Friday, 28 February 2020

Flipping (pesky?) bits

On 8 May 2019 Radiolab broached the subject of bit-flips (when a 1 or a 0  becomes a 0 or a 1), while listeners in the UK had the luxury of ignoring the still distant Brexit shenanigans of 31 Octob... sorry, January 2020 (in the vain hope that something would deliver us from the nightmare). So this Radiolab offering didn't air (sorry) on the BBC until February 2020.
<RANT>
And the latest is that unless the EU rolls over by June Boris is going to stamp his little foot and spend 6 months in intensive no deal planning, handing over the reins to the disaster capitalists of the far right.This really is the negotiating tactic of the schoolyard – if not the playgroup.
</RANT>
But about those bit-flips. In the words of the Radiolab blurb
Back in 2003 Belgium was holding a national election. One of their first where the votes would be cast and counted on computers. Thousands of hours of preparation went into making it unhackable. And when the day of the vote came, everything seemed to have gone well. That was, until a cosmic chain of events caused a single bit to flip and called the outcome into question.

Today on Radiolab, we travel from a voting booth in Brussels to the driver's seat of a runaway car in the Carolinas, exploring the massive effects tiny bits of stardust can have on us unwitting humans. 
The programme is worth a listen if you've the time, but I'll cut to the chase. In that Belgian election one of the voting machines gave an impossible result; more votes were cast in favour of one candidate than voters who'd used that machine.

There was a manual recount, and before you could say But I thought the Cold War was  over (it was the Communist candidate who'd had the Putinesque...
<ASIDE silliness ="11">
So that's what the name of that pasta dish means...
 </ASIDE>
...success) the IT people noticed that the difference between the manual result and the digital one was exactly 213 (or, as they said on the programme, 4096).

Not being a natural thinker in binary, but having some experience of the world of bits and bytes, when I heard "4096" I thought  Hang on, that's 4K' ...
<ROUGH_EQUIVALENTS>
A "binary thousand" is, in its approximation to 1000, a bit like a "metric mile" (or 1500 metres); that is, not. But close enough. Since 200 metres is just over a furlong, 1600 metres would be a lot closer.
<CALCULATION>
A metre is 4 inches more than a metreyard, so 9 metres is a yard longer than 9 yards (i.e. 10 yards). So 100 metres is 111 yards and change. People of a certain vintage will know a furlong is 220 yards, and there are 8 furlongs in a mile. So 1600 metres is just over a mile.
</CALCULATION>
But sports commentators still call the 1500 metre race "the metric mile".
</ROUGH_EQUIVALENTS>
What caused the bit to flip was a cosmic ray; and it happens all the time, often in a benign way (causing a glitch that can safely be ignored). Computer engineers know this, and know that there is no hardware "shield" that can save chips from this threat...
<CLOUD_CHAMBER>
Cosmic rays don't care about physical objects standing in the way.  The particles themselves are much too small to be seen directly, but there are several YouTube clips with instructions for making a cloud chamber – which lets you see the trails they leave (on the Radiolab  programme they liken them to the con trails that a jet aircraft leaves behind.)
</CLOUD_CHAMBER>
...Bit-flips happens more at higher altitudes (where the cosmic rays have higher energy); and it takes much less energy to flip bits in modern (miniaturized) chips.

So we're screwed, though not entirely. Where precision is a matter of life or death (say, in an aircraft's electrics) there is a way round the problem: redundancy. A life-or-death component is typically duplicated twice; and output commands (such as raise the nose or cut the engines) are voted on. Only a majority vote triggers action.

But redundancy isn't popular among designers of consumer electronics; who wants a smartphone big enough to house a lot of redundant processors?  Some bit-flips – a majority of them? – are benign, and usually go unnoticed. And, as for the rest, we'll just have to get used to the occasional computer blip (in fact, we already are accustomed to computers that misbehave; we just have to learn not to be so surprised when they do). Or (here's an idea) DO WITHOUT THE  WRETCHED THINGS.

Time I did a bit of prep for this:


Bye for now.

b

Update: 2020.04.30.12:10 – A couple of typo fixes. (Does nobody proof-read these things?)


.




Tuesday, 18 February 2020

Quarantine and epicentre

People who know a bit about the formation of languages learn  that they must beware of the Etymological Fallacy, which this Oxford Reference page defines as
The belief that an earlier or the earliest meaning of a word is necessarily the right one. That it is fallacious is illustrated by the fact that orchard once meant a treeless garden, treacle a wild beast, and villain a farm labourer.
The creeping Wikipedi-ization of the modern world has exposed us more and more to this tendency to hold that words must mean what they used to mean, which implies that meanings must never change. Still, a knowledge of where words come from can be fascinating.

