Advanced search
Publication Cover
WORD Volume 6, 1950 - Issue 1
Submit an article Journal homepage
681
Views
2
CrossRef citations to date
0
Altmetric
Original Articles

The Hypothetical Base in Romance Etymology

Yakov MalkielUniversity of California.
Pages 42-69 | Published online: 04 Dec 2015

  • Thomas' article appeared as a preface to his Nouveaux Essais de philologie française 1–34 (Paris, 1904); it was originally published in the Revus des deux mondes (Dec. 1902) under the title La Science étymologique et la langue française. Schuchardt's statement, clad in the form of an appraisal of G. Paris' work, is found in ZRPh. 28.50–5 (1904); cf. his earlier article Etymologische Probleme und Prinzipien, ZRPh. 26.385–427 (1902). The merits of the arguments of both opponents were weighed by M. Roques, Méthodes étymologiques, JSav. 419–33 (Aug. 1905) and especially by E. Tappolet in his inaugural lecture Phonetik und Semantik in der etymologischen Forschung, ASNSL 115.101–23 (1905), in which Thomas was characterized as a sober realist thinking in terms of present-day possibilities and Schuchardt as an inspired leader envisioning future potentialities. Wagner's article wae published in CN 3.5–26 (1943). Additional studies worthy of consultation include G. Paris, Un nouveau dictionnaire de la langue française, Revue des deux mondes 71.5.241–269,802–28 (Sept.—Oct. 1901), dealing with Littré's, Darmesteter's, and Thomas' methods of etymologizing, with apposite illustration of sense development from Latin to French (brume, tourner, timbre), W. Meyer-Lübke, Aufgaben der Wortforschung, GRM 1.634–47 (1909); J. Jud, Neue Wege und Ziele der romanischen Wortforschung, Wissen und Leben 9.270–9,320–8; E. Tappolet, Neuere Aufgaben der Wortforschung, GRM 13.130–41 (1925); L. Spitzer, Aus der Werkstatt des Etymologen, JbPh. 1.129–59 (1925); W. von Wartburg, Sprachgeschichte und Kulturgeschichte, Schweiz. Monatshefte 3.552–64 (1924) and the lecture Grundfragen der etymologischen Forschung, Neue Jahrbücher für Wissenschaft und Jugendbildung 7.222–35 (1931), with further elaboration in some of the best chapters of his controversial Einführung in die Methodik und Problematik der Sprachwissenschaft (Halle, 1943); E. Ga- millscheg, Zur Methodik der etymologischen Forschung, ZFSL 50.216–98 (1927), a rejoinder to L. Spitzer; H. Kuen, Pflichten des Etymologen, WS 20.184–9 (1939), a definition of minimum requirements in the light of rapidly collapsing standards of research in pre-war Germany; E. Seidel, Aufgaben und Methoden der etymologischen Forschung, BL 9.5–28 (1941), recommending sweeping reforms, see Wagner's reaction in CN 3.21–3. The latest presentation of Wagner's views on etymological desiderata is found in his article Sobre os nomes da moega nas linguas ibero-românicas, Biblos 24.1–19 (1949).
  • There exists a sizable corpus of researches on etymology based on a miscellany of sources transcending the Romance domain. Outstanding among the older sources is Thurneysen's monograph Die Etymologie: eine akademische Rede (Freiburg i. B., 1905). W. Schulze's note Etymologische Zweideutigkeit, ZVS 56.141 (1928–9), hardly bears out the promise of the title. Recently, Italian scholars have taken the initiative in formulating principles of linguistic reconstruction; see V. Bertoldi, Questioni di metodo nella linguistica storica (Naples, 1938–9; the revised edition is not accessible to me) and La parola quale testimone della storia (Naples, 1945; see my forthcoming review in Italica), V. Pisani, L'etimologia: storia—questioni—metodo (Milan, 1947), known to me only indirectly; G. Bonfante, On Reconstruction and Linguistic Method, Word 1.83–94,132–61 (1945); 2.155–6 (1946), and The Neolinguistic Position, Lang. 23.344–75 (1947).
  • Aside from the aforementioned methodological essays, carefully screened and classified material on etymological theory is scarce. Manuals of historical grammar all deal with phonology, based squarely on the identification of word-origins, without sufficient concern about the underlying processes of identification. Etymological dictionaries, in varying degree of detail, present established and controversial results of monographic inquiries; authors, as a rule, imply preferences of method by qualifying as adequate a limited number of explanations. Formulations of principles introduced into individual word-studies are frequently designed to fit a single situation, a given state of affaire; how much of such theorizing remains valid in the case of the elimination or withdrawal of the proposed solution of a concrete problem? And, granted that the putative ancestry of a single word should be found acceptable, does the implicit generalization necessarily carry commensurate weight? Authors inclined to be rash in capitalizing on a few felicitous findings have repeatedly been guilty of mutually exclusive claims. A tactical or strategic scheme in etymology, to be useful, ought to be based on a generous sampling of satisfactory derivations, collected at random, without any aprioristic view. Seasoned language students, in disentangling what French scholarship aptly calls “l'historique du problème,” act not unlike staff officers who, upon the conclusion of operations, reëxamine and evaluate the sequences of moves. This kind of retrospects may be out of place in texts and too cumbersome for inclusion in reference books. Studies in methodology, however, would be singularly deficient by omitting characteristic samples of antiquated approaches. The present-day explorer needs information to work out for himself the pattern of progress, which will help him to discern, within the body of currently accepted hypotheses, those in urgent need of critical reconsideration.
  • C. Michaëlis de Vasconcelos supported this derivation in Miscellanea Caix-Canello 136–7 (Florence, 1886); M. L. Wagner assumes a blend between Gal.-Ast. amorodo and Sp. lodaño < *lot-ōniu. In his Etimologías españolas, Rom. 29.334–79 (1900), Menéndez Pidal frequently upholds derivations initially formulated by Nebrija, Rosal, and Covarrubias; cf. Sp. basura ‘rubbish’ < *versūra (339), Sp. cicién ‘ague’ < accessiōne and Sp. en ridar ‘to stick, to incite’ < irrītāre (345–6); OSp. escorroço ‘indignation,’ akin to Fr. courroux ‘wrath’ (348); OSp. recel ‘bedspread of thin, striped cloth’ < *raçel < raça (363).
