Rate of Change

[Larry Trask]

--On Thursday, June 14, 2001 12:17 am +0000 X99Lynx at aol.com wrote:

[Joat Simeon]

> <<Eg., "pita" (Hindi) and "father" (English) are cognates, as are "pad"
> and "foot", despite having no surface similarity at all.>>

> If you said "pita" and "brassiere" have no surface similarity, well that
> would be different.  If you said "pad" and "philadelphia" have no surface
> similarity, that might work.  But "pita" and "father" clearly do have some
> prima facie structural similarities.  All these similarities and
> differences are clearly a matter of degree.  I would suspect that "found"
> cognates that truly have no recognizable resemblances are rare.

Well, maybe not that rare.  Romance is a rather shallow family, but look at
these cognates:

	French	Italian

	/u/	/agosto/	'August'
	/o/	/akkwa/	'water'
	/wazo/	/uttSello/	'bird'
	/Sa~/	/kampo/	'field'
	/SEf/	/kapo/ 	'head'
	/SE~/	/kane/	'dog'

And so on.  French /o/ 'water' arguably resembles the unrelated Basque <ur>
'water' much more closely than it does its Italian cognate, and Italian
/agosto/ 'August' arguably resembles the unrelated Basque <agorril>
'August' more closely than it does its French cognate.

I suspect that dissimilar-looking cognates are not particularly rare at
all.  After all, English 'head' is cognate with both the Romance words for
'head' given above, but it really does not resemble either.

But note that the last three words, while exhibiting little in the way of
resemblance, exhibit a systematic correspondence rather well.

> And I would suspect that any attempt to find a systematic pattern of change
> that included "pita" and "brassiere" as cognates would not get very far.

Oh, I beg to differ -- semantics aside.  First, the long form 'brassiere'
is now pretty much obsolete, and unknown to younger speakers.  So let's try
<pita> and <bra>.  An initial bilabial plosive, followed by a non-nasal
coronal, followed by /a/ -- by some standards, this is a very high degree
of similarity.  I know one or two not very respectable linguists who would
not hesitate to reconstruct *<pta>, with vowel epenthesis in one language,
and the changes *<pta> --> *<pra> --> <bra> in the other.  Of course, that
element <-ssiere> is just an accretion, and not to be taken seriously, like
the final elements of Spanish <cabeza> 'head' and <corazon> 'heart'.  ;-)

Now, the semantics of 'father' and 'bra' is a challenge, but perhaps the
most important Roman bra-factory was located in or near the Temple of Jove
the Father. ;-)

Sorry, folks -- I'm always like this at the end of the exam-marking period.
I hope to be back to my normal gruff self within a week.

Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK

larryt at cogs.susx.ac.uk

Tel: (01273)-678693 (from UK); +44-1273-678693 (from abroad)
Fax: (01273)-671320 (from UK); +44-1273-671320 (from abroad)