at the end in German) for part-of-speech determination. I just added the ability to try to re-split non-heads and to handle whole documents.

  84. @Paddy: Well, I didn’t say that “schräg” never has negative connotations, it just doesn’t necessarily have them, especially if it comes to the physical description of objects. “Schiefe Ebene” just is a traditional term of physics, I assume if it would be coined today, it would be “schräge Ebene”.

  85. PlasticPaddy says

    This was what triggered me because I gave Nachhilfestunden in physics to Abiturienten and schiefe Ebene came up. You are right; schräg and schief were doublets which moved apart but the fixed expressions stayed. There was some fluctuation in German for the physics term; at one time schiefe Fläche was also used. In English planum inclinatum > inclined plane ☺

  86. PlasticPaddy says

    Re schräg i thought French escroc was connected but wiktionary says this is ultimately a borrowing from Schurke. DWDS says schräg may be connected with PIE (s)krek from (s)ker “to bend”. Then Schurke < MHD scurgo is not so far. Maybe David M knows if there is anything here.

  87. The DWDS etymology for schräg reveals no connection with that of Schurke.

  88. David Marjanović says

    Schurke < MHD scurgo is not so far

    g > k is very far-fetched, rg > rk perhaps even more so. But there’s Scherge “henchman of an evil overlord”…

    Shark is supposed to be related to Schurke, however.

  89. Die Jagd auf den Schnurken

  90. Skew-whiff. No never heard it. Heard one word “skwiff” a bit. Definately never used “skew whiff”. I use crooked, lopsided or just skew. Unless I have led a very sheltered life I think it might be a colloquialism from somewhere up in the far North, that or it’s no longer much in use. UK Gloucestershire

  91. John Cowan says

    I’ve improved the German decompounder so that it does less harm, if perhaps a little less good as well. For purposes of natural-language processing, we don’t really want to chop off prefixes and suffixes. Processing PP’s example above now leaves all the words alone except Bedeutungs·anteil, but the two newspaper sentences are unchanged from before, so the second still mishandles the dreaded Ö-word.

    If anyone wants to play with it, it is written in Python 3 and is freely available, thanks to the GPL. A Dutch model (found in Tuggener’s original) is also in the repo, but alas none for Indonesian, another spacefreecompoundinglanguage. The models are in the repo, but the dictionaries must be obtained from the LibreOffice and OpenOffice sites. There’s a shell script to do that; it downloads and merges de-DE, de-AT, de-CH, and de-1901 dictionaries (some of the documents $EMPLOYER processes are very old leases).

    (I’m posting this because I’ve just heard from Don Tuggener; he’s impressed with my work, but agrees with me that a separate repo is the Right Thing.)

  92. January First-of-May says

    L. iūs (/i̯u:s/ or /ii̯u:s/?) > F. jus /ʒy/ > LG. /ʃy/ (probably) > Da. sky /sky?/.


    My own go-to example (…which I had yet to share) of what a few centuries of borrowing can do to initial consonants was Greek Theodoros with /θ-/ (< /tʰ-/) > Old Russian Fedor with /f-/ (because Old Russian didn’t have /θ/, and indeed neither does modern Russian) > Middle* Ukrainian Hvedir with /xw-/ or perhaps /xʷ-/ (because Middle Ukrainian didn’t have /f/ either, though modern Ukrainian gained it from loans**; /o/>/i/ is regular in this position).

    This one is far more ridiculous, though.

    *) this version persisted all the way to the time of Hvedir Vovk (1847-1918), though it was archaic by then and didn’t survive much longer
    **) this must have already happened in Old Russian, which initially didn’t have /f/ either; Byzantine loans apparently had /f/ both for original ph /f/ and th /θ/, though spelled differently until 1918

  93. David Marjanović says

    Probably only the most educated people could pronounce [f] for centuries (unless followed by a voiceless consonant or the end of a word). Allegedly it was still considered difficult to pronounce for lower-class Russians at the beginning of the 20th century.

  94. John Cowan says

    The final improvement to the decompounder is that it will not only refuse to break words in the wordlist, it refuses to accept any breaks that produce “words” not in the wordlist. So since it cannot break Öffentlichen, it perforce leaves it alone.

  95. John Cowan says

    Hugh Darwen, a computer scientist specializing in relational database theory, likes to refer to A Skew Wall that prevents us from seeing the beauty of Relationland. Mind-forged manacles and all that.