The word "quarantine" has had a long history, originally referring to a period of isolation at the time of The Plague (well  one of the waves of  one of them – it was all a bit muddled back them, TBH).

This page dates it (with annoying vagueness – the site is, after all, designed to drum up business for English Language summer schools, so it would be unreasonable to expect much in the way off academic rigour) to "a document from 1377"  though in an  earlier form that set the safeness-from-plague period to thirty days – una trentina. In an earlier post I briefly referred to "quarantine"‘s ship-of-possible-carriers origin, but I didn't place the first use as coined by Venetians trying to keep the plague out of Dubrovnik (as this page does if you care to read it through):
The word “quarantine” has its origins in the devastating plague, the so-called Black Death, which swept across Europe in the 14th century, wiping out around 30% of Europe’s population. It comes from the Venetian dialect form of the Italian words “quaranta giorni”, or “forty days”, in reference to the fact that, in an effort to halt the spread of the plague, ships were put into isolation on nearby islands for a forty-day period before those on board were allowed ashore.  
Source
My ingrained cynicism about the temptations of the Etymological Fallacy doesn't,  however,  prevent me from experiencing a frisson of "innocent merriment" ...
<PC_defence degree="set phasers to stun">
(at the risk of seeming to make light of a notably unfunny situation)
</PC_defence>
... when the Diamond Princess quarantine forced the word back to its roots (with people not  being allowed to leave a ship).

My other concern at the moment is the snowballing over-use of epicentre. Again, I'm not trying to argue that the word can only be used between consenting vulcanologists ...
<spelling_for_dummies>
(and the infernal machine wants me to write "volcanologists", but quod scripsi scripsi)
</spelling_for_dummies>
... but I just feel my lip curling whenever people use a big word just to make them sound serious. What's wrong with centre? What's wrong with focus?...
<digression>
There's that innocent merriment again. Focus, being derived from the word for fire, seems particularly apt when talking about an infection that causes, among other things, fever.
</digression>
...hub..., source ... There are many ways of avoiding epicentre; but it continues on its juggernaut way. This view of its growth in popularity (sadly out of date – the  latest data they've got is from 2008, and in the last twelve years its use can only have grown) comes from Collins:

 Anyway, it's time I got back to note-bashing for  our next concert:

<incidental_observation>
A fellow choir member has asked about "Jesu chare"  in the Pergolesi Buxtehude  – not having found chare in a latin dictionary.  There are two problems with this search. The first is that the phrase is vocative – addressing Jesus. "Dear Jesus" would be, in the nominative (just naming him), Jesus carus. The spelling of that second word points to the second problem. This recalls my "epicentre" rant; one of the mechanisms of language change is hypercorrection (trying to sound important by a misplaced display of "learning"). The introduction of h after c ...
<inline_ps>
I'm not referring to a /h/ sound following the /k/. The hypercorrect change is from /k/ to /χ/ (like the sound at the end of Bach). Trimalchio – that* character in the Satyricon – is (unwittingly) referring to the influence of Greek sounds on Latin. Greek slaves were common in the Roman world,  and there were Greek-speaking enclaves in what we now know as Italy.
</inline_ps> 
...is usually hypercorrect; this is satirized as early as the first century AD in the Satyricon; which has a social-climbing character who is mocked for saying chommodus rather than commodus; Oops, TMI.
</incidental_observation>

Anyway, it's lovely music. Don't miss it.

b

Update: 2020.02.18.15:45 – Added inline PS.

Update: 2020.02.19.12:15 – Added footnote

* Oh what a tangled web we weave
   When first practice to update with an inline PS

That character is mentioned later on. Sorry.

Update: 2020.02.21.10:20 – Corrected the composer; it was Buxtehude


Sunday, 9 February 2020

Where's the camera?

The many-BAFTA'd film 1917 features, about halfway through, a dog fight from which I've taken an illicit still showing the moment where my eyebrow jerked up (only one eyebrow – this is cynicism, not surprise).
Moment of untruth?
A flaming wreck is hurtling towards the soldier in the open at... what?  100 mph?...
<guess>
This a guess, but it can't have been doing less – possibly a lot more
</guess>
...and he turns and runs along the exact path of the crash . The plane is obviously disabled, so it isn't going to veer off that path. Presumably he thought he could outrun it – quick on their pins, these Tommies; or maybe he just had a death wish. Maybe, though, he knew the camera was behind his mucker in the barn, and wanted to be sure of creating just the right composition (which wouldn't have resulted if he'd done the sensible thing and run off on a perpendicular from the path of the doomed plane).