  • I do not discuss here the practices of medieval etymologists (traceable through Isidore of Seville to Antiquity) who concurred in identifying names and things. Ironically enough, etymology, when intrinsically at its lowest, was sufficiently respected to have exercised the strongest influence on neighboring fields of learning, particularly on scholastic thought and clerical poetry; see E. R. Curtius, Mittelalterliche Literaturtheorien, ZRPh. 62.473–9 (1942), and L. Olschki, The Myth of Felt 3–4 (Berkeley and Los Angeles, 1949).
  • Thomas, Mélanges de philologie française' 49 (Paris, 1927); Meyer-Lübke, REW3 no. 1414.
  • Thomas, Mélanges2 22–3, maintains this hypothesis in a slightly modified form; the trisyllabic structure of OFr. ançïen leads him to assume retarded development of *antiānus.
  • Old Spanish and Old Provençal congeners played a major rôle in Thomas' rehabilitation of Du Cange's etymon, see Nouveaux Essais 25; Mélanges2 5–6. Diez started from *acceptāre which has survived as the unchallenged base of Sp. acatar ‘to respect.’
  • In Rom. 49.389–95 (1923) Jud proposed Celt. *andebanno; the traditional view was upheld by Thomas, Mélanges2 34; Meyer-Lübke was silent in REW1 and REW.3
  • The discovery of the archaic French variant malabde led Diez (and, in his wake, Littré, Scheler, Brächet) to postulate the etymon *maleaptus. J. Cornu, in 1874, chanced upon the correct base malehabitus.
  • Cf. Plato's Cralylos and Cicero's De natura deorum.
  • According to G. Bertoni, Le origini delle letterature romanze nel pensiero dei Romantici tedeschi, AR 23.1–10 (1939), the thinking of early-nineteenth-century philologists was obfuscated by the arbitrary assumption of sharp cleavage between two cultural spheres in late Antiquity and the early Middle Ages: one, Latin (learned); the other, Romance (vernacular). Their interaction was overlooked and the agency of medieval Latin, as a language and a vehicle of literature, was ignored. The idea that Latin and Romance form a single, highly complex organism and that the terms ‘mother language’ and ‘daughter languages’, if taken literally, involve misleading images in their application to the Western Mediterranean, dawned upon scholars at a much later date, see G. Paris, Revue des deux mondes 71.5.264–5 (1901). Diez's influence on Scheler, Grandgagnage, Brachet, Littré (who also drew on A. Comte's positivism), Darmesteter, Michaëlis, Monlau, Reinhardstoettner, Caix, and others is known to have been very powerful.
  • See the ironical reference in Thomas Mélanges2 5–6 to Meyer-Lübke's classification of products, reputed or real, of *acceptāre, *accapit¯re, and *discapitāre. Schuchardt, the tireless opponent of Meyer-Lübke and Thomas, was equally prone to operate with dispensable reconstructions (for criticism, see the latter's Mélanges2 38–9, s. v. baillard). Meyer-Lübke's predilection for the asterisk is truly surprising; in REW3 (that is, as late as 1930–5), he unnecessarily starred cercius, sectōrius, and sessitāre (the last attested in Cicero); cf. Thomas, Mélanges2 57, à propos of Lat. cuculilō, and V. Garcia de Diego, RFE 13.389 (1926). Notice also that Meyer-Lübke, in doubt how to identify the base of Sp. tosco ‘coarse, uncouth,’ preferred the abstract scheme *tuscus, devoid of identifiable meaning (REW3 no. 9013) to Hetzer's excellent vicus Tuscus ‘name of an ill-famed suburb of Rome’; see J. Corominas, AILC 2.151–4 (1942–4). On Meyer-Liibke's increasingly refined interpretation of the term Vulgar Latin, see H. Meier, Über das Verhältnis der romanischen Sprachen zum Lateini-schen, RF 54.165–201 (1940), reviewed by B. Hasselrot, ZRPh. 62.195–6 (1942). On Meillet's changing attitude, toward reconstructed forms, see G. Bonfante, Word 1.87–8 (1945).
  • Menéndez Pidal's Manual de gramática histórica española7 (Madrid, 1944), whose original edition dates from the beginning of this century, also contains inaccuracies in regard to the starring of bases; in the case of pedregoso ‘stony’ (75), the author's belief that *petricōsus was an authentic formation led him to develop a hazardous hypothesis, see Lang. 25. 139–45 (1949).
  • E. Gamillscheg, Etymologisches Wörterbuch der französischen Sprache (Heidelberg, 1928), based on a series of earlier notes. For the demolishing analysis of a typical conjectural reconstruction by this author, see J. Orr, Linguistic Geography as a Corrective to Etymology, TPhS 81–91 (1948).
  • Thomas, Mélanges2 27–8, on *armōne, artemōne ‘top-soil’; Ibid. 19–20 on Lang, amelanche ‘fruit du néflier’.
  • Idem, Nouveaux Essais 6 (Greek ancestry of the hapax OFr. trupelin was assumed by the pre-Renaissance humanist Oresme).
  • Ibid. 9.
  • Such partial connections of apparent cognates (tentative in default of a demonstrable base) were practiced by Thomas in his Mélanges3 43 (Fr. bellicant, Occ. belugan ‘species of fish’), 54 (OProv. cadatz, Sp. cadarzo, It. catarzo ‘coarse, englanted silk’).
  • Thus, Menéndez Pidal, in tracing Sp. chichón ‘bump, bruise’ to abscessiōne (Rom. 29.345), assumes that the latter shared the secondary connotation of *‘tumor’ with abscessus. Thomas, Mélanges2, s. v. chaintre, argued that cancer, -eris ‘crab’ colloquially may have developed an acceptation attested solely in its diminutive cancellus (‘grate’). Gamillscheg, op. cit., in discussing OFr. robot ‘plane,’ unduly credits it with the preliterary connotation of *‘rabbit’ (see Orr's critique).
  • In addition to conjectural forms and meanings, there exist conjectural graphs, as when Menéndez Pidal, in interpreting engrare (Fuero de Sepùlveda), assumes an original (subsequently misunderstood) spelling *engre, from which he infers the existence of *engar(e) from yengo ‘freed’ (Rom. 29.378–9).
  • See Thomas, Mélanges2 30. There are yet other typographic devices to circumvent specific starring. Menéndez Pidal, Rom. 29.339, in tracing back Sp. asomar ‘to appear at the top’ to as-summāre, implies that sumāre alone is recorded. An example of implication through context is: “Für sard, cariaza wird zu Unrecht ein lat. carasea konstruiert,” see W. von Wartburg, ZRPh. 61.133 (1942). If all the bases are on an equally conjectural plane (e. g., where they pertain to a hypothetical parent language), it is, of course, possible to omit the asterisk (which derives its chief usefulness from the contrast it affords between real and inferred bases) and to resort to another conventional designation, such as the use of bold-face or small capitals, see Lang. 25. 61 (1949). Pioneer etymologists used qualifiers like quasi, gleichsam to set off reconstructions; thus, Covarrubias, anent OSp. recel: “Díxose quasi racel, porque está razada”; Diez, Etymologisches Wörterbuch der romanischen Sprachen 349 (Bonn, 1853): “Tronçon konnte freilich aus truncus erwachsen (lateinisch gleichsam trunciō, trunciōnis).”
