¹ The older grammarians frequently cite unusual and antiquated formations in -azo and -acho, no doubt also nonces, ghost-words traceable to misprints or misreadings, and artificial formations playfully coined by lexicographers. Neither in the latest, generously expanded edition of the Academy dictionary (1956), nor in the bilingual dictionaries of R. J. Slabý (1932) and E. B. Williams (1955), both based on independent, wide-ranging excerpts, does one find a trace of acicatazo (Diez, Förster), anchazo (Diez, Förster), arbolazo (Förster), asnazo (Meyer-Lübke), balconazo (Meyer-Lübke), bezacho (Diez, Förster), camisaza (Meyer-Lübke), carnacha (Diez, Förster), dulzazo (Hanssen), feazo (Wiggers), grandazo (Förster, Alemany Bolufer), grandonazo (Wiggers), haraganazo (Alemany Bolufer), herejazo (Wiggers), mujeraza (Diez, Förster, Nyrop), señorazo and simplazo (Alemany Bolufer). These formations are all transparent, giving the impression of authenticity, if not of currency. Downright errors include, in Diez : pigaza (which he himself later emended to picaza ‘magpie’; however, pic- and pig-aça both occur in Don Juan Manuel) and grimazo ‘wry face’ (which, unless spurious, at best represents an ephemeral Gallicism); in Meyer-Lübke: bastonazo and cordonazo, inadvertently listed among the augmentatives ; espinaza, a conspicuously rare Old Spanish formation (see n. 126, below) possibly confused with trivial espinazo; algazo, arbitrarily defined ‘Menge Algen’ (absent from Dicc. Hist.; on algazo as a facetious augmentative of algo ‘something’ [D. Sánchez de Badajoz] see UCPL 1: 9.398 [1948]); undefined espadetazo—only in A. and G. Doutrepont's translation—instead of either espadazo or espadillazo (a lapsus conceivably suggested by the preceding word, escopetazo); morrazo, dubious in the given acceptation (‘Morraspieler’; on dial, morrazo ‘puñetazo’ [El Bierzo], ‘coscazo’ [Salamanca], pl. ‘enfados, berrinches’ [Aragon] see BH 57.124-5 [1955]). In 1906, G. Baist subjected to penetrating criticism Meyer-Lübke's array of formations in -acho and -acha, some of them irrelevant or unauthentic, see n. 73, below. More harmful than the aggregate of such easily detectable blunders has been the failure of leading grammarians to place the quoted words in a forceful perspective by qualifying the rank they occupy within the total structure of the lexicon. Dulzazo, for instance, measured by any valid yardstick, is vastly less important than dulz(arr)ón ‘sickening, cloying’ and even than coll. dulzaino; asnazo fails to compare favorably with asnejón; v(i)ejazo (adj.) ‘oldish’, (subst.) ‘silly old fellow’ seems to trail behind vej-ancón, -estorio, -ete, -ote, v(i)ej-arrón, -ezuelo, Am.-Sp. viej-ón, -ito. Three more errors of omission permeate the older treatments: neglect of loanwords (as late as 1918 Alemany Bolufer failed to point out It. pagliaccio as the obvious prototype of Sp. pall-azo, -aso, payaso 'clown, laughingstock'), disregard of historical stratification (pocazo alone was segregated as obsolete by Diez and his closest followers), and lack of discrimination between genuine derivatives, possessing two neatly detachable morphemes, and mere relics, not analyzable by the average speaker, such as capacho 'hamper' (-acho lends itself to association, but truncated cap- makes no sense to the uninitiated).

² Only in rare instances (as in the case of 'peck') does a single English word suffice to render the Spanish derivative, with a heavy loss to its graphic suggestiveness : capirotazo 'fillip' (capirote; also pap-), espaldarazo ‘accolade’ beside ‘slap on the back’ (espaldar), puntillazo ‘kick’ (puntillas), trompazo ‘fisticuff, bump’ (trompa coll. 'protruding part of face'). In general, translations will vary with the context: bolazo may suggest a 'blow' or a ‘hit’, hurgonazo a ‘blow’ or a ‘thrust’, rodillazo a ‘push’ or a ‘blow’, alazo, hachazo, and tizonazo a ‘stroke’ or a ‘blow’, brochazo a ‘stroke’ or a ‘dab’, aguijonazo a ‘thrust’ or a ‘prick’, in accord with the shape and size of the tool, the momentum and the impact of the motion, the angle of the hit, the substance of the object affected, and the intent of the actor.

³ Modern French has recourse to coup (d'épée, d'épingle, de foudre, etc.); Sp. golpe de estado is a loan translation of coup d'état. In German, compounds do adequate service for all ordinary purposes (Blitzschlag, Degenstoss, Kanonenschuss, Säbelhieb), but fail to convey the gamut of overtones, ranging from mildly comic to downright grotesque, peculiar to Sp. -azo ₂; contrast Telephon (an)ruf with coll. telefonazo 'short, informal phone call made on the spur of the moment'. Danish operates with compounds or paraphrases: ‘stokkeslag’ = bastonazo, ‘φxehug’ = hachazo, ‘smæk med en dør’ = portazo (K. Nyrop, Spansk grammatik 31 [Copenhagen and Oslo, 1920]). Russian makes ample use of its reservoir of adjectives: ‘púšečnyj, ružéjnyj výstrel’ (beside 'výstrel iz púški, iz ružjá' = cañonazo, fusilazo), or resorts to the instrumental case: 'udár bašmakóm' (more idiomatically: ‘kablukóm‘) = zapatazo.

⁴ Pérez Galdós wrote in Misericordia, Ch. 3: ... que no te falte para el aguardentazo. Here are some colorful regionalisms accompanied by their original definitions: Am.-Sp. cacharrazo ‘trago de licor fuerte’ (DH), C.-Am. calazo (DH) ‘golpe o cachada que recibe el trompo en el juego’ (Guat.), ‘golpe, trago de aguardiente’ (Hond.), Dom. petacazo ‘trago de bebida fuerte’ (Henríquez Ureña), C.-Ast. (Cabranes) sorbiatazu ‘ruido de sorber’, from sorbiatar ‘beber a sorbos y haciendo ruido’ (Canellada). Cf. n. 61, below.

⁵ A border-line case is vistazo ‘brief look, glance’ (vista), which, to be fully appreciated, must be set against the phraseological background of clavar (fijar) los ojos, la mirada ‘to gaze’, lit. ‘to nail or bore one's eyes’.

⁶ The basic meaning of golletazo comes up for discussion immediately below. The sense development of ramalazo ‘mark or spot caused by disease or sickness’ follows a different line. Sp. fusilazo ‘heat or sheet lightning’ calls to mind Tig. fuzil ‘lightning’ and must be linked to an early evolutionary phase of *focīle (REW ³ §3399). Note also the Academy's twin definitions of candilazo ‘golpe dado con un candil’ beside (fig.) ‘arrebol crepuscular’.