Like a film scene I commented on here nearly 6 years ago, what mattered was the visual effect rather than any attempt at verisimilitude.

A not entirely unrelated case ...
<different_though>
(totally different, though; the similarity is only that the director's wishes for a satisfying visual composition were more important than any questions or doubts raised in the mind of the viewer)
</different_though>
...cropped up recently in the last episode of the Wisting series on BBC Four.
<spoiler_alert>
And if you are that way inclined, now's the time to bail out. Exposition of a cinematic cliché follows.
</spoiler_alert>
A lone policeman approaches a suspect 4x4, which conveniently enough has windows lighting the boot (not the ideal vehicle for a psychopathic kidnapper who from time to time uses the boot for purposes of  victim-conveyance),
Then the camera angle changes, so that it is inside looking out, as the policeman goes through a basic kidnapper's starter kit  (duck tape, plastic cable ties..., you know the drill).
At this point I started to wonder what could have been the point of this change. But I didn't have to wait long.

Right on cue, the aforementioned psychopath appears behind the policeman and knocks him out (but doesn't think to use his kidnapper's starter kit, to stop him butting in to the next scene).

There are times when cinematic cliché interferes with the story.

<autobiographical_note>
I'm reminded (irrelevantly of course, as this wasn't the director's fault). In the early 1970s I saw, at the Arts Cinema in Cambridge, Chabrol's Le Boucher. In one scene suspense was being ratcheted up by the images and the soundtrack –  I have no clear memory of the details; I think someone was looking for trouble in a gloomy outhouse. But what broke the tension was not Chabrol's soundtrack (a voice from out of the shadows, as I remember). It was the subtitle ( a case of legendum praecox?)
</autobiographical_note>

But I'm missing the cricket.


Friday, 31 January 2020

Hunting and pecking

There are things about smartphones that bother me when I see them in use. (I'm not a user myself, you understand: Ask not for whom the bell tolls, it tolls for islands of self-absorption who avoid eye-contact and court Repetitive Strain Injuries. The temptation would be too great.

Chief among these, apart from the standing invitation to be anti-social, is what happens to users' thumbs. Cradling the phone in the fingers of both hands and typing with two thumbs can lead to stenosing tenosynovitis, or ‘Trigger Thumb‘ (or even – less cryptically...
<aside subject="Trigger Thumb">
Trigger Thumb gets its name from the typical physical jerk and popping sound as the joint moves into/out of place.
<philological_observation>
When only specialists (in this case, specialists in orthopædics) have need of a term, they use what suits their needs – often leaving lay people wondering what they're on about. 
When an idea gets a wider use, as the needs of the users have changed, the term changes to reflect a new focus. "Trigger" Thumb referred to a characteristic sign of the pathology – what a diagnostician should look for; in medical terms, a sign (what an observer sees) rather than a symptom (what a patient feels). So when the same thing started being felt by a wider range of users, a more specific term was needed.
<autobiographical_note>

When my mother (whom saints preserve, [and they better had]) was working at Metal Box in the late '60s, when ring-pull cans were in development, they were called –  in the language of Metal Box technicians 'easy-open ends'. What mattered to those technicians was only the end of the can: specifically, that it was easy to open. 
Obviously that clunky name had to change, and the marketing people came up (in the UK) with 'ring-pull can'. 
<shared_language>
In the US the cans had 'pull-tabs' I gather from Wikipedia.
</shared_language> 
The new expression reflected what was important for the users
</autobiographical_note>
</philological_observation>
</aside>
...– "Texting Thumb").

Winged Words

I've been asked about the derivation of lurgy. A Google search for etymology lurgy leads to confusing results:
1950s (originally spelled lurgi ): used in the British radio series The Goon Show and probably invented by its writers, though possibly from an English dialect term.

So...? What's confusing about that? Well, click on the arrow for further information and you get this:

A 1950s coining with recorded usage going back to the beginning of the 19th century.

The word strikes me as owing something to India (look at the menu of an Indian restaurant; 'urg' is the sort of word-bit (that's morpheme, if you want the $10 word) you'd expect; murgh is Hindi for chicken, so it is not uncommon in that context. But my questioner said that the word cropped up in the context of an Indian asking what the word lurgy meant.

