From the tenth century on we find bruscus and brusca (44) for buscus and busca, but in the northwest of Italy bruca and bruga became the most popular words for "faggots, fuelwood, twigs," while in France we have the fuller forms bruscia, broca, etc. Where formerly buscalia, buscaria was used, we find in the south brucaria, bruguera, bruaria, brueria for "brush grown country." This group, spreading to the north into regions where there was no forest, has been applied to heathergrown land. The Welsh brwg "forest, brush, ferns," which is borrowed from the English, shows the meaning "forest" and "fern" side by side, because the underlying meaning is deserted land, brush grown land, etc. The early loss of s in bruscus, brusca is explained by the influence of the Lat. brocchus, broncus, bronchus "provided with teeth," while Lat. branca "paw" produces branca, branchia "fuelwood, faggots" (45) and then "branch." Similarly bropa results from a crossing of brosca with Lat. scopa "thin twig," (46) and under the influence of Lat. frondem arises bronda. (47) But brusca "firewood" has produced bruscare, bruxare, brusare, and French brûler, from an older brusculare.

We now can treat the Romance representatives of the group in a summary manner. We have "OFr. boschaille bois, boscheer, bouchoyer, bocheyer couper du bois, boschel, bochal, etc., buisson, bosquet, bocage, petit bois, bouche botte fagot, bouchage tas de fagots, boissier boisier, boichier qui travaille le bois, buschier abattre du bois et en faire des bûches, couper des branches d'arbres, busche, boize buche." If we keep in mind that boscare was the act of finding and knocking down the dry branches with the wooden stick, we see at once why we get "OFr. buschier, busquer, bucquier, bucquer, busser frapper, heurter, boucheter émouter, battre, mal traiter, busquer chercher." Forms with inserted r are: "broce, broche, brousse, brouce, brouse, etc., broussailles, hallier, petit bouquet d'arbres, broceron, brocheron petite branche courte, broçonner bourgeonner, reverdir, broisson rejeton." As early as the tenth century brustum means "forest fodder, young leaves, food for goats," (48) and from this comes "OFr. brost, brout, jeune pousse des arbres au printemps, broster brouter, brotage droit qu'on payait pour faire brouter ses bêtes quelque part, brotel, broteau taillis, broterie, brouterie lieu on broutent les animaux." From brusculum are derived "OFr. bruliau fagot, broussaille, brûlis portion de foret incendiée, de champs dont les herbes ont été brûlées, brusler brûler." Burjon "bourgeon," like "Prov. burca, burga, burja fourgonner, tisonner, aiguilloner, pousser, taquiner, cogner, heurter, bourjouna, broujouna, brouchona, brounchouna fourgonner, patrouiller, bousculer, Centre broquer, bruquer heurter, choquer" owes its origin to a contamination with Lat. brocchus, bronchus.

A similar wealth of forms is to be found in the Provençal: "bos, bosc, boch, boi, bo, etc., bois, bousca rechercher, quêter, gagner les bois, déguerpir, bouscaia, bouscalha ramasser du bois, bouscaio, bouscalho bois en général, toute sorte de bois, bouscaren, bouissaren forestier, bouscas grand bois, futaie, épaisse forêt, forêt affreuse, mauvais bois, sauvage, bâtard, bousco recherche, bois en général, bousquet bosquet, petit bois, bousa boiser, couvrir de bois, bâtonner, rosser, boustiga remuer, fouiller, bouleverser, busc broutilles ou brussailles épineuses, busca busquer, bûcher, buscaia glaner des buchettes, buscaio broutille, brindille de bois mort, busqueja ramasser du bois, bussa cosser, heurter de la tête; bro, broc, brot, boc branche morte, scion, bûchette, épine, dard, buisson, bâton, broco bûchette, broutille sèche, brous broussaille, broussa balayer, broussaio broussailles, bruyère en général, brousso bruyère à balais, touffe de broyère, broust brout, pousse d'un taillis ramée, brousto pousse, ramée, branche, taillis, pousse de chou, chou brocoli, brout brout, jeune pousse des arbres et des plantes, brouta pousser, vegeter, brouto rejeton, nouvelle pousse, broutouna pousser, bourgeonner, brusc bruyère commune, tige de bruyère, brusqueirolo petit champ de bruyère, brusquié taillis de bruyères, brusquiero champ couvert de bruyère, bruyère, bruga taillis de bruyères; broundas rameau dont on se sert en guise de balai, brounditoun, broutihoun petite broutille, petit rameau, brounca, abrounca, brouncha, brunca, bruca, breca, ruca broncher, heurter, bruia, brulha, broulha pousser germer, reverdir, bruioun, brulhou, brellou, broulhou, bourlhou, orboulhou bourgeon, rejeton de chou, brula, brusla, brunla, burla, bourla, brouda brûler, incendier, bruscla, baruscla, bourouscla, brucla, bristoula, brounzi, brousi, brausi, abrausi, brusi, brui, broui haler, brûler, brusa, bruza, bruisa brûler."

