Showing posts with label Tim Harford. Show all posts
Showing posts with label Tim Harford. Show all posts

Friday, 2 February 2018

Pedigree collapse

The other day my eye was caught by a BBC report:

Boris Johnson 'is descendant' of mummified Basel woman

more  here

Well GOLLY, I thought. A while ago More or Less dealt with a similar issue; in fact, the event was so Earth-shattering that Google will find it with the string Dyer Edward III.  1,560,000 hits. The Great Relatedness to Edward II Factoid leaves our Danny having to "take a moment" while the stupendousness (banality?) of the discovery sinks in. (Not quite banality – but what was extraordinary was not that he was a descendant: [after all, most people of English heritage are] but that his heritage could be documented.)

So when I heard about  the good woman of Basel's mummy I was underwhelmed. Huh, I scoffed, aren't we all descended from her? Well no; I had failed to consider the difference between the two cases: two or three centuries on the one hand, as opposed to  six or seven  – to give the Factors-of-2 magic its chance.

I was not the only sceptic. Stephen Fry tweeted:






But his tweet was met by a flurry of corrections (rebuttwals?): GNEURR Mr Smartypants Fry <eye-roll>, haven't you heard of pedigree collapse?






Nor had I. Fortunately, last week's More or Less explained (about 18 minutes in). Taking advantage, I guess, of  Stephen Fry's presence in Wogan House (or wherever) mentioned in this tweet...
 ⇒

... Tim Harford recorded a request from the genial polymath (from about 20'20") and went on to explain.


In his extreme case, if a man marries his sister they share a single maternal great grandmother (not four, as unrelated people have). Of course, very few of Edward III's descendants went in for this degree of inbreeding, but the general case is clear: the closer the consanguinity, the fewer the maternal great grandparents. This pedigree shows how 1st cousins marrying share only three maternal great grandparents.


When 1st Cousins M and F Marry

<digression>
This reminds me of the issue of pedigree's pedigree ... Hmm... [See PPS]
</digression>
But I'm missing this week's More or Less.

b

PS: A couple of clues:
  • I will do it again (despite this key skills deficit). (10)
  • Do they turn up trousers before droning on?. (8)
Update: 2018.02.05.12:30 – Added PPS on pedigree's pedigree

PPS  – As prefigured in my last digression., I've been looking into the derivation of pedigree. The version I was told 40+ years ago, and still favoured by many authorities (including the OED and Etymonline – but not having the wholesale  support of some current scholars ...
<example_scepticism source="Anatoly Liberman">
This offering from the OUPblog asks Does the current etymology of pedigree [pied de grue] have a leg to stand on?
</example_scepticism>
... ) is pied de grue. As Etymonline says
pedigree (n.)
early 15c., "genealogical table or chart," from Anglo-French pe de gru, a variant of Old French pied de gru "foot of a crane," from Latin pedem accusative of pes "foot" (from PIE root *ped- "foot") + gruem
According to this derivation, the pun with degree (in the sense of descent) was a happy accident of Middle English.

To quote that OUPblog again
In 1895 Charles Sweet, the brother of the famous Henry Sweet, and Round put forward the same explanation: according to them, the mark used in old pedigrees had the shape of a so-called broad arrow, that is, a vertical short line and two curved ones radiating from a common center, like three toes of a crane’s foot, with an allusion to the branching out of the descendants from the paternal stock.
So the jury's still out.

What we do know is that English crane and French  grue are related. Etymonline's entry for crane says:
Old English cran "large wading bird," common Germanic (cognates: Old Saxon krano, Old High German krano, German Kranich...from PIE *gere-no-, suffixed form of root *gere- (2) "to cry hoarsely," also the name of the crane (cognates: Greek geranos, Latin grus, Welsh garan, Lithuanian garnys "heron, stork"). Thus the name is perhaps an echo of its cry in ancient ears...
But what bothers me is this statememt from  the same entry:
Metaphoric use for "machine with a long arm" is first attested late 13c. (a sense also in equivalent words in German and Greek)...
And in French of course .  Scholars who question the crane's foot derivation of pedigree seem to be very concerned about the vowel: Isn't degree a more likely root for pedigree, preserving the /i:/ of degree? They seem to overlook the fact that the tongue positions for English /i:/ and French /y/ are very similar (if not identical –  many a language teacher uses this device for teaching /y/: Get ready to whistle and try to say /i:/); the main difference is in the lip-rounding.

OK, crane and grue are related. But if the first attested version of the lifting machine  dates from the 13th century, who thought of the metaphor first? Or is this a case of "convergent etymology".  Either an English builder thought
Pierre has named this machine after a bird that stands on one leg. That‘s a good idea. I‘ll do the same. Right, this hitherto unnamed device, which he calls une grue will henceforth be a crane
Or a French builder thought
Les Rozbifs ont donné à cette machine le nom d‘un oiseau qui se tient d‘une seule jambe. Formidable. Je ferai également. D‘accord: d‘ici en avant ce truc sans nomme, nommé outre-manche a crane sera une grue.
Well, no, that‘s silly. It‘s hard to imagine how people felt about language in the 13th century, when French, Latin, and English vied for... no, even that is an over-simplification. There was no "vying";  people just used whatever communication mode was suitable. And on a 13-century building site, with rulers setting impossible deadlines...
<Building_Times date="1250">
"I want a cathedral here in two years".  
"But sire, there  are not enough stone-masons in all Anglia."  
"Then get some from Francia. Two years I said.")
</Building_Times>
...the linguistic picture would have been very fluid.

b
Update: 2018.06.18.11:40 – The answers at last: ILLITERATE and WINDBAGS

Thursday, 9 January 2014

But I'm a good guy...