Lurgy/i was a dialect word that referred to laziness. Oxford's Lexico says this:

This was the word that Spike Milligan adopted for the Goon Show.

Spike Milligan was born in India, and in any case may have been exposed to *urg* words during his army service (and in a lifetime of Indian restaurants).  It doesn't seem to me impossible that these Indian influences led him to adopt an existing word that sounded somehow Indian (and looked it, in its lurgi guise).

That's what I think anyway; it's time to go though.

b

Update: 2020.02.01.12.30 – Added PS

PS Apropos of nothing, I've just heard on the radio yet another mis-stressing of Così fan tutte and I've thought that rather than just wincing (after all, the accent is there, and you don't need to be a professor of Italian to realize that it must do something) I should publicize this mnemonic: così means "like that" and it is stressed like that ("like that" – co). OK, as you were, it's now safe to go back to Classic FM.

Update: 2020.05.18.12.30 – Added missing bit of sentence in blue. Sorry.

Tuesday, 14 January 2020

That's one way of looking at it

I've been thinking about – among other things (hence my failure to add to the Harmless Drudgery mountain (slag heap?) for the best part of  two weeks) – obsession. In particular I've been thinking about what obsession, gunwale, and titanic have in  common.

People with an etymological bent (sic – I‘m reminded of Joni  Mitchell‘s ‘That  girl is twisted‘) – will be familiar with the random arising of questions such as "What‘s obsession got to do with sitting?‘
<grandmothers_egg_sucking>
(And I know it should be grandmothers', but the compiler wouldn't be able to handle apostrophes; I know the compiler is a figment of my imagination, but if a conceit‘s worth pursuing it‘s worth pursuing to the last syllabub of recorded time.) 
Words built from some variant of "session" (not -cession, which is a whole 'nother kettle of worms) include somewhere along the line the idea of sitting. At its simplest, for example, a court that  is  ‘in session‘  is sitting.  Session musicians "sit in". And the Holy See involves sitting on a particular sort of chair (whence the building that houses it, a cathedral, gets its name). 
<RC_note>
Ex cathedra pronouncements are reserved for when the Pope Really Really Means It.
<RC_note> 
Etymonline‘s entry for "obsess" explains further:
</grandmothers_egg_sucking>
So besieging – the now obsolete meaning of obsession – involves an army encircling a town and just sitting it out. (I discussed words to do with sitting a while ago, here.) If you think of the poor besieged  townsfolk,  who can‘t get anywhere, by any path, without coming up against the besieging force, you can see where the modern sense comes from: any thought leads to the same place  – the obsession (the besieging enemy).

So ob and sedere got together to concoct the military  meaning, and psychotherapy took the ball and ran with it – so successfully that the ‘besieging‘ idea withered on the vine: obsession isn't just a metaphor; it‘s a metaphor that turned into another metaphor with a totally different meaning.

Which brings me to  gunwale, commonly reduced to gunnel. Most  native speakers of English have met the expression "laden to the gunnels" (OK make that ‘about a third of us, with the other two thirds saying "packed to the gunnels"). The gunnel is a wide plank at the edge of the deck, and if a ship is laden to the gunnels its load is so heavy that the ship  is low in the water.

But if you peel back the superficial metaphor (and when I said "plank" I was giving the game away, as today wood need not  be involved and often isn't) you find out what the gun is doing. On a sailing ship intended for battle the edge of the deck was reinforced with an especially sturdy plank, which supported the cannon – the gun wale.
<further_reading>
Pick the bones out of this if you're interested in the wale bit.
</further_reading>
Like obsession, gunnel (in expressions such as "laden/full to the gunnels") started life as a metaphor, and was pressed into use as another totally different metaphor.

You can probably see where I‘m going with this; the story with titanic is similar. I‘ll just sketch out the bare bones:

Titans (powerful gods) 👉 titanic (=big and powerful) 👉 Titanic: big/powerful ship

Along comes an iceberg and one metaphor gets flipped on its head to make another: something that‘s titanic can either be big/strong/influential ("a titanic struggle") or it can have a capital T and be over-confident and doomed to failure.

There must be more such metaphors  that have been given a new lease of life as newly formed metaphors, but this has gone on long enough...
<autobiographical_note>
(as has this accursed backup)
</autobiographical_note>
.. and I  must return to the land  of the living.

b



Thursday, 2 January 2020

Once bittern, twice shrike.

