Etymology of "trouver" (summary)

Tue Nov 19 21:59:14 UTC 2002

Dear colleagues,

About three weeks ago, I solicited your opinions concerning the 
competing etymologies proposed for French "trouver" and Occitan 
"trobar".  I had in mind the debate, which took place about a century 
ago, between Hugo Schuchardt and the French linguists Gaston Paris and 
Antoine Thomas. Paris had reconstructed the Vulgar Latin proto-form 
*tropare via regular sound laws, and then proposed a somewhat 
farfetched semantic pathway ("compose [a melody]" > "invent" > 
"discover, find" ) to make the etymology work. Schuchardt revived the 
derivation from /turbare/ proposed by Diez, which required the 
postulation of irregular sound changes under the influence of the 
closely-related verb /turbulare/ > */trublare/ "stir up". On the 
semantic side, turbare underwent a meaning shift from "stir up" to the 
more specialized sense of "stir up [water] in order to drive [fish 
toward a trap or net]".  At the time I sent the message, I had 
encountered but a single mention of a third proposal, according to 
which the ultimate source of Old Provençal trobar and its cognates, at 
least in their specialized use to denote the composing of verses, 
singing, etc. (and of course, the derived nouns trobador, troubadour) 
is an Arabic word borrowed into the Romance dialects of medieval Spain. 
The text I had read was an unfavorable review (in Romania 1969) of a 
1966 paper by Richard Lemay (Annales Economies, Sociétés, Civilisations 
XXI: 990-1011), in which the word "troubadour" was traced to the Arabic 
root /D.-R-B/ “strike, touch”, by extension “play a musical instrument” 
(the postposed dot = "emphatic", pharyngealized coarticulation). The 
review was rather dismissive in tone, and did not inspire me to look 
into the matter further. Since submitting my question, I have received 
about a dozen responses, which fall into two groups: (1) those who 
believe that some form of the *tropare root is the most likely source 
of "trouver" &c, and who have never heard of the proposal of an Arabic 
source; (2) those who HAVE heard of the Hispano-Arabic hypothesis. Most 
of these latter find it credible or even "the most likely source" (Paul 
Lloyd). No one supported Schuchardt's /turbare/ etymology. What has 
come as a total surprise to me is not only the existence of this third 
etymology -- actually, set of etymologies, there is more than one -- 
but the curious disconnection between the two groups of respondants. In 
a second e-mail, Paul Lloyd wrote that his professor of Arabic (at U. 
of Pennsylvania) thought the Hispano-Arabic source "was the accepted 
etymology and was surprised that any Romance scholars doubted it".

The person most responsible for promoting this third proposal, at least 
among Arabists and literary historians, appears to be María Rosa 
Menocal, professor of Spanish at Yale. In two papers (Romance Philology 
XXXVI #2: 137-148 [1982], and Papers from the XIIth Linguistic 
Symposium on Romance Languages, pp. 501-515 [1984]), as well as her 
1987 book "The Arabic role in medieval literary history: A forgotten 
heritage" [University of Pennsylvania Press], Prof. Menocal has 
attempted to demonstrate, first of all, that the correct Arabic source 
is not the root identified by Lemay, but rather the nearly homophonous 
/T.-R-B/ “provoke emotion, excitement, agitation; make music, entertain 
by singing”. Secondly, she accused the scholarly community of Romance 
linguists of bad faith for their refusal to grant the Hispano-Arabic 
hypothesis the same airing in professional journals and etymological 
dictionaries as was accorded Schuchardt's /turbare/ etymology.  In her 
opinion, it was not a question of the relative plausibility of either 
Arabic etymon compared to the Latin ones under consideration; the real 
problem is "the intellectual framework and set of scholarly assumptions 
and procedures which led to the complete ignoring of this possible 
Arabic etymon" (Menocal 1984: 504).

It is essential to note that neither Lemay nor Menocal offer their 
Arabic etymon as the source for the Romance verb meaning “find”. In 
their view, this lexeme was already present in the Romance dialects of 
Spain and the Provence, with something akin to its modern meaning, when 
the Arabic root was borrowed. Homophony led to overlapping usage and 
eventual fusion of the two verbs, one indigenous (trobar1), one 
borrowed (trobar2) (Lemay 1966: 1009). One can easily imagine why such 
an etymology, in either  version, would meet with the disfavor of 
“mainstream” specialists. The semantic fields associated with /T.-R-B/ 
and /D.-R-B/ most closely overlap that of *tropare, in that all three 
roots could be employed to denote some sort of musical composition or 
performance, whereas they have no resemblance whatsoever with the 
meanings reconstructed by either Diez or Schuchardt for turbare. 
Therefore, the postulation of an Arabic source would compel rejection 
of the Latin etymon with the most impeccable phonetic credentials, and 
a meaning no more problematic, in favor of a hypothetical borrowing 
that would require additional phonetic assumptions, relating to the 
manner of its adoption into Hispano-Romance, to account for its 
attested forms (Lemay 1966: 1004-1007; Menocal 1982: 146-147). The 
proposed antecedent *tropare is not attested in Latin, but neither is 
there any compelling evidence that any derivative of either Arabic root 
was borrowed into medieval Hispano-Romance. The proposal has another, 
equally unfortunate, consequence. Having been pushed aside as the 
source of trobar2, trobador, etc., *tropare would be left to compete 
with turbare as the etymon of trouver/trobar1 “find” alone. On this 
reduced playing field *tropare would be at a distinct disadvantage, 
indeed, partisans of the Hispano-Arabic hypothesis would be almost 
forced to acknowledge turbare as the sole likely source of the 
homophonous verb trobar1. In other words, it requires overturning the 
stronger etymology in favor of the weaker one, and abandoning a single 
source for both senses of “trobar” for the less elegant solution of a 
split etymology.
One might question the extent to which the study of the Arabic 
influence on Hispano-Romance has been tainted by “the overtly 
anti-Semitic tendencies in Spanish history” (Menocal 1984: 504-505), or 
whether Romance etymologists have shown bad faith in refusing to 
discuss, in print at least, the merits of /T.-R-B/ or /D.-R-B/ as an 
antecedent of Old Provençal trobar. I have just given some of the 
weaknesses of the Hispano-Arabic hypothesis that might have motivated 
its rejection. Whether those are grounds for carrying on as though the 
etymon had never been seriously proposed is another question, one that 
I am in no position to answer.
Sorry for the rather wordy answer. My thanks to all those who 
responded: Miguel Carrasquer (who was also kind enough to scan & send 
me the entry on "trobar" from Corominas' "Diccionari etimològic i 
complementari de la llengua catalana"), Marc Picard, Laurent Sagart, 
Britt Mize, Mark Southern, Paul M. Lloyd, Carol Justus, Roger Wright, 
and Maria Rosa Menocal.

Kevin Tuite

************************************************************************
Kevin Tuite                                   514-343-6514      (bureau)
Département d'anthropologie                   514-343-2494 (télécopieur)
Université de Montréal
C.P. 6128, succursale centre-ville
Montréal, Québec H3C 3J7                         tuitekj at anthro.umontreal.ca
NOUVEAU! Page Web en construction:             http://www.philologie.com
************************************************************************
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: text/enriched
Size: 8312 bytes
Desc: not available
URL: <http://listserv.linguistlist.org/pipermail/histling/attachments/20021119/31a75f8f/attachment.bin>