⁷ If this analysis is valid, golpazo exemplifies a process catalogued as ‘suffixale Einreihung’ by Meyer-Lübke and as ‘inquadramento suffissale’ by Migliorini; for more detailed discussion see my article Diachronic hypercharacterization in Romance, in Arch. L. 9.79-113 (1957), 10.1-36 (1958). A parallel would be provided by batacazo ‘loud noise, violent contusion from a fall’, if this word were based—as the Academy assumes but fails to demonstrate (metathesis?)—on bacada, from baque ~< Ar. wāqi' 'falling'; cf. n. 63, below, and my comment on batatazo.

⁸ Here are Henríquez Ureña's respective definitions: ‘golpe que se recibe accidentalmente en la cabeza’ and ‘pequeño golpe en la cabeza por castigo’. On the other hand, Chil. Mex. caballazo ‘jolt, trampling with a horse’ (Román: ‘encontrón o golpe que da un jinete a otro’; [fig.] ‘acometida que, de palabra o de hecho, da una persona a otra‘) endows the charging horse with the power of a weapon (much as in this country an automobile sometimes acts as a lethal weapon).

⁹ In the historical perspective espald-ar-azo is based on espald-ar ‘backplate of a cuirass’, a term of chivalric culture. Nowadays speakers associate it most readily with espalda(s) ‘back’, a shift which relegates the segment -ar- to the position of a void nonfinal suffix. See my article Los interfijos hispánicos, in Estructuralismo e historia: Miscelánea homenaje a André Martinet 2.107-99 (Madrid and Tenerife, 1958). In the abnormal coinage of espaldarazo the preexistence of semantically adjacent espinazo ‘spine, backbone’ (involving -azo ₁ or even directly -áceu) may have played the decisive part. A further illustration is provided by the specialized meaning that cañazo (ordinarily 'blow with a reed') has acquired in Cuba, in reference to the gamecock: 'golpe que se da el gallo de pelea en las cañas o piernas'. This dichotomy -azo ₂ shares with its rival suffixes expressive of a blow: cf. pescoz-ada, -ón ‘slap in the neck’, cosc-orr-ón ‘bump on the head’ beside dial, coscazo, morrazo (see n. 1, above).

¹⁰ Off-hand one is tempted to class araño as a normal postverbal of arañar; granted that much, arañazo would be definable: (a) as an augmentative of araño, involving -azo ₁: (b) as an elaboration on araño, by means of a hypercharacterizing suffix (-azo ₂) ; (c) as a combination of (a) and (b), cf. the earlier dilemma posed by golpazo. In present-day usage arañazo is distinctly more common than araño; this situation opens up the additional possibility of analyzing arañazo as an immediate offshoot of arañar, and araño as a regressive formation. Even if the records should not bear out this last suspicion, the fact remains that in contemporary Spanish arañar and arañazo are allowed to team up, to the detriment of araño. In the case of picazo one must distinguish between three sources, semantically identifiable: ‘blow with a pike’ (pica), ‘stroke with a bird's beak’ (pico), ‘sting’ (picar). Here the first and the second variant, both morphologically unexceptionable, may have blazed the trail for the third, despite the strained relation between picar and picazo. At present the recoil from polysemy favors picotazo ‘peck’ and picadura ‘sting’.

¹¹ In this essentially synchronic survey it seems unnecessary to segregate with any degree of stringency borrowings (partial or total, presumable or proven) and (semi)learned formations, such as gigantazo, sanguaza, and terraza. These points will be taken up in the subsequent stratigraphie analysis.

¹² Examples in point include: (a) rambl-azo (beside -izo) ‘bed of a torrent’ (rambla 'sandy, dry ravine'), sarnazo 'malignant itch' (sarna); (b) coraza ‘cuirass’, fig. 'shell (of a turtle)' (cuero), hilaza 'yarn, fibre, uneven thread' (hilo), linaza ‘linseed’ (lino), oleaza ‘watery dregs in oil-mills’ (óleo), pelaza ‘chopped or beaten straw’ beside more descriptive paja pelaza (pelo), trigaza ‘short straw of wheat’ (trigo) ; (c) gallinaza ‘hen dung’, (orn.) ‘turkey buzzard’ vs. gallinazo ‘(male) turkey buzzard’ (gallina), hormaza ‘dry wall’ vs. hormazo ‘heap of loose stones’, dial, ‘house and adjoining orchard’ (horma), hornaza ‘jewellers’ furnace' vs. hornazo ‘Easter cake filled with hard eggs’ (horno), limaza ‘slug’ beside limazo ‘sliminess’ (limo 'mud'), vinaza 'wine drawn from the lees' vs. vinazo ‘very strong wine’ (vino).

¹³ Pelaza (alongside pelazga) ‘quarrel, affray, scuffle’, removed one step from pelo through its closer affinity to pelar ‘to pull out the hair’, pelear ‘to fight’, (refl.) ‘to scuffle, come to blows’, and pelea ‘quarrel’, obviously represents a case apart. For details see n. 125, below.

¹⁴ V. García de Diego, Elementes de gramática histórica castellana §200 (Burgos, 1914), quotes bellaconazo (bellaco 'scoundrel'), hambrentonazo (hambriento 'hungry'), machonazo (macho 'virile, male'), mujeronaza (mujer 'woman'), valentonazo (valiente 'brave fellow, bully'); cf. Cuervo's data cited in n. 41, below. Typical examples of the parallel pattern include hombr-acho, -achón 'heavy-built man' (hombre), pobl-acho, -achón ‘large, ugly village’ (pueblo), (coll.) porcachón ‘very dirty, hoggish’ (puerco), ric-acho, -achón ‘vulgar rich person’ (rico). Latent possibilities of further proliferation are unlimited, in particular with emotionally laden words such as mujeracha ‘vulgar woman’. Note hornachuela ‘hut, cabin’, from hornacha, the obsolete variant of hornaza (which itself is flanked by hornacero ‘crucible man’ and hornacina ‘vaulted niche’, while hornaguero ‘wide, spacious’, ‘coaly’, hornaguera ‘pit coal, hard coal’ and fornaguear ‘to dig for coal’ beside subst. fornagueo reflect fornacārius), and capachazo 'blow with a basket (lid)', where the quasi-suffix -ach(o) is followed by -azo ₂. Though in mostacho 'mustache' -ach(o) fails to act as a neatly detachable suffix, the stimulus for launching mostachón ‘kind of macaroon; diamond-shaped ornament’ in the last analysis may be traceable to this series. The ever-present possibility of a nasal insert, in anticipation of word-final n (see NRFH 9.266-73 [1955]), materialized in the case of corpa(n)chón ‘large body’ (cuerpo), in all likelihood through concomitant association with ancho ‘wide, broad’ (much as a slight measure of contact with its antonym angosto ‘narrow’ sufficed to transform OSp. lagosta ‘locust’ into langosta). The compound suffix -achón (var. -anchón) deserves joint study with -ajón, -ancón, -arrón, -astrón, -ejón, -etón, -icón, -ijón, and others similarly structured. In modern fiction one finds (adj.) bonachón ‘good-natured’ (Pérez Galdós, Doña Perfecta), corpachón (ibid.), hombrachón (Pardo Bazán), poblachón (Doña Perfecta), ricachón (Pérez Galdós, Misericordia [in direct discourse]): add from dialect speech And. holgachón ‘idler’ (Toro y Gisbert) and C.-Ast. picachón ‘pickaxe’ (Canellada), and observe the conflicting schemes of hombr-ach-ón and hombr-on-azo. The simplest way to explain -ancho, a rare variant of -acho (Ast. lebranchu = Ptg. lebracho ‘leveret‘) is by assuming back-formation from -a(n)chón; cf. Sp. lebr-ato, -atón.

