Philip Newton wrote:
> On 9/5/06, Henrik Theiling <[log in to unmask]> wrote:
>> Hi!
>> Carsten Becker writes:
>> > >>    2. werewolf / lycanthrope of some variety
>> >
>> > ayvengaryo (lit. "wolf-man")
>> Interesting.  Did you have a particular reason to decide to reverse
>> the typical order for compounding?  

Typical? Doesn't it rather depend upon whether a language forms 
head+attribute or attribute+head compounds?

>>All the languages in which I know
>> the word 'werewolf' compound it as 'man-wolf'.
> "Lycanthrope" is a counter-example :) (lykos, wolf; anthropos, human)

Yes, indeed. Greek _lykanthropos_ <-- lykos = wolf + anthropos = human 

French has _loup-garou_ (plural: loups-garous) where the 'wolf' element 
(loup) is clearly first. but _garou_ is, of course, not the French for 
'man'; it is from Old French _garoul_ which is from Frankish (a Germanic 
lang) *werwulf.

While Spanish has _hombre lobo_ (man wolf), it sister langs of Galician 
& Portuguese put the 'wolf' first, thus: Galician: lobisón; Portuguese: 

Italian _lupo mannaro_ also puts 'wolf' first (The etymology of 
_mannaro_ is uncertain. Some derive it from a Germanic mann- (man); 
others derive the phrase from an earlier *lup'umanario "humanish wolf").

Before someone asks, Latin merely has _uersipellis_ ("skin-changer")   :)

[log in to unmask]
Nid rhy hen neb i ddysgu.
There's none too old to learn.