  • Thomas, Mélanges2 266: “Les mots marqués d'une croix sont ceux qui, résultant de lectures erronées ou de coquilles typographiques, n'ont pas de valeur propre.” E. Hermann was among those who advocated the use of a wealth of symbols, each correlated to a different category of hypothetical formations.
  • J. Leite de Vasconcelos prepared a detailed study of the hosts of ghost-words in Fr. Joaquim de Santa Rosa de Viterbo's late-eighteenth-century compilation Eliccidário de palavras antigas, see RL 26.111–46 (1925-7). Hippeau, in an early edition of an Old. French text, misread cure as aite, which thence slipped into F. Godefroy's dictionary and penetrated nto the researches of such a keen mind as P. Meyer, until unmasked by Thomas; OFr. alun de boucaut ‘sort of alum’ contains a descendant of Uolcānus which should correctly read bouguanz; Fr. chambucle ‘black rust’ has suffered severely at the hands of lexicographers and printers; chambuche, chambuel(l)e, and chambruelle are some of its better-known disguises, see Thomas, Mélanges2 31, 48, 63. Fr. coriette, long attributed to Marguerite de Navarre, needs emendation to cornette ‘iron piece protecting the corner of a wall,’ see Th. Heinermann, ASNSL 180.49 (1942). False separation of words may lead to distressing mutilation: OFr. muse empastez was a subject of speculation until brought under control as muse en pastez ‘undecided’; OFr. chaux ‘piece of footwear’ baffled scholars before it was combined with the preceding word into souzchaux, which showed a familiar derivational pattern, see F. Lecoy, Notes de lexicographie, Rom. 70.341, 350 (1948–9). The tracing of Spanish ghost-words was initiated by R. J. Cuervo; see Rom. 29.574–8 (1900), on acudia. In a forthcoming study of the suffix-(i)ego in Ibero-Romance additional examples will be offered: indiego and indigüelo, seemingly suggestive of the Indies, are, in fact, disguises for judiego < Iūdaicu ‘Jewish’ and judigüelo, interpretable as a diminutive either of judiego or of judío < Iūdaeu; frailego from fraile ‘friar,’ abnormal on derivational grounds (in native Spanish words, the -ego variant occurs only contiguous to a palatal consonant), may owe its listing in authoritative dictionaries to an early misreading of frailēgo, i.e. frailengo, involving the formant known from abolengo ‘lineage,’ abadengo ‘abbatial,’ realengo ‘kingly.’ One measure of the thoroughness of lexical works such as A. Darmesteter's, A. Hatzfeld's, and A. Thomas' Dictionnaire général (1892–900) is the consistent abandonment of phantom-words; see G. Paris, Revue des deux mondes 71.5.823–4 (1901). On judiego, judigüelo, see also my contribution “Graeco-Latin iŪdaeus and iūdaicus in the Romance Languages” to the Oriental Studies Dedicated to W. Popper (Berkeley, 1950).
  • See Thomas, Melanges2 68,24,35–6. The historically satisfactory spellings for the words (ordinarily listed as schwène, antoit, avalies) would have been chevène (< *capitine, *capicine), entois (< enteser), and avalis (on the pattern of abattis, pelis, semis). On the inclusion of regionalisme by Littré (collected by himself in Normandy), see G. Paris, op. cit. 260–1.
  • One such case is Ptg. iguaria ‘tidbit,’ OJud.-Sp. yegüería ‘mess’ < iecuāria ‘giblets’ (cognate to iecur ‘liver’); see Lang. 20.108–30 (1944), 21.264–5 (1945). On cara, see RPh. 3.53 (1949–50). Coxus is recorded in CGL 3.46837. The lone variant cauicula (in lieu of clauicula ‘peg’), CGL 2.56335, has fathered untold Romance descendants. Very scantily documented is further tāliāre ‘to split’, which experienced a luxuriant growth in Romance.
  • Yet that contradictio in adiecto is implied by Thomas, Mélanges2 23, anent *antiānu > OFr. ançïen, likened to grātiōsu > Fr. gracieux and pretiōsu > Fr. précieux (divergent views of Darmesteter, Gröber, and Meyer-Lübke are quoted). A similarly awkward wording (of which the distinguished scholar surely would not approve at present) is found in Menéndez Pidal's Etimologías españolas, Rom. 29.343 (1900), à propos of Sp. corondel ‘column of writing’ < *columitellu < *columila < columen, columna: “Corondel es voz semiculta de amanuenses e impresores (cf. bajel, novel, doncel).” The termination -el points to Catalan- Provençal provenience; Catalonia contributed to Castile much of its terminology of craftsmanship.
  • Thomas, Mélanges2 48–9; the author's own preference was for *burriōne, from burra. Menéndez Pidal, with particular insistence in his didactically worded Manual, has often had recourse to such theoretically normal bases, in search for those factors that deflected the words from their straight course (titulu ‘title’ not > > *tejo; nātīuu ‘native’ not > *nadío; saeculu ‘century’ not > *sejo; cupiditās, *-ǐtia ‘greed’ not > *codeza' Manual7 12–4) and in criticizing etyma established by predecessors (Diez's* grūǐcula would have yielded *grueja rather than grulla ‘crane,’ Rom. 29.354). The conceivable effects of syncope obviated by counteracting forces are eloquently driven home by such formulaic suggestions of inequality as dolōrōsu not > *dorloso, hospitātu not > *hosdado (Manual7 75). In word-formation, starred forms exemplify, in W. von Wartburg's own words, “Bildungsmöglichkeiten, die unbenutzt bleiben: It. *librino beside ragazzino,” see ZRPh. 62.130 (1942).
  • See Thomas, Mélanges2 41; on the function of the suffix -ot in Walloon, see W. von Wartburg, ZRPh. 62.158–9 (1942). Similarly, OPic. ais auvereche is a derivative in -ārǐciu from auve, aube ‘paddle’ which, in Francien, would have yielded *auverez, -ece, see Thomas, Mélanges2 33–4; but did the type actually extend to the Seine basin? The reverse case is the reconstruction of a dialect form from an attested counterpart in the standard; thus, Menéndez Pidal, Manual7 6, infers the existence of the by-form *nūdus beside nodus ‘knot’ from the verifiable coëxistence of Octōber and (scantily recorded) Octūber ‘October.’