¹⁵ Cf. E.-Ast. ferronazu ‘golpe que recibe una peonza con la púa de hierro [ferrón] de otra’ (Vigón, Juegos 94); similarly, C.-Ast. espetonazu ‘skewering’, which Canellada derives by implication from espetu, although it calls to mind more readily espetón ‘spit, rake, ironprong’. From here -onazu, suggestive of a blow, spread vigorously in dialect speech, cf. C.-Ast. (Cabranes) abientonazu ‘sudden attack, fit’, from abientar ‘to winnow’, and especially sofronazu ‘rude answer, public display of contempt’ (farther west, according to Acevedo and Fernández, 'severe reprimand, exhibition of anger'), in preference to sofrenazo 'dressing down', lit. ‘sudden check of a horse’.

¹⁶ Aside from golletazo, the -et- segment is firmly soldered to the radical in bonetazo ‘blow with the bonnet’ (bonete), Per. Chil. camaretazo (Arjona) ‘explosion of a small mortar’ (camareta), Sp. chafarotazo ‘blow or stroke with a cutlass’ (chafarote), Dom. chancletazo (Henríquez Ureña) ‘hit with a slipper’ (chancleta), Col. Dom. fuetazo (Flórez, Henríquez Ureña) ‘blow with a horsewhip’ (fuete), Sp. pelotazo ‘hit with a ball’ (pelota), and Dom. toletazo (Henríquez Ureña) ‘blow with a club’ (tolete). Of these words, bonete, fuete, and pelota are no further analyzable Gallicisms, whereas the naïve speaker is likely to relate camareta to cámara ‘chamber’ (of a firearm), and chancleta to chanclo ‘clog, galosh’. In C.-Ast. llambiotazu ‘(noise made in) licking’ (Canellada), -ot- has been carried over from the verb llambiotar, the frequentative of llamber (= Cast, lamer). The gradual spread of -et- and -ot-is amenable to observation: Cast, picotazo (Doña Perfecta, Ch. 9: Dándose de picotazos y agitando las alas) contrasts with Segorbe picazo (Torres Fornés, Voces 268; cf. n. 10, above) ; concealed behind puñetazo (Lope de Rueda), dial. (La Lomba, León) puñitazo (C. Morán, BRAE 30.455b [1950], s.v. vidaya) are older puñazo, at present confined to Aragonese (Torres Fornés, loc. cit.; Borao; A. Kuhn, RLiR 11.233 [1935]), and synonymous puñete (DCELC 3.926a). The same conservative area shelters pisazo ‘heavy stamp on one's foot’ (Torres Fornés, Borao, Kuhn), whereas Castilian wavers between pis-ada, correlated to pisar ‘to stamp’, and pis-ot-ón, matching frequentative pis-ot-ear, and manages without a counterpart in -azo. Coll. Arg. timbrazo ‘loud ringing of the bell’ is flanked by timbretazo (Juan Goyanarte, Sur 224.118 [1953]), conceivably ‘loud repeated ringing of the bell’. Dom. cocotazo and Sp. capir-ote, -otazo beside papirotazo ‘fillip’, incongruously aligned with papirote ‘hood’, pose special lexical problems (see J. Corominas, DCELC 1.651a).

¹⁷ The family branch that sprouted from the doublet aguacha ‘stagnant water’ affords a derivational pattern equally, if not more, complex; see RPh. 2.63-82 (1948-9).

¹⁸ One such rare instance is C.-Arg. bolacear ‘to tell lies’, from a local figurative use of bolazo ‘lie, nonsense’ (Vidal de Battini). The Academy Dictionary records escobazar ‘to sprinkle with a broom’. W.-Ast. escachazarse ‘to break into small pieces’ (Acevedo and Fernandez) calls to mind cachazo Sal. ‘blow on the back of the neck’, Am.-Sp. (esp. Dom.) ‘thrust with horns’ (Henríquez Ureña), from cacho ‘horn’. Mex. balacera ‘hail of bullets’ contrasts with Arg. (Buenos Aires) baleo; see E. Montenegro, Anecdotario mejicano, La Prensa (B.A.), 25 Jan. 1959.

¹⁹ The same preference for variation that blocks the spread of -zazo (mazazo ‘blow with a mace’ [maza] seems less common than mazada) impedes the growth of -dada, -tada: witness the recession of codada ‘blow with the elbow’ before codazo and the undisputed prevalence of zapatazo (also, in its wake, of zapatillazo ‘blow with a slipper’, from zapatilla). As usually happens where resources are overabundant, subtle distinctions are allowed to creep in: thus, graphic escobazo ‘blow with a broom’ (A. Palacio Valdés: darle un formidable escobazo en la cabeza) and more subdued escobada ‘sweeping, movement of a broom’ are kept apart, but may regionally coincide, as has happened in Chile and Argentina, with escobazo emerging as the victor. Only the equation of -azo ₂ = -ada in the domain of blows explains the transfer of -azo to gatazo ‘artful trick, cheat, deception’ beside ‘large cat’.

²⁰ From pescuezo ‘neck’ the rival derivatives, as expected, are pescoz-ada and -ón ‘slap on the neck’ (see n. 9, above) ; observe again the avoidance of *-zazo, as against pestorej-azo, -ón.

²¹ See E. Gamillscheg, Zur Frage der Auswahl bei der suffixalen Ableitung, [Dietrich] Behrens Fe6stschrift 56-76 (Jena and Leipzig, 1929; ZFSL Suppl. 13).

²² D. Alonso, BRAE 33.386 (1953), illustrates dar un portazo ‘to bang the door’, dar (un) cerrojazo ‘to slam the bolt’; note dar (pegar) un batacazo ‘to fall down badly, (polit.) be overthrown’. The relationship of batatazo to etymologically obscure batacazo raises a separate problem overlooked by Corominas (DCELC 1.424a). Add coll. de bolazo ‘carelessly, hurriedly’ and modern a prueba de pinchazos ‘puncture-proof’. Cervantes' aforementioned a garrotazos was not subordinated to any verb; M. Alemán favored daránte codazos y rempujones (Guzmán de Alfarache 3.274, ed. Gili Gaya). Cf. W. Beinhauer, Beiträge zu einer spanischen Metaphorik, RF 55.16, 190 (1941).

²³ In her trail-blazing study of Spanish doublets, C. Michaëlis [de Vasconcelos], Studien zur romanischen Wortschöpfung §15 (Leipzig, 1876), correctly opposed aguazar to aguachar, but complicated matters by dragging in dubious formations (lapazar [?] beside lapachar 'marshy terrain'), juxtaposing words at present known to be unrelated (mozo 'young' ~ mocho 'blunt'), and failing to disentangle the true relationship between (learned) capaz 'capable' and (vernacular) capacho 'basket (lid)', (native) capuz 'cowl, hooded cloak' and (imported) capucho ‘cowl, hood’, plus the correlated verbs encapuzar and encapuchar. See notes 77, 81, 82, below.