Since the corresponding Italian words have evolved out of the French terms, we naturally find them at first in the north. Beginning with the twelfth century we find recorded broco "twig, faggot" (49) and crossing with frasca we get frusca, (50) so later that Ital. bruscolo, fruscolo fuscello are synonymous, while broco develops further into sbrocco, sprocco. The forms with and without r interchange in the dialects, hence Genoese brusca bûsca = buscare "to seek" and bruscare "to brush," while Friul. busca = busca, busco, brusco, bruscolo, and brusçhâ and buschâ = dibruscare, bruscare, buscare. If we now consider that Ital. busso, buscio, fruscio at the same time mean "noise," it becomes at once clear that not only all these belong to the busco group, but that also French bruit is to be included in it. The development of "noise" is universal in Europe, as has been seen under vasta. Spanish brotar "to form buds," brote "bud," bruza "brush," buscar "seek" have developed from the French, and a careful search will in the Romance languages reveal a much larger number of words that belong to this group.

I now turn to the second part of the expression arbustum vitatum. In Langobard Italy terra vitata is opposed to "terra campiva, aratoria, prativa, ortiva," to field and garden land, and also to "terra olivata, selvata," olive grove and forest. (51) Much older and much more popular is the use of vitis in the sense of terra vitata, "vites meas petia una in fundo bonate quem ego emmi" (745). (52) Although now and then the combination terra cum vitibus occurs, (53) from which may be assumed that it has the original meaning of vine-covered land, such expressions as "ipsa petiola est de uites et uacua terrola," (54) "petiola una de terrola cum uite et uacuum," (55) show that vites means "brush grown land."

This vitis, which occurs in Langobard documents since the beginning of the eighth century, occurs at least as early in Frankish documents. In a spurious Merovingian document of the sixth century, which is, no doubt, not earlier than of the eighth century, the pertinence runs as follows, "haec omnia cum mancipiis desuper manentibus, mansis, domibus, aedificiis, curtiferis, widis, campis, vineis, silvis, pratis, pascuis, aquis aquarumve decursibus," (56) and again, "tam mansis, colonicis, aedificiis, silvis, ingrediciis, widis, campis, pratis, pascuis, aquis aquarumve decursibus." (57) Pardessus reads incorrectly for it unidis, campis, (58) while Quantin has the significant reading vuidis, campis. (59) In another, most likely spurious, document, which is said to be of the year 499, but which exists in a copy of the thirteenth century, we find the same pertinence, "haec omnia cum mancipiis desuper manentibus, mansis domibus aedificiis cultiferis mudiscapis vineis silvis campis pratis pascuis aquis aquarumve descursibus," (60) where, of course, mudiscapis is a misread vuidis campis or widis campis. The position of this vuidis campis between curtiferis and vineis does not permit any doubt that we are dealing here with some kind of a plantation, and the identical juxtaposition of vitis and campum of the Langobard documents shows that what in Italy formed two distinct words and concepts here begins to be conceived as one idea, so that in the last quoted document, but not in the first two, campis is once more repeated before pratis. The constant use of vuidis campis in the pertinence has led to regarding this as a compound. Now, as vasta has led to forestis, and bustum to bruscus, etc., and uacatum to vacartum, (61) so widis campis has changed into widriscapis, (62) wadriscapis, (63) watriscapis, (64) watriscafis, (65) vatriscafis, (66) wardi scampis, (67) uuatriscapud, (68) wardriscapis, (69) warescapiis, (70) wariscapiis (71) quadriscapis, (72) quadris campis, (73) wastris campis, (74) votris campis. (75) Ducange also records wadiscabum and waskium. From this variety of forms has arisen OFr. warescais, warescait, warescape, wareschel "tierres vagues, lieux destinés à la pâturage publique," (76) and to these are to be added Wallon. warechais, wareschaix, Liège wercha, Mont. warescaix, waréchaix, wareskaix "pâturages communaux, vaines pâtures."

