Letters playing leapfrog

... or 'metathesis' as we say in the business - it's what relates 'wrought' ('strong' past form of) to 'work' (displaced in more recent English by the regular  past form 'worked'), and is why Chaucer called birds bridde and a widow a widwe. On the radio this morning (or yesterday morning, by the time I publish - and OK, it was Woman's Hour) Nigella Lawson pronounced 'mascarpone' à la Delia Smith - /ma:skǝpǝʊni:/. And this pronunciation accounts for all the 'mask a pony' jokes - if you haven't heard any yet, you can easily add to the corpus (delicti?).

Now I'm not a stickler for pretentiously 'correct' pronunciation of foreign words. But even with purely British English phonemes the word would be /mæska:pǝʊni:/. So I'd expect Nigella Lawson, who studied for a year in Italy (many years ago - but she has recently given live TV/radio interviews in Italian to at least ('when I split an infinitive it stays goddam split') use an English pronunciation that respects the Italian spelling. 'Jumping Jesophohat', I thought, 'so Delia Smith's not the only one, she just happened to be the first I heard - maybe this is metathesis.'

In a not-so-recent discussion in the UsingEnglish forums. We were discussing the expression the dog's b@llocks. I suggested:

I've heard from a fairly reliable source (Stephen Fry I think) a derivation for 'dog's bollocks' that is not mentioned on that Phrase Finder page. Some commodity (it may have been the toy construction set, 'Meccano') was listed in a catalogue as '<whatever> - Box (standard)/ Box (deluxe)'. From this we get two idioms; "bog standard" and "dog's bollocks".

Two for the price of one - neat! I'm not sure I believe, but I'm impressed.

This was questioned: one contributor said the idea involved 'a linguistic jump', and I replied:

'Linguistic jump'? Speaking as a student of philology, I can say that it's hardly a jump at all. Consider the French guirlande and the Spanish grinalda. We can ignore the u, as it just keeps the g hard.

So we've got French

G + I + R + L + A + N + D + <unstressed final vowel>

versus Spanish

G + R + I + N + A + L + D + <unstressed final vowel>

The beginning and the end are the same, but four of the middle five phonemes are in different positions, and the only 'stable' one changes in quality (it's nasalized [in French, because the consonant that follows it has changed - clarification for blog]). In language development, phonemes jump about.

I first noticed this, before I embarked on this field of navel-gazing, in a sea shanty, Bring 'em down:

Up the coast to Vallipo,
Northward to Cally-o
Them Vallipo girls I do admire,
They set your riggin' all afire!
Them Vallipo girls puts on a show ...

I shall draw a veil (sail more like) over the details of the show. Valparaíso => "Vallipo";  Callao => "Cally-o". This leads me to a bit of home-grown folk etymology, which I expect to be dismissed by readers who decry the efforts of the imaginary body CANOE - the 'Committee to Ascribe a Naval Origin to Everything'^† ).

In a sea battle, the part of a ship where you could do most damage to an enemy ship was the poop (Spanish  popa) - the seat of 'intelligence'. A ship whose poop you had blown off could be said to be con ninguna popa - 'with no poop' - directionless, not unlike a dumb barge. Just give that phrase to the sailors who used 'Vallipo' for Valparaíso, and Bob's your hare-brained philologist.

However this is unproven, and - I think - unlikely. The n separated from p by an unstressed vowel could easily assimilate to the p, giving -mp- (because p is bilabial and so's m: there are about 400 English headwords* that  include the consonant  cluster /mp/ and few  - fewer than 100, more than half of which are 'un-' + <initial-p>' - that include /np/. [That volume of my dictionary - consonants - is a long way off but consonant clusters starting with a sonorant will be fairly high up the priority list.] Also, a few of the ones that have the spelling '-mp- don't have the phoneme 'p' - for example 'emphasis', 'emphysema'..., which have /mf/. And if an accident of word-building throws up such a cluster [as in 'unpleasant'] in informal speech the assimilation takes place, giving that word the same first syllable as 'umpire'; similarly a phrase like 'in present company' often starts with an /ɪmp/.) If anything, the derivation is suspect because there is too little change in the sounds. And g>c - that is, /g/ to /k/, is a change in the wrong direction.†

But I'll leave anyone who's that interested to do some research - starting from that Wikipedia link to Lenition  and pursuing this line of thinking to a more reliable source. In short, as in Latin acute ('sharply') being at the root of Spanish agudamente, a /k/ 'should' become a /g/ (and a /t/ a /d/ etc etc...), not vice versa - as for the -mente suffix, that would be a meta-digression. When I've got some reasonable progress on the book under my belt, though...it's extraordinarily interesting and well worth waiting for.  This digression (which doesn't involve metathesis) has already gone on far too long. As has this post - Christmas tree to light.

b

Update:1212.13:17.15 A few tweaks to the 'nincompoop' digression.)

Update: 2013.10.02.16:20
Footer updated.

Update: 2013.10.27.19:25
Added PS: 
*PS
.... in my dictionary of choice (but only because it is installed and lets me do useful wildcard searches): the Macmillan English Dictionary.

Update: 2014.04.22.15:05
Added PPS: Metathesis crops ups all over the place: I've been thinking recently about the derivation of Simnel Cake. Here's what Etymonline says:

"sweet cake," c.1200, from Old French simenel "fine wheat flour; flat bread cake, Lenten cake," probably by dissimilation from Vulgar Latin *siminellus (also source of Old High German semala "the finest wheat flour," German Semmel "a roll"), a diminutive of Latin simila "fine flour" (see semolina).

Just thought I'd mention it...

Update: 2014.09.29.14:15
Here's that -MENTE digression...

Update: 2014.10.02.14:25 – Added this PPPS
†PPPS I've come across what looks, on the face of it, like a counter-example (to the tendency of a /k/ to evolve into a /g/, rather than vice versa). When she was about a sixth of her present age (though she's never allowed to forget it –  aren't families wonderful?)  my daughter used to call spaghetti /pəsketi:/ (another case of metathesis, as it happens). The /g/ has become a /k/ though, not by a magical reversal of that lenition I linked to before but by another linguistic process – assimilation: briefly the voicing of the /g/ assimilates to the voicelessness of its new neighbour, /s/, and so becomes /k/.

Update: 2015.08.10.15:45 – Added this P⁴S

P⁴S A common source of metathesis in everyday speech is in the bastardization of borrowed words. People who should know better (#GBBO passim) often call crème pâtissière 'crème pâtisserie' , and you  don't have to listen to Classic FM too long to hear Cavallería Rusticana called /kævǝ'li:ri:ǝ.../. In both these cases, and often elsewhere, liquidity of phoneme-position (appropriately as /r/ is a liquid in the argot of phonologists) tends to favour phonemes near /r/.


Update: 2018.06.08.11:25 – Added this P⁵S

P⁵S
† This imaginary body is mooted in the penultimate paragraph of that piece on the derivation of brass monkies.

Update: 2018.10.03.09:25 – Added repair in red. (Over the years something got accidentally deleted.)