²⁴ In chosen instances the doublets are near-synonymous, with -azo, it would seem, more heavily emphasizing the excessive size and -acho more neatly accentuating the coarseness or wretchedness (cf. the respective derivatives from hombre, mujer, and pueblo). In others, the semantic gap is enormous (e.g. hornazo ~ hornacho; also, contrast buey-azo with as-n-acho ‘rest-harrow, cammock’, a plant name synonymous with gat-una, -uña) and explicable only as a result of bifurcation at a very early stage. Both -azo ₂ and -az, a variant of -azo ₁, beside -acho are amenable to combinations with interfixes: coll. dicharacho ‘low, slangy expression’ (dicho 'saying, saw') and vivaracho 'sprightly, frisky' (vivo) echo lengu(ar)az ‘long-tongued, garrulous’ (lengua), montaraz ‘born or raised in the mountains, wild, untamed’ (monte), and testarazo ‘blow with the head’; on espaldarazo see n. 9, above. In the latest Academy Dictionary asnacho is flanked by the phytonyms arnacho and asnallo.

²⁵ This statement is based on careful perusal of medieval prose and poetry representing a balanced assortment of genres: glossaries, heroic poems, didactic treatises, historiographie accounts, city ordinances, etc.

²⁶ Hence the self-explanatory counterparts of the Spanish formations previously quoted: aguilhoada, agulhada, alabardada, alfinetada (from hybrid alfin-ete 'pin'), arcabuzada, basto-nada, bolada, canhonada (the last three immediately recognizable as Castilianisms on account of -l-, -n-), cordoada, dardada, flech- and frech-ada, fuzilada, garrochada, navalhada, picada, sapatada, tesourada, testada, trabucada, trancada, etc. D. Alonso, BRAE 33.381-3, contrasts Sp. latazo with Ptg. dial, latada (Viana do Castelo, Vila Real). Note furthermore bicada (bico 'beak'), bordoada (bordão < bordõ 'staff, stick'), cacetada (cacete 'club, cudgel'), chicotada (chicote 'whip'), espingardada (espingarda 'rifle'), facada (faca 'knife'), ferrada (ferro '[piece of] iron'), joelhada (joelho 'knee'), maçada (maço 'mallet'), pancada (panca 'wooden lever'), paulada (pau 'stick'; cf. cha-l-eira 'tea pot'), porretada (porrete 'club with one rounded end'); note the frequent contiguity of -et- (-ot-) and -ada (see n. 16, above). A few of these formations show a semantic and phraseological proliferation reminiscent of Sp. -azo ₂; thus, pancada, aside from broadening its scope to denote any kind of 'blow' or ‘knock’ and from referring to the ‘ticktock of the clock’, has become a common member of fixed combinations: pancada de dinheiro ‘[hitting of] jackpot’, pancada de água ‘sudden shower’, pancada do coração ‘throbbing of the heart’; de pancada ‘at once, suddenly’ (cf. Sp. de golpe), ter pancada ‘to be cracked’, às três pancadas ‘slovenly’.

²⁷ C. von Reinhardstoettner, Grammatik der portugiesischen Sprache auf Grundlage des Lateinischen und der romanischen Sprachvergleichung 125 (Strassburg and London, 1878), adduced canhonaço ‘cannon shot’, mosquetaço ‘musket shot’, and pistolaço ‘pistol shot’ (function succinctly defined as ‘die Folge des Primitivums‘). C. S. R. Collin, Étude sur le développement de sens du suffixe -āta ... dans les langues romanes 161-2 (Lund, 1918), observed that Portuguese heavily favors -ada (cf. the offshoots of carabina, escopeta, fuzil), oscillates between the two possibilities in the cases of mosquete and pistola, and contrasts iterative canhonada with noniterative canhonaço. Neither scholar questioned the native status of -aço in this particular usage or observed its striking limitation to the discharge of firearms. This restriction, incidentally, could be most plausibly explained by the assumption that, during the crucial half-century of the union of the two countries under one crown, Portuguese soldiers familiarized themselves with these terms in Spanish training camps.

²⁸ Examples in points are bolaço, manotaço, and tromp-aço (-ázio), which Brazilian dictionaries assign to the extreme south (Rio Grande do Sul).

²⁹ V. García de Diego, Elementos de gramática histórica gallega 186 (Burgos [1909]), cites 18 such formations. Of these, aldabazo ‘sharp knock at the door’ (aldaba 'knocker, clapper'), porrazo 'blow, knock' (porra 'bludgeon'), trastazo 'whack, blow' (trasto, -e), and varazo ‘stroke with a rod’ (vara), also balazo, garrotazo, ramalazo, trompazo, and zapatazo are analyzable as straight Castilianisms, and agullazo, espadelazo, jringazo ‘injection with a clyster’ (= Sp. jeringazo), navallazo, taleigazo ‘blow with a sack’ (= Sp. talegazo) may count as superficially disguised invaders. Sp. anguilazo ‘stroke with a port-rope's end’ (from anguila 'eel') and latigazo 'lash' are likely to have served as models for Gal. lampreazo (Carré Alvarellos: 'golpe dado con un látigo, vara, vergajo o cualquier objeto que se ciñe al cuerpo'), from lamprea 'sea lamprey'. This leaves as only residue the most original formation, lostregazo ‘relámpago de marca mayor’ (Cuveiro Piñol), ‘relámpago grande; vareada, golpe rápido dado con una vara’ = vardascazo (Carré Alvarellos), beside lóstrego ‘relámpago’ and the corresponding verb lostregar. Lostregazo lends itself to the same multiple analysis as, on the one hand, Sp. golpazo and, on the other, Sp. arañazo and pinchazo (see n. 10, above). (Gal. aldabazo may be a misprint for aldrabazo, cf. n. 36, below; Carré Alvarellos lists only aldraba, aldrabada.)

³⁰ Collin, loc. cit., contrasted Sp. abanicazo ‘blow with a fan’ with Cat. ventallada, and Sp. naranjazo ‘blow with an orange’ with Cat. taronjada. This opposition supplements the suffixal divergencies provisionally assembled by Meyer-Lübke, Das Katalanische §79-§102 (Heidelberg, 1925). Cf. T. Forteza y Cortés, Gramática de la lengua catalana 108 ([Palma de Mallorca] 1915).

³¹ M. L. Wagner, Historische Wortbildungslehre des Sardischen 25 (Berne, 1952): Logud. Campid. [sali̯áttsu] < Sp. salivazo ‘expectoration’; Campid. niáttsu < Sp. nevazo.

³² A triangular relationship obtains in the case of hormazo ₁ ‘blow with a last or block’ : hormazo ₂ ‘pile of stones’ : hormaza ‘dry wall’ (term of masonry).

³³ F. Diez, Grammatik der romanischen Sprachen 2.255 (Bonn, 1838) : ‘Auch die Wirkung eines Werkzeugs bezeichnet es’; 2d ed., 2.293 (1859), with 9 instead of the original 3 examples; 3d ed., 2.314 (1871). Before Diez, C. F. Franceson, Grammatik der spanischen Sprache 74 (Berlin, 1822), discriminated between -azo, strictly augmentative, and -acho, both augmentative and derogatory. V. Salvá, Gramática de la lengua castellana ² 31 (Paris, 1835), observed a number of noteworthy formations, including bonazo, corpanchón, poblachón, ricacho, vivaracho, aguilucho, and bobonazo, but failed to organize his material in any really cogent way.