  • Locura ‘(fit of) madness’ and tristura ‘(fit of) ruefulness’ seem to have branched off travesura ‘prank, frolic, caper,’ based on trānsuersus ‘cross’. This type of starring is frequent in discussing non-existent and inconceivable sequences of words, e.g. G. Gougenheim, Les pronoms interrogatifs que et quoi, FM 17.87 (1949): “On ne peut donc dire ni *que de neuf savez-vous? ni *quoi de neuf savez-vous?” It also applies to rejected reconstructions, as when W. von Wartburg traces Fr. groseille ‘currant’ to Germ. Krausbeere rather than to Gall.- Rom. *acricella, see ZRPh. 62.219–21 (1942), and to derivatives impossible of coinage, as when parasynthetic Sp. desalmado ‘soulless, impious’ is specifically said to require no assumption of either *desalma or *almado, see Menéndez Pidal, Manual7 237.
  • Thomas, Melanges2 56–7.
  • Similarly, if Occ. art means ‘fishing equipment’ (cf. Sp. arte ‘net’ among fishermen) and Med. Lat. artes ‘nets’ is recorded precisely at Nimes (a. d. 1352), why not assume with Thomas (Mélanges2 29) OProv. *art which, ironically enough, may have slipped, as a technical term, through the meshes of otherwise well-informed lexicographers (notice the comparable sense development of ingenium in several vernaculars)? Leon, bisgo, Cat. bizco ‘squint-eyed’ were aptly connected with *uersǐcu ‘turned, shifted’ by means of the assumed links *viesgo and *viesco, cf. Ptg.-Gal. vesgo and topon. Puenteviesgo, quoted by Menéndez Pidal, Notas para el léxico románico, RFE 7.31–3 (1920). Sp. majuelo, not as a term of viticulture, but standing for ‘thornbush producing a red berry,’ reflects *mojuelo < mulleolu ‘reddish’ (Tertullian) rather than malleolu, see Menéndez Pidal, Rom. 29.348 (1900). Between ambūrō, -ěre ‘to burn’ and Sal. Ast. aburar, Ast. amburar one is tempted to insert, without excessive risk, *a(m) būrāre (Menéndez Pidal, RFE 7.6–7). Equally tenable are uespertīliō ‘bat’ > *vesperteyo > Ast. esperteyo, possibly through the influence of (d)espertar ‘to awake at night’ (Manual7 8, where the concomitant factor is disregarded) and apothēca ‘drugstore’ > *abodega (on the evidence of bodega ‘wine vault, cellar’) > OSp. abdega, Ptg. adega, see Rom. 29.334.
  • Thomas, Mélanges2 22–3.
  • A. Alonso, RFE 13.9–10 (1926); Y. Malkiel, RFH 8.136–41 (1946). Equally untenable is G. Baist's series rubru ‘red’ > * rovro > *rouro > Ptg. louro, Sp. loro ‘dark,’ ZRPh. 7.120 (1883), cf. Menéndez Pidal's comment: “implica cambios un poco anómalos en cuanto a la forma y muy violentos en cuanto al sentido” (Rom. 29.357–8). Yet Menéndez Pidal's intermediary forros are not invariably reliable, either, as he himself, with characteristic integrity, concedes. The gaps between hērēditāte ‘heirdom’ and OSp. herdad and between trīticu, Hisp.-Lat. tridigo (11th century) ‘wheat’ and OSp. trigo can be bridged in several ways (Manual' 154, 183). Subject to caution are the proposed series cīmice ‘bedbug’ > *cimce (rather ḱimḱe) > *chince > chinche (Rom. 29.345); apĭcula ‘little bee’ > *abeg'la > *abeyla > abeja (Manual7 47); septe(m) ‘seven’ > *siette > siete (Ibid. 142); capitāle > *cabidal (why not cabedal?) > cabdal caudal (Ibid. 161). Distinctly inaccurate, in my considered view, are the series titulu > tidulo *tidlo, *tildo > tilde (Manual7 12; the ending -e gives away tilde as an intruder from outside, a congener to, but not a descendant in direct line of, tidulo) and *sēricāriu ‘bird with a silken plumage’ > *sīricāriu > Sp. silguero, jilguero ‘linnet’ (Ibid. 197; the form serguero is attested; in other words, the shifts -erg- > -irg- > -ilg- occurred during the Spanish phase of the development, so that *sīricāriu implies an unwarranted projection of the process into Antiquity).
  • Thomas, Mélanges2 22–3.
  • V. Bertoldi, Questioni di metodo nella linguistica storica 84, posits *ebriaculus as a transitional stage betweeen μ∊μαιĸυλoν, μιμαιĸυλoν ‘species of plant’ (Arbutus Unedo L.) and Sic. mbriaculu. Thomas, Mélanges2 54–5, inserts Lat. catabola (cf. parabola > παραβoλή) between ĸαταβoλή ‘laying the foundation of a building’ and Lyonn. cadola ‘small hut’ (presumably borrowed from Occitanian). L. Spitzer, in opposing C. C. Rice's derivation of Sp. cansar ‘to tire’ from ĸἀμπτ∊ιν (Lang. 13.18, 19.154–6), disapproves of such assumptions on principle, see Lang. 14.205–6 (1938).
  • Thus, *acūtiāre ‘to sharpen’ has been inferred from the recorded name of the agent acūtiātor and the consensus of Romance offsprings (Fr. aiguiser, Sp. aguzar, It. aguzzare), see Menéndez Pidal, Manual7 4. Regional semi-learned *benevis > beneficiu has been extracted from OFr. abeneviser, postverbal abenevis, Med. Lat. benevisum, beneviser (Le Forez), see Thomas, Mélanges2 44. ONorm. ameschier ‘apple-tree,’ emmeschier ‘cherry-tree’ presuppose *amesche, *emmesche, which some scholars trace to Damascus and others to domesticus, see my Studies in the Hispanic Infix -eg-, Lang. 25. 178–80 (1949).
  • Gers, bezougneto ‘small bill-hook’ has been analyzed as a merger of bezougno (the local counterpart of Fr. besogne) and *bezoulheto, related to Occ. bezeuch, bezoui, Béarn. bedoulh, Arag. bodollo, Gers, bousoulh, bousoun < Celt, vidubiu ‘wood knife,’ see Thomas, Mélanges2 45–6.
  • See her Beiträge zur Geschichte der Romanismen (Halle, 1934). A similar line, on a less original level of scholarship, was followed by Mildred K. Pope in her simultaneously published text From Latin to Modern French (Manchester, 1934).
  • Thus, Thomas, Melanges2 27, lists *arm as the (expected) product of armu(s) ‘shoulder (of an animal)’; yet the only form on record is the plural ars ‘flanks (of a horse).’
  • This would apply to Southern and Eastern (including Northeastern) French, Northern Italian, Raeto-Romance, and Astur-Leonese-Galician. The term ‘type’ is used somewhat differently by W. von Wartburg, ZRPh. 62.131 (1942): “… höchstens vielleicht die Warnung, dass… die Form *theodisconem nur als Typus, nicht als so gesprochen aufgefasst werden darf.” In that context, ‘type’ is tantamount to ‘derivational scheme,’ ‘structural pattern’ and compares with ‘formation’ much as ‘phoneme’ does with its concrete realization ‘sound.’ This use of ‘type’ is widespread.