Boom boom. That is all.

Nearly 7 years ago, with the chutzpah of a fairly sprightly (at the time) recent-notcher-up of the big 60, I wrote this:
I had a difference of opinion with MrsK the other day. We were in the seventh circle of la città dolente, or PC World as it is more commonly known, looking for a new laptop. In defence of one I pointed out that it did not have Windows 8 (which to me made it preferable). She wanted to know why this was an advantage, and I said that with any new operating system there's more to go wrong; tried and trusted software is no longer supported. 
This was further evidence of my defeatism, she said. Why expect things to go wrong? She asked a passing school-leaver if there were any known support issues with application software (I'm paraphrasing here, you understand) and the answer was, surprisingly enough, that everything was hotsy-totsy with Windows 8. 
Well, twenty years of working with software engineers (actually, 19¾ - HP took the penny-pinching precaution of shafting me 3 months before they would have had to fork out for a 20-year award) has taught me that if anything can go wrong with new software it will. This was true of Windows 95, and with everything since. Working in 'Support', which I did for many years, involved me almost daily in fixes and workarounds and you-just-can't-do-that-any-more when people tried to get existing application software to play nice with a new operating system. 
So everything, I feared, was not hotsy-totsy. To quote Ogden  Nash it was coldsy-toldsy (and Google, incidentally, has just asked me whether I mean 'cold toddy'). New operating systems are great when all the dependencies work, but with each new operating system there are more dependencies; there's more to go wrong. I hold no candle for Windows 7; give me Windows NT 4.1 any day. But for me it's preferable to Windows 8 (just as Windows 8 will be preferable to - saints preserve us - Windows 9).
My fear in the last line was ill-founded. For reasons best known to the Microsoft marketing department, 'Windows 9' is The Operating System that Never Was. I wonder why... It's not as if it were Windows 13, or Windows 666 (due in some future century, perhaps)..

But earlier in 2019, Microsoft decided to pull the plug on Windows 7 (the home, at the moment, of Knowles family computing – although there are outposts of more recent operating systems on various less benighted devices). On some date in mid-January 2020 support will be withdrawn. This could be relatively painless, if things just stop being fixed. I suppose Anti-Virus software could be a problem. But, realistically, what are the odds against  hackers bothering with a ten-year-old system? We revert to the old tried and tested system: Security-by-Obscurity – as they used to say in the world of VMS.
<OpenVMS>
Not that that was the whole story. OpenVMS was a much more robust system than some I could name, and not a prey to nearly as many viruses as we have become inured to in the M$ world. But the fact remains that being surrounded by lower-hanging fruit, malware-wise, is a fairly good guarantee against infection.
</OpenVMS>

A less benign outcome seems possible though, in the light of the behaviour of tech firms in recent years, punishing users who are content with things as they are. For the TEMERITY of not upgrading, they are deprived of even what they had; that'll teach them not to genuflect at the altar of Perpetual Upgrades. Per me si va nella città d' aggiornamenti perpetui (which is the way I imagine Dante might have said perpetual upgrades).
<digression>
Hmm. What a charming way Italian has with the word for 'upgrade': aggiornare. Think of giorno. A bit like the Creation, in Haydn's version: And there was [wait for it] light.
</digression>

Still. Sufficient unto the day is the evil thereof. Perhaps everything will be hotsy-totsy, as St Matthew might have put it.  And I have two weeks to find out.

<seasonal_novelty>
I'll sign off with something that came to me recently. It may have been the most middle-class of cracker jokes, remembered from Christmas 2018, or it may be original (not to say meaningless to most):
Q. What sort of hair conditioner does Santa Claus use?
A. Ho-ho-ho-ba. [Bou-boum and indeed tsh].
Like I said, meaningless to users of the – possibly more common,  – /ʤә'ʤәʊbә/ version (which is fine by me if that's what fleauts your beaut  – I'm not going to kick off the new year by laying down the law about talking proper (although  I haven't found a dictionary that recognizes the /ʤ/ version – just saying).
</seasonal_novelty>

Frohes thingummy.

b

Thursday, 19 December 2019

It depends what you mean by "bee" and "bonnet"

The latest issue of Third Age Matters has the usual contribution by professional stick-in-the-mud Eric Midwinter, to which a sub-editor has contributed the ridiculous idea of theft (as if a meaning, whatever that may be, could be "stolen"):