All these words belong to the north of France. Otherwise the older form widis, vuidis has spread over a very large territory. In the Chanson de Roland we have once voide place, (77) and once voide terre. (78) In the latter case we have a precise rendering of Ital. terra vidata, and from a large number of documents of the thirteenth century we know that wide, voide terre was the technical expression for "wasteland." This is also the case in the Chanson de Roland, where the word never occurs again, although one would expect it, if it already had the current meaning of "empty." Besides, the last quoted verses sound like an imitation of the documentary pertinence "cum viis et semiteriis et vidis terris et arboribus et planis." Godefroy quotes several passages in wich wide terre means "pasturage." (79) However, the wides terres are not exactly pastures, but fallow land overgrown with brush, used as pastures. They are opposed to pleines terres "cultivated lands," as may be seen from a discussion of Philippe Beaumanoir in the thirteenth century. (80)

For vuide terre we also find vaine et vuide. (81) This vaine is only a learned adaptation to vana, but has in reality arisen from gain. (82) Gain means "extraordinary, unexpected gain." In Godefroy there are a number of quotations in which he falsely translates gain by "fruit de la terre, recolte," where it should have been rendered "regain, rowen." In the lines:

Si a veu en une pleigne

Berbiz qui paissoient gain (Renart VIII. 175).
gain can only refer to "grass." For "autumn" Godefroy records waym, wain, vain, win, gayn, gain, vayng, gaain, and he adduces a large number of forms from the dialects. (83) As the original meaning of gain was forgotten and it meant any kind of gain, there was formed in French, first recorded in Littre in the sixteenth century, the pleonastic regain for "rown," for which we have the dialectic, Picard. reguin, rouain, Wallon rigain. To these are to be added Morvan "regâmer repousser, pousser de nouveau, Normand revouin regain, revouiner possuer comme le regain, revoiner reverdir, Comtois (Fourges) ruwain," (84) Berry "reguiner pousser en regain." (85) From the Norman revouin comes Engl. rowen. later popularized to roughings.

The distribution of this word in the south is particularly interesting, because it shows to what wild transformations, utterly baffling philology and phonetics, a word may be subjected. Levy records gaim in the Provence. In Berry we already find regouiver for reguiner. In the south the case is much worse. In Lyons we get the forms reviouro, revioulo, revuro, revivro, revioro, reviula "regain." (86) After this one will not be surprised at the extravagances recorded in Mistral. Here we get revieure, rouibre, rouire, rouibre, rebouibre, reboulbre, roubibre, roudibre, gouibre, reboulibre, rebouribre, reboulume, relubre, rebouchouire, reboujouire, reboussouire, rourieu, bourieu, abourieu, aurieu, bouirieu, vourieu, vouri, voueiri. (87) One sees plainly how, beginning with gain in the north of France, the forms get worse and worse as they proceed south. If we turn to the Atlas linguistique de la France (N° 1139), one can get a clear idea how it has happened. Gain occurs only sporadically in the northwest, in Côtes-du-Nord, Orne, D.-Sèvres, while in Belgium forms arising from wain are universal. If a straight line is drawn from Bordeaux to Châlonssur-Saône, we get to the north of it the regain group. From Belgium, where the rewain froms are exceptional, the rewain words go straight south, here and there alternating with wain. Towards Switzerland begin the corruptions which produce the Provençal forms. In Doubs regain, rewain, wain meet. In the southwest of Switzerland we get rekwai, while in Jura and Doubs we have rewain. Towards the east rise the forms rekwa, rekwar, rekor, reko, and these go south to Savoie and Ain. In Isère rekor stands side by side with revur, and the latter explains the other corruptions of the south. In the south, along the coast, and along the Pyrenees, in Landes and Gironde, that is, at the periphery, derivations from regain are exceptional. Here we have the freely formed reprin or retaille, and Covarubias was certainly right when he derived Span. retoño from the synonymous retallo. If this map may be safely trusted as representing a historic tradition, and there is nothing to contradict such an assumption, the whole group must have had its birth in the north, spreading in two streams, one directly to the south, the other along the western border, equally to the south and towards Switzerland, taking from there a westerly direction. We are, therefore, compelled to assume that an intensive use of the rowen proceeded from the north. French gaim, waim has entered Italy in the form guaime, and guimeau, waymal has produced there gomireccio, grumereccio, again proving the assumption that the group proceed from the north of France. The same confirmation is found in Germany. MHG. amat, üemet, Swiss amad, German Emde, Oehmd, Ohmet, Ohmt "rowen" are transformations of the French waim, with a leaning towards the MHG. mat "mowing," and as Fr. guimeau has led to Ital. grumereccio, so it has also produced German Grummet, with a possible popular derivation from "grün" and "Mahd."




Endnotes