³⁴ P. F. Monlau (d. 1871), Rudimentos de etimología, included in his Diccionario etimológico (1856, 1883), see reprint (Buenos Aires, 1941), p. 116; the author separates -azo from -aza, a procedure marking an indirect progress, inasmuch as the critical masculine variant thus becomes semantically less burdened: ‘Aumentativa ... y otras veces connotativa de la idea de golpe dado con algún cuerpo o instrumento, o del efecto causado por este golpe’. Reinhardstoettner, Grammatik 125 (see n. 27, above). P. Förster, Spanische Sprachlehre 204-5 (Berlin, 1880), offers 14 examples of ‘Bezeichnung der Wirkung eines Instruments’.

³⁵ W. Meyer-Lübke, Grammatik der romanischen Sprachen §414 (Leipzig, 1894) : ‘Merkwürdig ist daneben nun die Bedeutung Schlag, Wurf, Hieb, Wunde, die mit einem Gegenstande beigebracht wird ... eine Bedeutung, die sich wohl wiederum aus der ursprünglich adjektivischen der Ähnlichkeit herleitet’.

³⁶ J. D. M. Ford, The Old Spanish sibilants (dissertation submitted in 1897), [Harvard] studies and notes in philology and literature 7.55 (1900). García de Diego, Elementos de gramática histórica gallega 186, with 18 rather carelessly selected examples (see n. 29, above): for aldabazo read aldrabazo; jringazo is the presto form of xiringazo (Carré Alvarellos) ; id., Elementos de gramática histórica castellana 196, a passage reproduced without alterations in Gramática histórica española 239 (Madrid, 1951). É. Bourciez, Éléments de linguistique romane §349b (Paris, 1910), claimed the absence of -azo ₂ from Portuguese, a challengeable statement (in later editions he illustrated Portuguese substitutes with cutilada ‘cut with a knife’ and escopetada 'gunshot'). F. Hanssen, Gramática histórica de la lengua castellana §310 (Halle, 1913), classed the augmentatives animalazo and dulzazo, also the collective melaza, with bastonazo and escopetazo as so many imitations of the Hispano-Latin legacy. In his treatise on Castilian derivation and composition (§47), based squarely on the current Academy Dictionary, J. Alemany Bolufer, BRAE 5.70-1 (1918), repeated the outworn definition of -azo ₂ (19 illustrations), observing the erratic relation of test-ar-azo to testera '(animal's) forehead, front face', likening semantically anomalous espaldarazo to pescozada, pointing out the affinity of -azo ₂ to -ada, and making mention of such suffix chains as -onazo and -otazo. A. Zauner, reticent in the original edition of his Altspanisches Elementarbuch 102 (Heidelberg, 1909), quoted in the 2d ed. (1921) cuchillaço on a par with pedaço (p. 90), implicitly staking out the first claim for -aço ₂ as a productive medieval suffix. J. Huber, Altportugiesisches Elementarbuch 243 (Heidelberg, 1933), chose two singularly infelicitous examples: on the level of Old Portuguese neither ameaça ‘threat’ nor pedaço ‘piece’ was any longer decomposable. J. Oliver Asín, Iniciación al estudio de la historia de la lengua española ³ 182 (Zaragoza, 1939), expressly connected the suffix of estac-, golp-, porr-, tranc-azo with -āceus (an equation repeated in 6th ed. [Madrid, 1941], p. 169). Among the authors of other textbooks, E. Gorra (1898) merely observed the fission of -āceu into -aço and -azo (Lingua e letteratura 64), Ford (1911) placed fogaza < *focācea alongside words ending in -eza, -iza, and -uza (Old Spanish readings, p. xxxvii), W. J. Entwistle (1938) referred incidentally to the spread of augmentative Ptg. -aço to Germanic toponyms (The Spanish language 281), R. Lapesa (1942) reported special uses of -azo and -ada without etymological commitments (Historia de la lengua española 261-2; cf. the elaboration in 3d ed. [1955], p. 340) ; and R. K. Spaulding (1943) completely dodged the issue (How Spanish grew). See further G. B. Pellegrini, Grammatica storica spagnola §171 (Bari [1950]), who recognizes a bridge leading from animalazo to bastonazo, and J. M. Piel, RF 67.375 (1956), who, surprisingly, regards verbal abstracts, arrived at from pejoratives, as the terminal point of the development.

³⁷ J. Wiggers, Grammatik der spanischen Sprache ² 207-8 (Leipzig, 1884). The interpretation of gatillazo is erroneous.

³⁸ In a way Wiggers inadvertently followed in the footsteps of Nebrixa who, in the suffix inventory of his Gramática de la lengua castellana (Salamanca, 1492), Bk. 3, Ch. 4, had aligned rosado ‘rose-colored’ and violado ‘violet’ (adj.) with cevadazo ‘pertaining to barley’, linaza, mostaza 'mustard (seed)'—capriciously derived from mosto 'must'—, and trigazo (see ed. I. González Llubera [Oxford, 1926], p. 83). What caused Wiggers' slip was probably the functional near-identity of -azo ₂ and -ada, the latter patently descended from -āta; between these rival suffixes he drew with razor-sharp precision a border line which Collin later refused to recognize : ‘Die Endung -azo fasst an dem Stammwort die einmalige Wirkung ins Auge, während die Endung -ada die einmalige Bewegung bezeichnet’.

³⁹ S. Puşcariu, Lateinisch ti̯ und ki̯ im Rumänischen, Italienischen und Sardischen, JIRSL 11.134 (1904).

⁴⁰ For criticism see E. Gamillscheg, Grundzüge der galloromanischen Wortbildung, in Gamillscheg and Spitzer, Beiträge zur romanischen Wortbildungslehre 38 (BAR 2.2; Geneva, 1921).

⁴¹ From modest beginnings in the original ed. of his Apuntaciones críticas sobre el lenguaje bogotano §601 (Bogotá, 1867-72)—and as late as the 4th ed., §712 (Chartres, 1885)—there developed a full-fledged discussion of the relation of -azo ₂ to -ada in the 6th ed., §865 (Paris, 1914); furthermore the 4th ed. (§211) contained a statement later apparently excised, on the chain -onazo (see notes 14 and 15, above) : baladronazo ‘unrestrained braggart’, barbonazo ‘solemn gray-beard’, bufonazo ‘master clown’, fanfarronazo ‘reckless blusterer’, glotonazo ‘insatiable glutton’, etc. But in the two versions of his article Disquisiciones sobre antigua ortografía y pronunciación castellanas (RH 2.17 [1895] and the undated revision published posthumously in Obras inéditas 393-4 [Bogotá, 1944]) Cuervo refrained scrupulously from injecting the issue of -azo ‘blow with ...‘ into the general debate centering about the transmission of -azo (i.e. -azo ₁) < -āceu and -izo < -īceu. His interest (1901) in the American, especially Argentine, use of -azo as an augmentative (cruelazo, guapazo, juertazo [from fuerte], patriotazo, superiorazo in Ascasubi, Aniceto el Gallo) failed to induce him to break his silence (see Disquisiciones sobre filología castellana 306 [ed. R. Torres Quintero; Bogotá, 1950]).