For many years...
<digression>
(I was about to write "sixty-odd", but that would imply that when the word was coined in 1958, I [extraordinarily gifted child that I was] immediately adopted it into my seven-year-old's working vocabulary; improbable, I think)
</digression>
...I have used the word (ABused it, from the point of view of Mr Midwinter) in the way suggested by the New York Times interview quoted by Etymonline














It's a quite simple bit of word formation. As aristocracy is rule by the best (aristoi) and theocracy is rule by God (theos), meritocracy is rule by people judged on merit. But nous avons changé tout cela as le bourgeois gentilhomme was told, and as Mr Midwinter would have it; strictly, though, the Midwinter version would be "almost everyone has changed it, but they're wrong".

Michael Young, writing in an introduction to the 1994 edition of his work, wrote:


The case against meritocracy (the one and only true meaning of meritocracy, in the view of Midwinter and his ilk) is that people who get into a position of power on the basis of merit then protect their progeny regardless of merit. In other words, meritocracy works once, and thereafter arrogates power and influence to the privileged, regardless of merit. To insist that this flawed sort of one-time meritocracy-followed-necessarily-by-mediocracy-in-perpetuity is the only sort – the TRUE meaning of "meritocracy" – deprives the word of any useful meaning.

Over time, meanings change. Here I give lots of instances of  change, sometimes complete reversals, finally getting round to this explanation  of "backlog":
The words back and log were first fused together (to use an appropriately fiery metaphor) in the late seventeenth century. They referred to a log placed at the back of a fire. Such a log was desirable; it was a Good Thing. It protected the fire from going out. But about two hundred years later it was used metaphorically to mean a Good Thing in the commercial world: a stock of unfulfilled orders. 
Here's where the reversal in polarity happened, possibly influenced by another meaning of log. The metaphorical ledger ... could be the record of a Bad Thing – work that hasn't been done and gets more and more embarrassing as more is added to the mountain faster than it can be done.

But no sane observer scans the airwaves, searching obsessively for people using "backlog" to refer to a Bad Thing and complaining that they've stolen its meaning. Like Aung Sang Suu Kyi, flipping from fêted to fœtid overnight.

Mr Midwinter thinks "meritocracy" is another case of such a flip – but from Bad Thing to Good Thing, rather than vice versa. I think he's simply wrong; that "meritocracy" does what it says on the tin,  and the satirical background to the first use of the word doesn't affect that meaning the notion that the word can only ever be used to refer to one side of a multi-faceted argument in the original work of fiction that gave it its first airing more than 60 years ago is frankly ridiculous.

(Not that this prevents Mr Midwinter from writing to any organ that will publish him, accusing various public figures of this "abuse":   with the search string Eric Midwinter meritocracy I get over 14,000 Google  hits –  never, in the history of human apiculture have so many bonnets attracted so many bees, vainly swatted at by so few).

On a more festive note, I'm reminded of words I sang in last Saturday's concert, in Ralph Vaughan Williams' charming Fantasia on Christmas Carols:

God bless our generation who live both far and near.

This was not Roger Daltrey's My G-G-G-Generation. I was not singing God bless all baby-boomers  but nobody else. Words change their meanings, and when someone's  meaning is clear  it's unproductive  not to say patently absurd  to insist that they are among the great majority who are all out of step.

b

Update: 2019.12.20.12:50  Added PS

PS I have belatedly found a source for Midwinter's article online. Here's a taste:
It is just over 60 years since Michael Young, co-founder of our marvellous U3A movement, published his perceptive and prophetic text The Rise of the Meritocracy.
He had foreseen that social mobility worked fully only once, recruiting a new oligarchy of power, influence and wealth which, like all former ruling castes, then ensured the power, influence and wealth was bestowed on its children. 
Social mobility does not obey the law of gravity; very, very few go down the social ladder to be replaced by an urgent climber-up. It is not normal to find the children of the privileged among the homeless or the customers of food banks – and vice versa. 
I fear his word has been abused to the point of it sounding like a good thing, not least in the mouths of prime ministers, including a speech by Theresa May some months back in which she used the word “meritocracy” 20 times, with perhaps unconscious irony, as if it were something for which to aspire. 
This led me to write a correcting letter to the press and I quickly evade the charge of political bias by pointing out that Tony Blair was an even worse offender, to the point where a despairing Michael Young, a short time before his death, published an article pleading with the then prime minister to refrain from using the coinage “meritocracy” as the opposite of what it actually meant.