  • A case in point is *mordäcia ‘tongs, clog, bridle bit, clamp’ (REW3 no. 5678), which from Southern France extends all over Catalonia, Castile (mordaza, mordacilla), and Portugal; it is modeled on tenāx, -ads ‘tongs’ (REW3 no. 8638).
  • Thomas was known for the value which he attached to provisional types established exclusively on the basis of phonological speculation; cf. his statement on *anatolius ‘blind worm’, Rom. 40.107 (1911), endorsed by Wartburg, FEW 1.92–3, and on *nasiāre; “je crois que l'ensemble des formes romanes postule *nasiāre, mais mon credo s'arrête là”) as the base of Bourg. Morv. Berr. (n)aiger, (n)aizer ‘to steep (hemp),’ see Mélanges2 9–10. The same orientation is discernible in Thomas' attempt to link Prov. anar ‘to go’ to the hapax annāre ‘to last one year’ (Mélanges2 21–2) and in the vindication of the meaningless scheme *tropāre ‘to find,’ see Tappolet, ASNSL 115.113–7 (1905), and L. Spitzer, Rom. 66.1–11 (1940–1).
  • On this point (forcefully stated by Thomas, Nouveaux Essais 13) there has been considerable disagreement; apart from Renaissance scholars (G. Budé in France, J. de Valdés in Spain; their forerunner was N. Oresme) who made efforts to reduce French to a Greek nucleus, specialists in our own ranks have allowed their vision to be blurred by various manias and phobias, especially in regard to Iberian, Celtic, Teutonic, and Arabic elements of the Romance lexicon.
  • This is especially true of functional words like etiam ‘also,’ autem ‘however,’ tandem ‘finally.’
  • See Thomas, Mélanges2 55, s. v. cagouille ‘volute on top of the ram of a man-of-war’ and ‘snail.’
  • Ibid. 34–5, à propos of Lyonn. avair ‘swarm of bees’ (< *apāriu, from apis ‘bee,’ or = aveir ‘cattle’ < habēre ‘to own’?).
  • Menéndez Pidal, Manual7 306; V. García de Diego, Manual de dialectología española 170 (Madrid, 1946).
  • See Thomas, Rom. 33.213 (1904); REW3 no. 541; FEW 1.107.
  • Fr. dial, aranchier ‘to overturn, to lean one's back against’ and Sp. derrengar ‘to break the back’ jointly militate in favor of V. L. *arrē;nicāre, *dērēnicāre, from rēnēs ‘kidneys,’ see Thomas, Mélanges2 24–5; REW2 no. 7206; Lang. 25.147 (1949). OFr. chenevuis (mod. chènevis), Norm, canebuche, It. canapuccia ‘hemp-seed’ postulate V. L. *canapūtia, -ium (Thomas, Mélanges2 67–8). CAst. baltar ‘to knock down,’ Parm. baltar ‘to winnow, to sift’ point to V. L. *uallitāre, see Parodi, Rom. 27.204 (1898); Menéndez Pidal, RFE 7.36 (1920); REW3 no. 909. OFr. frailler, Camp. fraŷare ‘to suffer bruising,’ Ast. frayar ‘to squeeze, to break’ testify to V. L. *fragulāre (from fragor ‘breaking, crashing’), see Menéndez Pidal, RFE 7.12–3 (1920). RRom. kurám, Sp. corambre ‘hides, skins’ vouch for the early coinage of *coriā;men, from corium ‘leather,’ see Menéndez Pidal, Rom. 29.344 (1900). Reflexes from Rumania down to Portugal militate in favor of *dēpänäre ‘to reel (yarn),’ from pānus,-ī (REW3 no. 2569) and *crēdentia ‘faith, belief’ (UCPL 1.41–86), the latter indubitably a neologism propagated along with the new religion from one early Christian community to another. *Pampanu (for pampinus) ‘young yine branch’ and *cofanu (for cophinus) ‘basket, hamper’ were current prior to the disintegration of the Empire, on the testimony of Spanish and Italian (Manual7 77,131).
  • In addition to areal configuration there exist other norms for the dating of recordless Latin formations. One important criterion is the assibilation of the voiceless velar stop before front-vowels. The by-form *pūlica, from pūlex, -icis ‘flea’ (cf. Sp. pulga) is unquestionably old; the evidence is less conclusive in the case of OSp. amizdat, Sp. amistad, Ptg. amizade ‘friendship,’ from *amīcitāte in lieu of amīcitia (Manual7 155,157,162). Early metathesis must be assumed in the development of EAst. escripia, CAst. esquirpia (the latter with regression to the initial sound sequence) ‘side board (of a wagon)’ ‘texture of rods placed around the bottom of a wagon’ < *scripea < scirpea ‘tissue of rushes’ (Menéndez Pidal, Rom. 29.350). Metaphony seems to justify the early dating of *uariola (from uarus ‘scab on wounds’) > OSp. veruela > Sp. viruela ‘pock’ (Manual7 38–9, 68). The complete disappearance of the primitive strengthens the case of *gemellïciu > OSp. (e)mellizo ‘twin’ (Ibid. 124). Morphological considerations buttress the assumption of the early rise of *uīstu ‘seen,’ *mouitu ‘moved; motion’ > Sp. muévedo ‘aborted fœtus,’ OSp. muebda ‘movement, instigation,’ *uersūra ‘sweeping’ > Sp. basura ‘rubbish’ (Ibid. 76, 78, 118, 136).
  • Cf. Thomas, Mélanges2 70–1; in other cases, the author has carefully refrained from projecting into Latin such processes as were characteristic of Gallo-Romance, see Ibid. 41–2, s. v. Occ. aguio bastaresso. On the distortion of the historical perspective inherent in such a projection, see Lerch, RF 60.647–84 (1947), reviewed by G. Gougenheim, FM 17.156 (1949); M. L. Wagner, VR 9.299–302 (1947) and Biblos 24.8–9 (1949). Menéndez Pidal's Manual, although repeatedly brought up to date by its tireless author, nevertheless has carried over from the late 19th century many bases gratuitously attributed to late Antiquity: Sp. corredera ‘upper grinding stone’ < *curritōria, Sp. abrevadero ‘watering place’ < *adbiberātōriu (65); Sp. desahuciar ‘to declare (a patient) past recovery’ < *disaffīdūciāre (130); *incaballicare ‘to ride,’ *incominitiāre beside *excominitiāre ‘to begin,’ *deexpergitāre ‘to wake up’ (332). Why assume *conti(n)gēscere ‘to happen’ (182), if OSp. (a) contecer coëxisted with the archaic variant contir, cuntir < conti(n)gere, see St. Phil. 38.442–4 (1941)? Why posit *attestificāre ‘to testify’ (74), if Sp. atestiguar perfectly fits the recorded base testificārī, with a-, a notoriously fluid element, easily classifiable as added on in the Middle Ages, see RR 32.278–95 (1941)? Why vacillate between *disdignāre (144) and dēdignārī (330) as the base of Sp. desdeñar ‘to scorn,’ if the latter etymon, aside from its merit of authenticity, is also morphologically satisfactory, with dē- > des- as in deinde > OSp. desende ‘afterwards,’ dētectāre > Sp. desechar ‘to reject’ (see my contribution to the forthcoming F. A. Coelho homage volume)? Is there reason to assume that *alaundula ‘lark,’ *rehinnīntulāre ‘to neigh’ (189) were altogether pronounceable formations within the phonemic system of Latin, plebeian or literary? Is the evidence for *tōnsōria beside tōnsōria ‘scissors’ (197, 231) unshakable, or can the presence of forms in -eira in territories distinguishing between -eiro and -oiro be attributed to the diffusion (and subsequent slight adaptation) of a central form, whose wedges in the direction of the west and northwest are understandable in the case of a word designating an industrially manufactured tool? It is noteworthy that in his real masterpiece of linguistic research, the Orígenes del español (1926), Menéndez Pidal became much more parsimonious in establishing hypothetical bases.