  96. David Eddyshaw says

    Probably only the most educated people could pronounce [f] for centuries (unless followed by a voiceless consonant or the end of a word)

    On the other hand, although in native vocabulary Kusaal [h] occurs only as a non-word-initial allophone of /s/, nobody seems to have any difficulty with the initial consonant of the (very common) loanword hali “up to, even.” Nobody seems to say [salɪ] or even [alɪ].

    On the other other hand, not many English speakers can manage a syllable-final [h] or even a word-initial [ŋ] … perhaps if we had more loanwords from Dagbani …

  97. PlasticPaddy says

    If the sound is recognisable in non-initial position, it can normally be produced, i.e., Hey, Ngover! There just seem to be some syllabification rules, causing e.g. Spanish to say espejo and estrella insted of *speclo and *stella. Your example of h is paralleled in Irish with the loan word halla.

  98. @J1M, I mentioned already that I confuse some Arabic f’s with th (the Arabic system is /w/, /b/, /f/).
    It can explain Russian f for th: they could think that putting your lower lip between the teeth is less unnatural than stickign your tongue in there. (Try to touch the upper [rather than inner] surface of your lower lip with the cutting edge of your incisors and you obtain th-like sound)

  99. @DM, it is associated with bilabilar adn approximant articulations of /v/, I think.

    Accordingly /w/ or even /u/ is expected in such dialects where /v/ would be devoiced. They say, there are also dialects where /x/ appears in such positions (contact with /v/-dialects or not?) and… labio-velars:-/

    Conversely, it is the position before a vowel where your bilabial is comfortably fricative. (but the above is not anything I actually know well)

  100. We have discussed a few times how there seem to be different kinds of phonotactic constraints. Some sounds or sound sequences may not occur in a given language/dialect, but they are not difficult to produce for speakers. Yet some other phoneme sequences can be quite challenging to produce for native speakers of languages in which they are disallowed.

    I think this is probably related to another observation I have made about accents. In first or second grade, I remember my teacher saying that having a different accent meant that someone pronounced certain sounds in a slightly different way. This clearly wasn’t wrong, but it did not fit with the way I thought about accents at the time. (I played around with accents from a very young age, especially English, German, and north Indian.) It made it sound like an accent could have unrelated differences in its realizations of all its phonemes, which did not feel quite right to me. Rather, there seemed to be a correlation between the peculiarities of most or all of the phonemes in a given accent. Eventually (after about a decade of occasional consideration) I decided that the issue was that different accents usually have slightly different ways of opening the whole vocal track. The “neutral” position of the mouth is slightly different for speakers with different accents, and phonemes are largely formed by variations around that neutral position. This gives a distinctive cast to all the sounds a speaker utters, even those that are non-phonemic, like grunts.

  101. I agree. I’d mentioned something like this here a few years back. I’ve occasionally passed for having a native Spanish accent, (though always from a vague somewhere other than the home of the native speaker I was talking to.) I find that when I hold my mouth a certain way, all the pieces of the accent fall into place.

  102. Lars Mathiesen says

    To further Brett’s evidence, I find that when speaking Swedish I push my lower jaw forward just a fraction of an inch compared to my Danish “voice”. (As a wild guess, that may be why Swedes find it almost impossible to pronounce Danish gade with the +ATR it needs).

    For German it’s more a difference of intonation. In singer’s terms, German uses chest voice and Danish uses head voice. My English is not very different from Danish, there are just many more voiced segments.

  103. David Marjanović says

    German uses chest voice

    I bet that varies a lot across German.

  104. Andrej Bjelaković says

    ” The “neutral” position of the mouth is slightly different for speakers with different accents, and phonemes are largely formed by variations around that neutral position. This gives a distinctive cast to all the sounds a speaker utters, even those that are non-phonemic, like grunts.”

    This is a very well known phenomenon, and it goes by various names, such as oral posture, articulatory setting, etc.
    Here’s a helpful video by the voice and speech coach Erich Singer:

    There’s also a relatively well known book by the phonetician John Laver, The Phonetic Description of Voice Quality, which deals with this issue among others.

  105. Lars Mathiesen says

    varies: I mean when I speak German, trying to sound like the Germans I’ve met.