So you're allowed to cheat...?

In one of the many review of the year/predictions programmes aired in the past week, Tim Harford inveighed against infographorrhoea (he didn't call it that, but I can't quote him verbatim as I can't recall the programme's details): the tendency to splatter documents with eye-catching but meaningless and/or misleading infographics. (It was very funny and right on the money; I hope it'll make it to Pick of the Week.)  Ironically, as I write a fatcat developer is being interviewed on You and Yours, trying to defend just such an infographic.

A few days ago, I saw and retweeted (from the hip – will I never learn?) this:


It's making a good point. Government priorities are wrong-headed in a way that in less socially pregnant contexts would be laughable.  Stopping tax avoidance and evasion is the LOW-HANGING FRUIT – easy wins for a Chancellor needing to save a billion or two.

But having retweeted it I began to feel a bit uncertain about its integrity. The dark blue circle looked like a pretty dodgy quarter of the red circle. And even if the dark blue circle was the right size, the
light blue £16 billion circle, which should be slightly over half the size of the £30 billion one, was barely a third of the size (if that it's hard to tell, as the 'circle' isn't circular).

The little circles are so small that it's impossible to tell precisely what they are supposed to represent. The yellow one, which conveniently for my schoolboy arithmetic – should be one hundredth the size of the red one, is obviously far too small.

Just for the record, I had a bash at it with Visio. My additions aren't exact, as the configurations of our two graphics tools are bound not to match:


The arithmetic may not be perfect; in particular I ignored the possibility of negative roots for radii of circles, which –  in the view of someone who, to quote my maths master, 'took the soft option of becoming a Poet rather than a Plumber' (as for Doc Lewis there were only two sorts of people in the world, Plumbers and Poets) –  wouldn't make much sense in this context (though maybe they do in black holes...?). But my new circles are a lot closer to reflecting the truth of the statistics.

They do however fall into the trap they were supposed to, of taking the largest figure as a basis for the scale for the whole diagram. There are two scales here; a decent infographic would take the largest two circles as bases for two scales, and show the other three as proportions of each. This is grist for a possible future update. But I must get on otherwise the proofs of the hardcopy of #WVGTbook will arrive before I've got the index sorted out.


L'envoi (French for what what HR trainers call 'take-homes')


The fact that the bad guys fight dirty, using 'infographics' full of omissions and distortions, doesn't justify the good guys in doing the same. I'm pleased to add my weight to that of people looking for a fair deal for benefits claimants, but I regret retweeting an infographic that was more than just 'economical with the truth'; it was plain fallacious.

Report from the word face

I'm nearing the end of the initial trawl though the speech sounds used (in the one dictionary I'm using), ready to create an index (useful for the Kindle edition ['This phonemic sound can be represented by...', with immediate links to example words], but not for the hardcopy one [plenty of 'rhyming dictionaries' already on the market do the same job better]). The monophthongs are as given in the standard IPA 'phonemes of English' chart; well, they couldn't really not be. But the diphthongs, triphthongs, and indeed tetraphthongs  (if there is such a word) are far more numerous. I'm not sure how to present this, without muddying the waters for EFL/ESOL students...Hmm...

b

Update: 2014.01.12.17:30 – Added PS:
PS You'll notice from the footer that I have overridden the 'perhaps...?' and am working on an index. For the reasons I gave in my last para, though, I'm not sure what value it adds; so, rather than edit all the footers to make a claim that I'm still unsure about, I'm leaving it for now. (And if anyone can teach me to use pages rather than posts in Blogger – which I suspect would save me a lot of time – please feel free....)

Update: 2014.05.16.17:00 –Updated header:

 Mammon When Vowels Get Together V5.2: Collection of Kindle word-lists grouping different pronunciations of vowel-pairs. Now complete (that is, it covers all vowel pairs –  but there's still stuff to be done with it; an index, perhaps...?) 

And here it is: Digraphs and Diphthongs . The (partial) index has an entry for each vowel pair that can represent each monophthong phoneme. For example AE, EA and EE are by far the most common pairs ov vowels used to represent the /i:/ phoneme, but there are eight other possibilities. The index uses colour to give an idea of how common a spelling is, ranging from bright red to represent the most common to pale olive green to represent the least common.

Also available at Amazon: When Vowels Get Together: The paperback.

And if you have no objection to such promiscuity, Like this.

Freebies (Teaching resources: over 41.050 views  and 5,700 downloads to date**. They're very eclectic - mostly EFL and MFL, but one of the most popular is from KS4 History, dating from my PGCE, with nearly 2,100 views and nearly 1,000 downloads to date. So it's worth having a browse.)

** This figure includes the count of views for a single resource held in an account that I accidentally created many years ago.