  • If France was the center of radiation for words pertaining to medieval spiritual culture, it is more difficult to determine the focal points for the spread of terms relating to material civilization (migratory words, parole girovaghe, Wanderwörter). On one such case (Fr. brenèche, It. vernaccia, Arag. garnacha), see Thomas, Mélanges2 50–1. A few pertinent problems are illustrated by Bertoldi, Questioni di metodo 87–136. The hazards inherent in such interpretations are discussed in my essay on Sp. bernegal ‘kind of container’ and its congeners, see Lang. 25.165–78 (1949).
  • The reconstruction of extinct forms inferred from borrowings into neighboring languages where they happen to have persisted is as fascinating as it is risky. Submerged Western Balkan-Romance has been pieced together from scraps of evidence extracted from Albanian and Serbian; Basque throws light on pre-literary Ibero-Romance; Latin borrowings in Celtic and Teutonic are central to the comparative investigations of J. Jud; an Old Picard form has left no traces except a casual reflex in the neighboring Teutonic dialect of Western Flanders, see W. von Wartburg, ZRPh. 62.159 (1942). The scholar requires much control over his imagination in inferring hypothetical bases from borrowings; L. Spitzer's untenable derivation of Fr. console ‘pier-table’ from It. *consola (instead of from older Fr. consolateur), AR 23.92–3 (1939), exemplifies the danger of facile approach, cf. E. Lerch, AR 24.167–87 (1940), and W. von Wartburg, ZRPh. 62.152 (1942), with a reference to P. Barbier.
  • Thomas, Nouveaux Essais 11.
  • Word-studies have recently made such headway that correction of hastily established etymologies through phonological analysis is as feasible as is the modification of prematurely formulated correspondences through penetrating inquiry into the development of a single key-word. A. Castro's intensive examination of the products of petra ‘stone’ and cathedra ‘chair, hip’ in Ibero-Romance toponymy has shown the speciousness of the accepted equation Lat. -tr- > Sp. -dr-, see RFE 7.57–60 (1920). What was once dubbed ‘phonetic law’ might more fittingly be labeled as the farthest advance of a sound in a characteristic environment from its initial position; many formations are retarded in the course of their advance or swerve from their straight course; others succumb to analogical influence which may necessitate withdrawal from the advanced front-line. If, under comparable circumstances, a single word (in most cases an inconspicuous one, overlooked in the process of leveling) maintains its original form by remaining exempted from counteractive influences, then it alone may serve to indicate the point of farthest advance and, although numerically an exception, it represents the original norm in evolutionary terms. One solid identification of base and product (Sard, báttile, -i ‘caparison,’ ‘strata di saiale che mettesi sul dorso del cavallo’ < coartile, quactile) has enabled M. L. Wagner, CN 3.13–5 (1943) to reverse a time- honored ‘sound-law’; on the method of the anomalous form in linguistic reconstruction, see G. Bonfante, Word 1.133–4.
  • See the explicit statements by W. von Wartburg, NJbWJ 7.222–35 (1931), and, in greater detail, Einführung.
  • Cf. the eloquent complaint of E. H. Sturtevant in his review of R. G. Kent, The Forms of Latin, AJPh. 69.349–51 (1948).
  • Thomas, Mélanges2 11–2.
  • On this suffix, there exists an elaborate monograph by C. S. R. Collin (Lund, 1918).
  • Thomas, op. cit. 29–30; cf. my article Atristar ∼ entristecer: Adjectival Verbs in Spanish, Portuguese, and Catalan, St. Phil. 38.429–61 (1941).
  • Thomas, op. cit. 15–6, 69. OFr. espois, Fr. épais ‘thick, dense’ was, in turn, modeled on espoisse ‘thickness,’ much as Sp. lobregura, originally ‘moisture and darkness (as in caves),’ under the pressure of tenebregura and negregura ‘darkness’ (with which it shared most of the word body and one component of the complex meaning), experienced a narrowing down of the connotation and effected a similar change iii its primitive lóbrego, see Lang. 25.159–65 (1949). The interplay of gros ∼ groisse is responsible for the permutation of Fr. groselle ‘currant’ (< Krausbeere) into dial, groiselle, see Wartburg, ZRPh. 62.220 (1942).
  • Thomas, Mélanges2 32–3, 70; cf. ibid. 20–1, 41, on (h)amula ∼ (h)amella ‘small bucket’ and margula ∼ *margella. See Rom. 29.342 on fībula ∼ *fībella > Sp. hebilla ‘buckle, clap’; Manual7 227–8 and UCPL 1.269 (1947). On -ǐculus ∼ ĭcellus, see E. Gamillscheg, ZFSL 62.229–51 (1939). A last symptom of the vitality of -ulus, -ula was the coinage of *tenula (for tenus, -ūs or -oris ‘cord, snare’) > ORioj. tienila (Rom. 29.372–3); -ulāre remained longer alive, witness *misculāre ‘to mix’; *tremulāre ‘to shiver.’
  • Knowledge of protracted vacillation between -ānu and -āneu, -āre and -iāre paves the way for Sp. temprano ‘early’ < *temporānu (not temporāneu), Sp. domeñar ‘to tame, to subdue’ < *dominiāre (not domināre; coëxistence of dominus and dominium may have had a catalytic effect). See my forthcoming study of duendo, duende in the A. M. Huntington homage volume.
  • Thomas reconstructs a wholly plausible *cancerēus (Mélanges2 65), in accord with what is known about the spread of -eus; his *maculentāre ‘to cover with spots’ > OFr. maillenter ‘to soil,’ Morv. aimaillenter ‘to bruise, to crush’ (ibid. 10–1) seems to carry conviction. It is less easy to accept his series Lat. acer, V. L. acrus (Appendix Probi) ‘sour’ > Gall.-Rom. *acrācius > *acrāciolu > Lang, agrassol ‘currant-bush’ (ibid. 8–9) in view of scant information on the coalescence of -āceu and -iolu, -eolu. Wartburg's new derivation from Germ. Krausbeere gives the coup de grâce to a hypothesis inherently fragile: agrassol is a local blend of Fr. groseiller with the word for ‘sour.’
