This is highly ingenious, and I'm perfectly happy to accept that we may be dealing with two historically distinct words, but the solution still has problems. Latin /vj/ normally gives /d3/ in French, as Keith notes. But it is hard to see how we can argue from the resultant /d3 + z/ (/z/ = analogical plural marker (quadrivia being plural already)? new nominative singular marker?) to the /ks/ required by the etymology suggested. The problem is specifically the requirement for /d3/ to become /k/.
For those names that can be assigned to quadrifurcus rather than quadrivia, I'm inclined to believe the traditional story. Keith is right to say that the /k/ would regularly disappear between /r/ and /z/~/s/, but "from the middle of the 12th C on .... there showed itself a marked tendency to introduce the final consonant of the uninflected radical before flexional s ..." (Pope, From Latin to Modern French, p. 462 (para. 1242)). So what might have been spelt <-furs> or similar before 1150 might afterwards come to be written <-furcs> or similar. The point is here that the <c> would have continued to be audible (or visible, at any rate) in non-nominative case-forms, and got transferred to the nominative (with its final <-s>). Either the reintroduction of the <c> would have been both orthographic and pronounced, or the new orthographical form would have come to have a spelling-pronunciation
That leaves us with the loss of the second /r/. For me that's much less of a problem; it's just the loss of the second of 2 /r/s in succession (as in "Febuary", but the other way round) - a normal kind of dissimilation.
Richard
--------------------------------------------------------------------------
Richard Coates
Professor of Linguistics ~ Professor of Onomastics and
Director of the Bristol Centre for Linguistics at UWE
http://www.uwe.ac.uk/hlss/llas/bcl/index.shtml
Hon. Director, Survey of English Place-Names
(w: http://www.nottingham.ac.uk/english/ins/epns/ <http://www.nottingham.ac.uk/english/ins/epns/> )
Contact details:
e: [log in to unmask] <blocked::mailto:[log in to unmask]>
t: +44 (0)117 328 3278 (internal 83278)
f: +44 (0)117 328 2295
h: Room 5E26
Dept of Languages, Linguistics and Area Studies
University of the West of England (Frenchay campus)
Bristol BS16 1QY, UK
w: http://www.uwe.ac.uk/hlss/llas/staff_coates_r.shtml <http://www.uwe.ac.uk/hlss/llas/staff_coates_r.shtml>
________________________________
From: The English Place-Name List [mailto:[log in to unmask]] On Behalf Of Keith Briggs
Sent: 15 June 2009 10:08
To: [log in to unmask]
Subject: [EPNL] carfucs
The treatments of carfax in EPNE (s.v. carrefors) and VEPN are unsatisfactory, especially after the points made by Sayers in N&Q 55,131 (2008). But Sayers does not separate the French and English evidence, and when this is done, I think the picture is even clearer than in his presentation. Here are the main points:
1. To start with what is certain, French carrefour `crossroads' comes from Latin quadrifurcus, with regular loss of final -k, and regular retention of final -r. It has not changed its form since the 13th century. As far as I know, it has always meant `fourways'.
2. There is no reason to assume that carfax is the same word.
3. There is another OF word, lost in ModF - carroge, carouge etc., and this is probably from quadrivium. For the -g-, cf. cage from cavea. It seems to have early lost its association with four, and could mean an intersection of any number of roads. This word gives numerous place-names: Carrouges, Carouge, Charroux etc.
4. At some point these words were confused, and a blend arose in which the -f- from carrefour got introduced into carouge. This is the essence of Sayers' idea, and it seem to me better than assuming that carrefours lost its last -r, because that does not explain the stop. The first case of this new word, and probably the only one from France, is au carrefoz de treis chemins in the early 12th century Gormont et Isembart. Here we also have an example of a sense `threeways'.
5. In England, this blended word proliferated. The earliest examples are these two glosses, not cited by Sayers:
carfukes compita 13th TLL i 67
carfucs triviis 13th TLL ii 48
We have another 'threeways'. The only difficulty is to understand why the -g- (/d3/) became devoiced. This is the only bit of the whole story which is unclear to me. MED has these examples in which a different number of ways than four is meant:
Vppon a buschement they ordeyned hem þere at a carfox..where as vij weyes weren vsyd (15th)
Thei enbusshed hem a-gein a carfowgh of vj weyes. (15th)
The MED headforms are:
carfouk (n.) Also carfuk(es, -fax, -foix, -fox, carfough
AND has:
carfour, carfouk, carfouke, carfu, carfuc, carfuke; qarfou, qarfouke; quarefur, querefur; quarfour, quarrefour, quarrefur, querefu (quartfur TLL ii 27; pl. quarruges; quaterfukes)
(I suspect that carfu, qarfou might be 'curfew'. Carfou is a recorded spelling for this in France.)
6. Oxford's Carfax was le Quarfuk 1297-8. I suggest that because this place was so famous, it became commonly thought that the meaning was `fourways', and it was re-associated with quadrifurcus under academic influence. The modern termination -fax is simply fucs-avoidance. For most of its history, the word had -u-, -ou- in its last syllable.
Keith
TLL = T. Hunt, Teaching and learning Latin in 13th-century England (3 vols, 1991-3)
PS: Could Quar(r)efurs (YW 2.65) c.1190, c.1200 just be `quarry furze'?
________________________________
This incoming email to UWE has been independently scanned for viruses by McAfee anti-virus software and none were detected
This email was independently scanned for viruses by McAfee anti-virus software and none were found