  106. PlasticPaddy says

    @lars, dm
    I think “chest voice” is most apparent in the dipthong “au”, it is the difference between typical pronunciations of Dutch “vrouw” and German “Frau” or between English how and German hau…

  107. On the other other hand, not many English speakers can manage … even a word-initial [ŋ] … perhaps if we had more loanwords from Dagbani …

    Or Māori: Ngaire Marsh (the actor/author), Ngarahoe a mountain (that’s 4 syllables). In fact in Māori (and other Pacifica languages), ng- should always be pronounced syllable-initial: To-nga, Wha-nga-nui

  108. David Marjanović says

    I think “chest voice” is most apparent in the dipthong “au”, it is the difference between typical pronunciations of Dutch “vrouw” and German “Frau” or between English how and German hau…

    The difference I find salient between English and most of German is that the English MOUTH vowel starts with an actual front [a] (except in accents where it’s moved to an arguably even fronter [æ]), while the German version mostly starts with a central vowel even if /a/ is front in the same accent.

  109. David Eddyshaw says


    Yes, I was thinking about Ngaio Marsh (and secretly hoping that you would turn up with some Maori …)

    Do New Zealanders who can’t speak Maori nevertheless have enhanced powers of pronouncing word-initial /ŋ/?
    (The presence in personal names gives me some hope that it might be so …)

  110. DM: I think the MOUTH vowel is what I found most striking in he accent of the singer Nico. The word clown in The VU’s femme fatale comes to mind. It might even be a back vowel, not a central one.

  111. >This is a very well known phenomenon


  112. @DE (thanks for correcting my spelling: yes there’s lots of personal names starting Ng-. The actor Ngaire is Dawn-Porter; I mixed that one up.)

    Do New Zealanders who can’t speak Maori nevertheless have enhanced powers of pronouncing word-initial /ŋ/ ?

    They certainly don’t have enhanced powers of pronouncing word-medial /ŋ/. So it’s Toŋ-a, woŋ-a-nooey (the word-initial Wh- in the spelling should be pronounced more like aspirated f- or bilabial fricative — but we had a whole acrimonious debate about whether there should even be a ‘silent’ h in the spelling, let alone whether it should get pronounced other than w-. And this all going on in which language where silent h following w appears in all sorts of spellings? But see also N.Island Dialects at the wiki. Jeepers it’s a minefield.)

    It’s complicated: There’s South Island dialects of Māori where the /ŋ/ has morphed into k. So (N.Island) Ra-ngi = ‘cloud’ becomes S.Island Ra-ki, as in Aoraki (Mt Cook) the cloud-piercer.

    A very commonly-occurring word in government/governance circles is Ngai or Ngati meaning clan or tribe (approximately — I’m committing cultural howlers even drawing that equivalence). So Politicians and Officials generally manage to pronounce those.

    Less frequent words tend to get pronounced with a vaguely nasalised g- or k-. Or just n- . I suppose that’s ok. In Tūhoe and the Eastern Bay of Plenty (northeastern North Island) ⟨ng⟩ has merged with ⟨n⟩. says the wiki. I think they’re in complementary distribution.

  113. David Marjanović says

    I think the MOUTH vowel is what I found most striking in he accent of the singer Nico.

    Never heard of him. Do you have a link?

  114. David Eddyshaw says

    It saddens me that the Young People of Today are unfamiliar with classical music.

  115. “Femme Fatale” (YouTube). I too am saddened.

  116. Stu Clayton says

    # Reviewer Richard Goldstein describes Nico as “half goddess, half icicle” and writes that her Velvet Underground vocal “sounds something like a cello getting up in the morning”.[18] #

  117. @Andrej Bjelaković: Thanks! I figured it must be a well-known phenomenon, but without knowing the right terminology, it was hard to search for evidence of it.

    While searching the Internet for information can be very useful, it’s much harder to use it to find the correct wording for something you can only describe.* There is the example here, with articulatory setting, and another one I’ve encountered recently is with misheard song lyrics. There’s a catchy tune with a chorus that repeats what sounds like “cenote” a bunch of times. Since oceanic cenotes are a running joke with me and my sons, I would like to know what song that was, but it hasn’t been easy to find.

    * Google and other search engines are good at finding popular sites that match search terms, but they are not able to field more grammatically complicated questions, like, “What is it called when you…?” Ask Jeeves advertised itself as using the Internet to answer users’ questions, but it really didn’t and just searched for the terms that made up each query.