  • Typical examples were *abbrachicāre, *fallitāre, *pellicicāre, *quālāneus, suggested by authoritative scholars as bases for Sp. abarcar ‘to contain, to comprise,’ faltar ‘to fail,’ pellizcar ‘to pinch,’ and OSp. calaño ‘similar.’ Did Menéndez Pidal, in supporting Nebrija's derivation of sanguisuela, -juela ‘leech’ from *sanguisūgēla (allegedly derived from sūgere ‘to suck’ on the analogy of cand-ēla ‘candle,’ quer-ēla ‘complaint,’ suād-ēla ‘persuasion’), take into account the fact that -eia was an unproductive suffix in Late Latin (Rom. 29.370; cf. J. Corominas, AILC 2.149–50; Y. Malkiel, Lang. 22.316)? The sterility of -undus (as against -ibundus) casts suspicion on the derivation of Sp. orondo ‘vain, proud,’ ‘hollow, swollen’ from *aurundus, advocated on the assumption of a secondary meaning (“”swollen by the gentle breeze like the sails of a boat'), see Rom. 29.361; Manual7 71. J. Corominas' discovery of the by-form jorondo has opened up different avenues of approach, see AILC 1.154–60 (1941).
  • See the excellent discussion of l'ajoux < la joue ‘cheek’ and l'assure < laçure ‘lacing’ by Thomas, Melanges2 12–4, 30–1, with full account of the processes of agglutination and deglutination of the article.
  • There is no space here to present an elaborate restatement of the various norms used by linguistic geographers. Some workers have striven to evolve fixed norms applicable, in the proper sequence, to any language or group of languages (M. Bartoli, J. H. Bonfante). The majority are satisfied with working out a scheme valid within the confines of a given territory throughout a more or less flexible period of time; the chief emphasis is placed on the clear indication of major and minor focal points, recess areas, and channels of transmission, in harmony with the political, economic, and social configuration of the chosen Fzone (J. Jud, K. Jaberg, V. Bertoldi, A. Dauzat, J. Orr).
  • Romanisches etymologisches Wörterbuch3, p. x.
  • Notice in this connection Ptg. leirāo ‘dormouse’ < *glēriōne, Sp. lirón < *glīriōne (rather than *glīrōne, as indicated in the Manual7 127); the suffixes -ō, -ōnis and -iō, -iōnis may represent blends of a Latin formant with an element traceable to the Mediterranean substratum, see Bertoldi, Questioni di metodo, 204–5, 255–8.
  • Thomas, Mélanges2 40–1; *berbicīle was substantivated at an early date.
  • Ibid. 68–9: Occ. cabes, OFr. chavessot; Occ. cabotz, Fr. chabot < chaboz; Jur. chevasson, Rhod. chavasson.
  • Ibid. 66–7; products include Berr. chebiche, Dauph. chavisso; Rouerg. cobis; Lyonn. chavassi.
  • Ibid. 23–4 and under errata; M. L. Wagner, Biblos 24.14–8 (1949). Cf. *nīdāle ∼ *nīdāriu ‘pertaining to the nest,’ originally qualifying ōuurn ‘egg’ (Menéndez Pidal, RFE 7.34–5); *sērðtinu, -a ‘belated’ (> Cast. Ast. seruenda, Ast. sebreñu) ∼ *sēruculu, -a (> Sp. serojas ‘pieces of dry wood, leaves falling from the trees’) ∼ *sērotinālia (> OSal. serondaja ‘late grains,’ OSp. çarandaja ‘dry leaves, crumbs falling from the table’), Menéndez Pidal, Rom. 29.371; RFE 7.27–9; *cannica > Gal. Ptg. canga ‘yoke,’ Ast. canga ‘narrow passway between rocks’ ∼ *cannābula > Arag. cañabla, canaula ‘collar for cattle,’ see C. Nigra, ZRPh. 27.129–30 (1903); R. Menéndez Pidal, RFE 7.25–7; crepita, *crepta > Sp. grieta ‘crevice’ ∼ *(re)-excreptiāre > OSp. rescrieço > Sp. resquicio ‘chink, slit,’ quicio ‘eye of a door hinge’ (RFE 7.24); *budōne (from buda ‘swamp grass’) > Seg. Vail. Sal. Extr. bodón, bodonal, bohonal ‘bog, quagmire’ ∼ *budētu + āle > Pal. Gal. topon. Buedo, buhedal, Extr. budial (RFE 7.20–2); *stuppāciu > OSp. estropaço ∼ *stuppāculu > Sp. estropajo ‘mop swab’ (Rom. 29.352); *caducula (from cadus ‘wine-jar’) > Lit. cadolla ‘small cavity opened in a rock to gather rain water’ ∼ *cadōceu > Zam. ca(d)ozo ‘swampy ground in a river bed,’ OSp. cadozo ‘deep spot in a river or lagoon’ (RFE 7.24–5); *gemellica (from gemellus ‘twin’) < OSp. emelga, Sp. dial. (e)melga ‘even division in a piece of tillable land’ ∼ *gemellīciu > OSp. (e)mellizo ‘twin’ (Rom. 29.337–8); *pedicus ‘pertaining to the foot’ (REW3 no. 6352) ∼ *pedicellus ‘itch-mite’ (REW3 no. 6349) ∼ *pedicullus ‘leaf-stalk’ (REW3 no. 6351); *cappāneu ∼ *cap(p) āceu ‘basket,’ see Bertoldi, La parola quale testimone della storia 75–4 (Naples, 1945), REW3 no. 1643.
  • Thomas, Mélanges2 46–7. Comparable mutual support is lent by OSp. peaño ‘footwear,’ Sp. peldaño ‘step of a staircase’ < *pedāneu and OSp. aledaño, aladaño ‘contiguous’ < *latāneu, from latu(s) ‘side’ (Rom. 29.335); by WAst. abanigar ‘to move,’ Các. envanguear ‘to place insecurely,’ Sal. abangar ‘to warp’ (of wood) < *uānicāre and Ast. allancar ‘to get stuck’ < *planticāre beside Fr. lâcher, Occ. lascar, It. lascare ‘let loose’ < *lassicāre, see G. Tilander, Étymologies romanes, St. N. 19.229–309 (1946–7); by *comperāre (for compareāre) > Sp. comprar ‘to buy’ and *sēperāre (for sēparāre) > It. scevrare, Ast. enxebrar ‘to segregate’.
  • A complicated and hazardous derivation is Béarn. arroumera ‘to wind into balls’ < *romellāre < *lomellāre < *glomellāre < glomerāre; on the other hand, Lyonn. charolesse ‘road suitable for the passing of wagons’ < *charraresse < *uia carrāricia is supported by It. carrareccia, see Mélanges2 28–9, 66. An example of a long chain of successive assumptions in Spanish is tolondro ‘bump from a blow’ < *torondro < *torondo < *torundu < toru ‘protuberance’ (Rom. 29.373–4).
  • And yet such constructions are found in the best writers. To explain a few regional variants (For. chambucle, Lyonn. charbuclio), Thomas, Mélanges2 62–3, had recourse to *carbūsculus, an assumed by-form of carbunculus. Can subsequent modification not be involved?
  • Cf. Thomas' derivation of OFr. adcier < *ad-aciāre, preferable on numerous grounds to earlier fanciful reconstructions (*acaciāre, *agriāciāre, *alligātiāre, *exaciāre, Teut. *hwat-jan, *azjan), see Mélanges2 1–2. An attempt was recently made to trace Sp. rastrojo, Ptg.res- tolho, Cat. rostoll ‘stubble’ to *rōstruculu(m) 'little sharp end'in preference to restruculu(m), see RPh. 1.209–34 (1947–8). OFr. ferner ‘to blame, to punish, to reprimand’ was long connected with *ferīnāre on the basis of inaccurately defined meaning (REW3 no. 3252; FEW 3.465); F. Lecoy, Rom. 70.333–6 (1948–9) argues convincingly in favor of *ferulāre.