⁴² This, at least, is what follows from M. L. Wagner's digest (ZRPh. 64.354 [1944]) of a passage identified so inaccurately (or, if an antiquated edition is involved, so awkwardly) as to be out of reach.

⁴³ Thus, J. Leite de Vasconcelos, Estudos de filologia mirandesa 1.458 (Lisbon, 1900), commented on Mir. spad-ag-aço ‘slash, sword-stroke’, reminiscent of Ptg. espad-ag-ão ‘large sword’ (on -ag- as a variant of -eg-, -ig- see Lg. 25.139-81 [1949] and F. Lecoy's review in Rom. 72.140-2 [1951]); J. Moreira, Estudos da língua portuguesa 2.164 (Lisbon, 1913), produced from a play by A. Ribeiro Chiado tarraço ‘container used in milking’, the rare intermediate form between tarro 'id.' and (cant) tarraçada 'bowlful'; J. J. Nunes, Compêndio de gramática histórica portuguesa 387-8 (Lisbon, 1919), focused attention on lod-aç-al ‘bog, swamp’, fig. ‘dissolute life’ (cf. lod-ac-ento 'miry'). Occasionally -aço represents the second component of such a cluster of morphemes, as in coll. amig-alh-aço, -ão 'bosom friend', amig-alh-ote ‘doubtful friend’ (in the absence of *amigalho); see C. Michaëlis de Vasconcelos, Lições de filologia portuguesa 47-8, 71 (Lisbon [1946]; the lecture course was actually delivered in 1911-2).

⁴⁴ J. H. D. Allen Jr., Portuguese word-formation with suffixes 6-8, 72 (Lg. Diss. 33 [1941]) ; cf. the statements on -echo (32), -icho (50), -ocho (62-3), -ucho (66), my objections in Lg. 18.53, 55, 61 (1942), and J. M. Piel's review in Biblos 22.372-4 (1946), concerned with the segregation of OPtg. (a)negaça ‘bait’ = OSp. (a)ñegaza < Ar. naggāza (A. Steiger, Contribución a la fonética del hispano-árabe 217 [Madrid, 1932]), also with the nondiminutive function of -iço and the controversial reading of parvoechão in Gil Vicente. See also n. 47, below.

⁴⁵ O. Motta, O meu idioma ⁸ 61-2 (São Paulo, 1937), with a reference to Jud.-Ptg. (< OSp.) cansacio (Samuel Usque) and to the contrast between augmentative -aço and diminutive -acho, a point made independently by E. C. Pereira, Gramática expositiva: Curso superior ²² 167-8 (São Paulo, 1927). Michaëlis, Lições 42, included -aço among the deverbal suffixes on the strength of cansaço, setting it aside from denominai -aço (58, 74, 76)—as had done before her J. Cornu, Die portugiesische Sprache, in G. Gröber, Grundriss der romanischen Philologie 1.748 n. (Strassburg, 1888).

⁴⁶ After rejecting earlier conjectures by Meyer-Lübke ('man kann sich nicht vorstellen, dass die lateinische Funktion von arēnāceus 'sandartig' usw. in diesen Bildungen weiterleben kann, schon deshalb nicht, weil -azo-Bildungen, die diese Funktion hätten, im Spanischen sonst nicht vorkommen; solche Adjektive kommen nur in der gelehrten Form -áceo vor ...' [what about OSp. cevadazo, trigazo, etc.?] and, allegedly, by Cuervo ('aber wenn auch schliesslich -ada und -azo nahezu gleichbedeutend werden, so kann doch nicht -azo auf -ada beruhen, ohne dass für die -azo-Bildungen ein Vorbild bestanden hätte'), Wagner posits golpazo as the model word, without bothering to date its appearance (ZRPh. 64.353-4 [1944]). Corominas, DCELC 2.743b (1955), offers a casual comment on -azo vs. the paraphrase golpe de ... s.v. golpe, but fails to date golpazo, a silence which in his dictionary usually implies the absence of the given word from Nebrixa (1495).

⁴⁷ Review of Allen's dissertation in ZRPh. 67.458-9 (1951). The author is reproached for his failure to segregate the function here called -azo ₂ (i.e. -aço ₂ in the west).

⁴⁸ Such an attitude of studied restraint characterizes G. Weigand, Spanische Grammatik 96 (Halle, 1922). Boselli, La grammatica spagnola 162-5, etymologizes augmentative and derogatory -azo, but, counter to his general practice, withholds any parallel remarks on -azo ₂.

⁴⁹ Stimulated in all probability by G. Flechia's pioneer study (1874), Meyer-Lübke, Romanische Formenlehre §414, unhesitatingly subsumed Fr. lavasse ‘shower’ and Prov. lavaci under lauātiō ‘washing, (water for) bathing’. Nor did he rule out the possibility of such equations as It. popolazzo ‘mob’ < Low Lat. populātiō and It. minuzzo ‘small piece, bit’ < (dē-, dī-)minūtiō ‘lessening’, analyzable as starting points for autonomous -azzo and -uzzo (§420). But he expressed concern over the discrepant outcome of strictly erudite Praefātiō ‘(religious, legal) formula’ > prefazio and datiō ‘giving’ > dazio—a concern not necessarily shared by those who operate with different degrees of learnedness, discernible within a well-defined spectrum of transmission levels (see Lg. 31.284-6 [1955]). In the following years S. Pieri and C. Salvioni supplied neatly classified shreds of evidence from Italian dialects (see RPh. 2.72 [1948-9]). Subsequently Gamillscheg, Grundzüge 46-7, distilled two semantic clusters: (a) OFr. (es)lavace ‘shower, sudden flood’, surrounded by Verd. Chal. mouillace ‘light rain’, Middle Fr. plouvasse ‘shower’, Norm, foudrasse ‘violent gale’, and Poit. lacquasse ‘small puddle’ (lac × lavasse? Or, an alternative left unmentioned, aquātiō ‘fetching of water, watering place’ with the agglutinated definite article, plus lac?); (b) Fr. crevasse, Prov. crebasa ‘crev-asse, -ice’ < crepātiō ‘creaking, crackling’, flanked by its synonym OFr. fendace, Vosg. fodes, Morv. fendasse. To these two residual groups Gamillscheg added two isolated relics : (c) OFr. liace (liaz) ‘bundle, wisp of straw attached to a stick’, Fr. liasse ‘bundle of rags’, Prov. liassa < ligātiō ‘binding (up, together), tying’; (d) OFr. (learned) generace ‘race’ < generātiō ‘procreation’ (in the first three instances hypercharacterization of the feminine by means of -a may be projected into Proto-Gallo-Romance). Shortly afterward (ca. 1923) W. von Wartburg admitted the substitution of -āceu for -ātiō in the progeny of aquātiō (FEW 1.117), a pattern of suffix change so subtle as to have eluded G. Cohn's attention in his almost exhaustive study of such shifts at the preliterary level (1891). G. Rohlfs, Historische Grammatik der italienischen Sprache und ihrer Mundarten 3.255, 259 (Berne, 1954), grants that -ātiō underlies -azzo in It. andazzo ‘epidemic, fashion of a transient kind’, codazzo ‘train of attendants’, guazzo ‘pool, mess of water splashed about’, svolazzo ‘fluttering, excessive ornamentation’ [cf. svolazzare ‘to flutter, flit about’ and the Spanish parallels adduced in n. 18, above], Lucch. tremolazzo (Ven. tremasso) ‘shivering’, but selects -ūceus (as in pannūceus ‘ragged‘) as the immediate prototype of -uccio, -uzzo.