  118. Stu Clayton says

    Blue C-Note


    You did it for the broads, I did it for the paper
    I got a hundred styles, got a hundred flavors,
    Fuck the police, hundred middle fingers,
    And you a square nigga, blue c-notes
    All of ’em singin’, blue c-notes
    All of ’em Franklins, blue c-notes
    Spend without thinkin’, ice on both pinkies
    Mobile phone ringin’, blue c-notes


  119. John Cowan says


    Which is still better, I suppose, than the rest of the Anglosphere’s Toŋ-ga.

  120. “Still better”? There’s nothing wrong with it at all. It is reasonable to expect a decent approximation to a name in a different language than one’s own; it is unreasonable, and in fact fanatical, to expect or demand exact duplication, even when that violates the sound structure of one’s own language. If an English-speaker chooses to make the effort to learn to pronounce syllable-initial /ŋ/ (or a Welsh lateral fricative or any other such exercise), that’s a fine thing, but it’s a supererogatory virtue.

  121. David Marjanović says

    Well, I’m unfamiliar with most music, and I didn’t grow up with Nico’s because she died when I was 6…

  122. Not your fault at all; we’re just exclaiming “eheu fugaces, Postume, Postume…”

  123. David Marjanović says

    By “I’m unfamiliar with most music” I mean I’m not representative – if the rest of my generation is familiar with Femme Fatale, I wouldn’t necessarily even know. (Though they clearly don’t talk about it a lot.)

    What is unsettling is the young colleague who didn’t know who Luke’s father was.

  124. Now then, DM, what do you think of Nico’s pronunciation of down and clown?

  125. >it saddens me that the Young People of today are unfamiliar…

    I was young a mere thirty years ago and the closest I come is this obscure tribute to the ancients.

  126. David Eddyshaw says

    What is this “REM” of which you speak, child? Is it some sort of Velvet Underground tribute band?

  127. An Erich Maria Remarque cover band, perhaps?

  128. David Marjanović says

    Now then, DM, what do you think of Nico’s pronunciation of down and clown?

    Oh! Yesterday I was so tired I forgot to watch the video. 🙂 The pronunciation is very odd – it’s just “ah”, not a diphthong at all. I don’t have a geographic explanation for it.

  129. You can hear a rounded off-glide at the end, if you listen closely. But yes, that’s not something I have heard in any German regiolect.
    On the other hand, in this live version , she has a bog-standard German “au” (at 00:43).

  130. David Marjanović says

    In down, yes. In clown she also has a diphthong, but she cuts it off before she reaches [ʊ] – it only gets to [ɒ] or thereabouts.

    For eu it’s common to end before it reaches [ɪ], i.e. around [ɛ]; but for au to end early, let alone so early, is not something I’m used to.

  131. “she also has a diphthong, but she cuts it off…”

    Such a protophoneme would look good in reconstructions:)

  132. John Cowan says

    even when that violates the sound structure of one’s own language

    Which Toŋ-a does not (cf. siŋ-er), though To-ŋa does. We have discussed Lydia Nyongo’s name in this connection.

  133. Which Toŋ-a does not (cf. siŋ-er), though To-ŋa does.

    Which was exactly my point.

  134. David Marjanović says

    Lupita Nyong’o?

    Which Toŋ-a does not (cf. siŋ-er), though To-ŋa does.

    …Yes, but syllable boundaries are only audible if the syllables have different pitches (and everything between the vowels is voiced, which is the case here). They’re more a phonological than a phonetic phenomenon.

  135. ktschwarz says

    Stan Carey mentioned skew-ways again on his blog: Irishisms in City of Bohane.

  136. ktschwarz says

    mollymooly said: “regardless of whether the OED entry dates from 1933, it must have been revised at some point since 1977.”

    mollymooly, you should get on the subscription list to the Variorum OED if you aren’t already!

    1933 was the date when skew-whiff was published in the first Supplement, but more often than not, Burchfield’s Supplements expanded and improved on the 1933 entries. As you deduced, what you see under skew-whiff in the OED2 (and still online) is from his 1986 Se-Z Supplement. The 1933 Supplement had only two quotations, the earliest one from 1754 as shown, and an uninformative reference to “1896 Warw. Word-bk.” Burchfield dropped that one and added eleven more quotations, eight of them from actual use, not dictionaries.

  137. Edward Bear says

    It apparently means exactly the same as “askew” or “skewed”.

Speak Your Mind