  • Overscrupulousness has occasionally led to an unnecessary increase of hypothetical bases. Though the oblique case, in all probability, is based squarely on the Latin accusative, there is hardly any need to star *culmine ‘summit,’ *sulphure ‘brimstone,’ *ūbere ‘teat, udder’ (Manual7 140,215), which are recorded forms. Why not adopt a more flexible formulation of the Vulgar Latin case system, allowing for the persistence of the ablative as a substitute for a few abnormal accusative forms? Cf. M. A. Pei, RR 28.241–67 (1937), 30.189–91 (1939). Similarly, is there any valid reason for spelling the base of Sp. cañaherla ‘fennel giant’ *cannaferula rather than canna ferula (Ibid. 241); cf. pīnī pullu < Sp. pimpollo ‘sprout, shoot,’ päut pullu > Sp. pavipollo ‘young turkey,’ rāmī pullu > It. rampollo ‘offspring’ (Ibid. 238; it is dubious whether Menéndez Pidal's bases pīnuspullus, pavuspullus are felicitous).
  • Thomas, Nouveaux Essais 21, 37; cf. Fr. allier ‘net to catch partridges,’ traced to *ālitārium (from āles, -itis ‘winged, swift’) by Ménage, yet to āiārium by Thomas (Melanges2 16–7) on the evidence of Sp. alero. Similarly, OSp. hojalde, Sp. hojaldre ‘puff paste’ cannot go back to *foliandrinu (Puigblanch), but only to foliātile (Rom. 29.335); OSp. ahaxar, Sp. ajar ‘to crumple, to rumple,’ in which J. Brüch was inclined to see a descendant of Goth. *aljan, *af-aljan, see ZRPh. 36.577 (1912), actually perpetuates facula ‘faggot, torch’, see Menéndez Pidal, RFE 7.9–12 (1920). In both cases improved interpretation was achieved with the aid of ancient forms.
  • However, the accuracy of this derivation has seriously been questioned by F. Lecoy and J. M. Piel. Other examples may be more convincing. Occ. grifouni'e ‘noise of the tempest, roaring of the rough sea’ might have been traced to an artificially erected base but for the rich cluster of variant forms (broufounié, brefounié, brafounié) which led Thomas straight to OProv. brefania < Epiphania, a holiday commemorated in a mood of loud hilarity (Mélanges2 51–2). Fr. dial, ambersac ‘wallet’ might have suggested a conjectural combination involving ambō ‘both’ but for close congeners like havresac, habersac, haubresac, which all point to Germ, habersack, hafersack, borrowed during the Thirty Years War. Two plants named ivrogne in regional French seem to presuppose * ēbriōnia, from ēbrius ‘drunken’; comparison with cognate designations (in one case, Fr. dial, abrone, avrone, lavrone, abroigne; in the other case, Occ. liborna) leads to the authentic etyma abrotonu and liborna, the latter modified through the deglutination of the initial consonant mistaken for the definite article, see Bertoldi, Questioni di metodo 84–5.
  • Thomas, Mélanges2 7–8.
  • Ibid. 69.
  • On the obsoleteness of this type of research, see Wartburg, Einführung 105–7; Wagner, CN 3.15 (1943).
  • For a person trained in rational thinking it is not obvious that the same word should signify the cub of an animal and the new shoot of a plant (Thomas, Mélanges2 69–70, on Lat. pullus and Occ. cadel, Fr. dial, chiau < OFr. chael < catellu); that a fish should be named after a bird, because its snout has the shape of a beak (Ibid. 3–5); that the name of an agricultural appliance may become a term of navigation (Ibid. 12); that overseas sailors confuse their native plants with others never seen before (Ibid. 36–8, on the contamination, probably in Smyrna, of Low Greek βαλανiδια ‘oak’ with Occ. avelanedo ‘hazel-nut grove’); that clusters of worms used to fish eels should be likened to a bunch of grapes (on Occ. bouiroun, Cat. botiró < bortyōne, see Bertoldi, Questioni di metodo 265); that a word for ‘cheek’ like Fr. joue may serve to designate an infinite variety of symmetrically built tools (Thomas, Mélanges2 12–4). The elimination of the reconstructions *camice, *camite as bases for Fr. dial, chaintre ‘furrows traced at the end of a field in the reverse direction’ (Mélanges 63–5) was made possible only by Thomas' inquiry into the range of images suggested by cancer and canceling ‘crab.’ The etymology of Spanish words for ‘prodding animals, especially dogs’ (OSp. açomar, Sp. azomar < summu and azuzar < sūrsu) was suggested by the inherent image known from a familiar romance (jarriba, canes, arriba!; see Rom., 29.339). Sp. escamondar ‘to prune’ was traced to hypothetical etyma (Cabrera: < *es-commundāre; Diez: < *escami-mondar; Parodi: < *ex-capu-mutāre) until Menéndez Pidal observed that ‘clearing a table from crumbs’ and ‘clearing a tree from useless branches’ were two activities closely associated by naïve speakers of Spanish; the way was paved for the authentic base ēscam mundāre (Rom. 29.347). Images are subject to diffusion much as phonemes and morphemes, and the starting-point of a successful pictorial equation cannot always be easily ascertained; see B. Migliorini, Calco e irradiazione sinonimica BICC 4.14–28 (1948).
  • *Sessicāre is accepted by C. Michaëlis de Vasconcelos, Miscellanea Caix-Canello 156, and RL 3.186; L. Spitzer, Notes étymologiques, RFE 13.116–7 (1926); Meyer-Lübke, REW3 no. 7879; on *subsedicāre, see J. Storm, Rom. 5.184 (1875). The Gallo-Romanie base *sedicāre hae incomparably stronger foundations (REW3 no. 7782: OFr. siegier, OProv. setjar).

Reprints and Corporate Permissions

Please note: Selecting permissions does not provide access to the full text of the article, please see our help page How do I view content?

To request a reprint or corporate permissions for this article, please click on the relevant link below:

Order Reprints Request Corporate Permissions

Academic Permissions

Please note: Selecting permissions does not provide access to the full text of the article, please see our help page How do I view content?

Obtain permissions instantly via Rightslink by clicking on the button below:

Request Academic Permissions

If you are unable to obtain permissions via Rightslink, please complete and submit this Permissions form. For more information, please visit our Permissions help page.

Related research

People also read lists articles that other readers of this article have read.

Recommended articles lists articles that we recommend and is powered by our AI driven recommendation engine.

Cited by lists all citing articles based on Crossref citations.
Articles with the Crossref icon will open in a new tab.