⁵⁰ A. Bello, Gramática de la lengua castellana §§206, 208-9, at first focused on -azo as a member of a set of augmentative suffixes (with some of which it combines), then enumerated the patterns of semantic coloration (ugliness, clumsiness, frivolity, etc.) that this basic function may assume. He thus failed to distinguish adequately between denotation and connotation. Monlau, loc. cit., defined (m.) -azo, except where it suggests a blow, as strictly augmentative, yet (f.) -aza as an ending ‘connotativa de la idea de aumento o abundancia, pero corno que acompañada de la idea accesoria de menosprecio o de inferior calidad’. J. Bouzet, Grammaire espagnole 34 (Paris, 1956), centering attention on anatomical terms, analyzed -azo as purely augmentative in ojazos ‘big eyes’ and bigotazos ‘thick mustache’, but as implying a pejorative nuance in manazas ‘big, coarse hands’, piernazas ‘long, ugly legs’, orejazas ‘disproportionately big ears’. The derogatory meaning outweighs the augmentative component in Portuguese, see Baldinger's comment (in severe criticism of Allen) on doutoraço, ministraço, poetaço, senhoraço and the parallel he draws between Ptg. -aço and Fr. -aille (ZRPh. 67.458-9); cf. n. 47, above.

⁵¹ Bourciez, loc. cit., described Arg. cruelazo ‘terribly cruel’ and patriotazo ‘ardent patriot’ (Buenos Aires) as superlatives (cf. n. 41, above). Hanssen, Gramática §310, produced Chil. calorazo ‘suffocating heat’ and tantazas from A. Orrego. Wagner, Lingua e dialetti dell'America spagnola 139 (Florence, 1949), likened Cast, (coll.) exitazo ‘big, splashy success’ [cf. G. Bombenerfolg] to Ur. comodazo ‘quite easygoing’ and gustazo ‘tremendous kick’ (in the dialogues of C. Reyles' novel El gaucho florido). Lapesa, Historia ³ 340, quotes Am.-Sp. amigazo, cansadazo, lindazo, paisanazo. In peninsular Spanish, augm.-derog. -azo is by no means extinct, especially on the colloquial level ; examples from Pérez Galdós, Misericordia: bribonaza (Ch. 19), gorrinaza (fig.) ‘big sow’ (Ch. 13), dióse el gustazo de ... 'indulged in the great pleasure of ...' (Ch. 22), humazo (Ch. 40), morazo 'big Moor' (Ch. 38), perrazo (Ch. 14). Other languages, less richly endowed with suffixes, may have recourse to a gamut of adjectives (cf. Dan. stor skælm ‘bribonazo’, tykt græs ‘herbaza’, tæt rφg ‘humazo’, langt ben ‘pernaza’, gement fruentimmer ‘mujeraza’, to quote Nyrop). Note also the German compound adjectives bärenstark, blutjung, mausetot, riesengross, steinreich, involving colorful comparisons (ur- produces a similar effect in conjunction with -alt, less so with -wüchsig; cf. also erzfaul, erzdumm, in which erz-, however, represents a 17th-century adaptation of Erz-< Gr.-Lat. archi-).

⁵² For full documentation and localization (unfortunately, in terms of political rather than dialectal frontiers) see C. E. Kany, American-Spanish syntax 51-2 (Chicago, 1945 and 1951). (The upper-half brackets here mark the projection into the standard of regionalisms phonetically transcribed; geographic labels do not invariably imply that the formations are restricted to the respective territorial units.)

⁵³ Apuntaciones críticas ⁶ §865 and §866; on Am.-Sp. campan-azo, -ada see also D. Devoto, Fil. 3.48-9 (1951). Cuervo argued that -ada symbolizes the effect of a blow, whereas -azo, commonly found in designations of shots, suggests its powerful momentum, in harmony with the preponderantly augmentative character of -azo—a line of reasoning not free from aprioristic assumptions. Gaznatazo (= -ada) ‘blow on the throttle’ (gaznate) and pestorejazo (= -ón) ‘blow on the (fleshy) back of the neck’ (pestorejo) were summoned as witnesses to the passive function of the suffix. Cuervo's merit, as usual, lies in the exceptional neatness of his analysis on the lexical level (cf. Col. cocotazo ‘slap on the back of the neck’: relation of coc- to cog-ote; chicotazo ‘crãck of a whip’: chicote classed as a nautical term; fuetazo 'id.': fuete recognized as a Gallicism; hybrid guacarazo 'sudden blow': from metathesized Quechua huaraca ‘sling’ [with hua- > gua- like hue- > güe-, cf. UCPL 11.91-105]; lamparazo ‘drink, swallow’, a derivation rooted in the metaphoric phrase atizar la lámpara 'to refill a wine glass'). Note also his comments on pepazo 'bullet shot', from Col. pepa = Cast, pepita ‘pip, small seed’ (§910), planazo ‘golpe dado con el plano de la espada’ (§505), suelazo ‘tumbling onto the floor’ (§254), notable on account of the preservation of the diphthong, hybrid totazo ‘bursting’, beside totear (§986; cf., on morphological grounds, cimbronazo ‘shivering’ beside cimbrearse, §505), and varejonazo ‘blow with a long, thick rod’ (§721), which shares the sequence -onazo with cimbronazo. Cuervo seems less penetrating in his statements referring to higher levels of abstraction, and his historical focus is not sufficiently sharp, at least not for the preclassical period: thus, he might have attributed Fray Luis de Granada's preference for lanzada ‘thrust with a lance’ to linguistic inertia (the word is well-attested in medieval texts) and, concomitantly, to the tendential recoil from the unattractive sequence -zazo (see n. 19, above).

⁵⁴ Collin adduces castañazo ‘blow with a chestnut’; the Academy recognizes only castañetazo ‘blow of a Castanet’ (from obs. castañ-eta [Cervantes] = mod. -uela, borrowed by numerous languages), ‘cracking of a chestnut in the fire’ (from castaña), fig. ‘cracking of the joints’. The second acceptation exemplifies the amalgamation and spread of -etazo (see n. 16, above). In Buenos Aires castaña(zo) is tantamount to ‘slap in the face’.

⁵⁵ Examples: candilazo ‘blow with an oil lamp’ ~ candilada ‘oil spilt from a lamp’, escobazo ‘blow with a broom’ ~ escobada ‘sweep, quick sweeping’, hachazo ‘blow, stroke with an axe’ ~ hachada ‘axe’ (which, if authentic, would match az-ada ‘hoe, spade’ < ascia 'carpenter's axe'), naranjazo 'blow with an orange' ~ naranjada ‘orangeade, orange juice’. By way of alternative device for blocking homonymy Collin presents -ar-ada as the ‘augmented’ by-form of -ada: llamarada ‘sudden blaze’ (from llama < flamma) vs. llamada ‘call’ (from llamar < clāmāre) and pulgarada ‘fillip, pinch’ vs. pulgada ‘inch’ (in the last analysis, both traceable to pollex -icis ‘thumb‘); cf. the comment here on espald-ar-azo in n. 9, above. For fuller treatment of this side issue, viewed from a different angle, see Estructurolismo e historia 2.107-99.

⁵⁶ As many as 140 instances of Catalan derivatives in -ada expressive of ‘blow’, H. R. Kahane informs me, have been tabulated, without exemplification, by O. Bernhard, La formación de nombres por sufijos en catalán (Zurich, ca. 1940). The same statistical survey lists 22 cases of augmentative and 1 of pejorative -as (= Sp. -azo ₁). T. Forteza y Cortés, Gramática catalana 108, illustrates the augmentative function with casaça, grandolaç (f. -aça), homenaç.

⁵⁷ Note in particular Boselli, La grammatica spagnola del XX secolo ³ 164-5, and Bouzet, Grammaire 35.

⁵⁸ Spanische Umgangssprache 149-50, 267 (Berlin and Bonn, 1930), affording excerpts from contemporary light comedy (P. Muñoz Seca, E. García Álvarez, etc.); and 2d ed. (Bonn, 1957), p. 174, where Beinhauer unfortunately stops short of recognizing the progressive coalescence of -etazo, despite references to bayonetazo ‘bayonet thrust’ (bayoneta, semantically removed from Bayona), palmetazo ‘blow with a ferule’ (palmeta, at a distance from palma), puñetazo (closer to puño than to puñete), silletazo ‘blow with a chair’ (more readily associated with silla than with sillela 'small chair') ; see notes 16, 26, 36, 54, above, and n. 61, below. Beinhauer's illustrations include botellazo 'blow with a bottle' (botella), punterazo ‘hit with a pointer’ (puntero), reglazo ‘slap with a ruler’ (regla), and varazo ‘stroke with a rod’ (vara); on the last-mentioned cf. n. 29, above.

⁵⁹ Among the less common formations adduced note cantazo ‘stone's throw’ (canto), cornazo ‘butt’ (cuerno)—with a monophthong, probably in imitation of cornada ‘thrust with horns’, cornal ‘strap for yoking oxen’, etc.—, correazo ‘blow with a strap’ (correa), garrochazo ‘prick with a goad stick’ (garrocha), coll. leñazo ‘cudgeling’ (cf. coll. leña ‘caning’, a semantic replica of pal-iza). Wagner extends the area which Cuervo assigned to planazo and suelazo and extracts numerous provincialisms from vocabularies. Among these And. jardaso (⌜haldazo, faldazo⌝) ‘tumbling so awkwardly as to get caught by the skirt’ owes its comic effect to the fact that a garment is pictured as a cause (indeed, the tool) of its wearer's severe predicament.

⁶⁰ See his article Calco e irradiazione sinonimica, BICC 4.3-17 (1948) ; included, in a slightly retouched form, in Saggi linguistici 11-22 (Florence, 1957).

⁶¹ Taking his cue from Cuervo's analysis of lamparazo (see n. 53, above), Wagner effectively assembles the jocular formations in which -azo suggests a ‘swallow, gulp’ (usually of a strong alcoholic beverage): Am.-Sp. cacharrazo (cacharro), Cub. (Ortiz) cañangazo (caña, with an unexplained interfix) and coñacazo (coñac), Hond. (Membreno) farolazo (farol), Cub. (Ortiz) ginebrazo, from ginebra ‘gin’, and guarapetazo, from guarapo ‘fermented juice of sugar cane’, of unknown exotic origin (DCELC 2.814b; note -etazo), And. (Alcalá Venceslada) jarupazo (jarupe = Cast, jarope 'syrup'), Am.-Sp. lampazo, Mex. mecatazo (mecate < Azt. mecatl 'rope'; patterned on the figurative use of latigazo). Cf. n. 4, above. On such colorful terms for ‘drunkenness’ as lamparillazo, loba(zo), tranca(zo), zorr-ón and -azo, see also H. Kröll, Designações portuguesas para ‘embriaguez‘ 22, 91, 133, 85 (Coimbra, 1955; reprinted from RPF), and M. Alvar's forthcoming review in RPh.

⁶² Thus, one misses the elucidation of the relation of Murc, zamarrazo ‘hard blow, reverse’ (Sevilla) to zamarr-e-ar ‘to shake’ (cf. n. 53, above, on Col. tot-azo ~ tot-e-ar and cimbr-on-azo ~ cimbr-e-ar) ; of Sp. costal-azo (-ada) ‘blows received in toppling over’ to cuesta ‘slope’; of Murc, santigü-eñ-azo ‘pummeling on the right and on the left side, as if crossing oneself’ (Sevilla) to santiguarse (blend with santiagueño, from Santiago?); of Sant. Murc, testerazo to testa (see notes 24 and 36, above). Wagner recognizes two legitimate analyses of encontronazo, of unequal plausibility, but only one of arañazo (from arañar). Counter to his assertion, coll. telefonazo is not peculiar to Ecuador alone.

⁶³ The direct derivation of batacazo from a verb as weakly rooted as batir ‘to beat’ (see n. 7, above; the vernacular counterpart is pegar) and of And. (a)gachapazo ‘falling down with one's whole body’ (Alcalá Venceslada), Sp. gachapazo ‘blow’ from agachar ‘to lower, (refl.) crouch’ is certainly misleading. A distinctly better example would have been so-frenazo.

⁶⁴ One sees scant justification for including here (naut.) bandazo ‘breaking of a wave on the side of a ship’ (banda), only distantly related to weather phenomena.

⁶⁵ There is no point in dwelling here on Diez's mistaken identifications, some of which he himself corrected in revising his grammar. His indisputable merit lies in having firmly grasped the underlying pattern of dichotomy.

⁶⁶ Among Diez's followers, Reinhardstoettner, Grammatik 125, assigned to -aço, -acho either augmentative or diminutive force (bichaço ‘big animal’ beside lebracho 'young hare'), without expressly labeling the suffixes as semantically differentiated doublets. In 1890 (Romanische Lautlehre 435) Meyer-Lübke, influenced by G. Baist's and A. Homing's statements (1880-3) on OSp. lech-uza, -ucha 'owl', traced mochacho to *mochaço. Four years later (Formenlehre 464) he reaffirmed the bifurcation of -āceu (mindful of the dual reflexes of furnāceu 'furnace') and set apart the closely aligned group of the names of cubs in Portuguese: borr-acho, lebr-acho, lob-acho, m(u)-acho (referring to 'pigeon', ‘hare’, ‘wolf’, and 'mule'), arguing that in much-acho the suffix was 'conceivably diminutive'. Alemany Bolufer and Menéndez Pidal adopted the core of Diez's and Meyer-Lübke's analysis; Bourciez remained noncommittal.

⁶⁷ On toponymic remnants of -ascu, -uscu see P. Skok, ZRPh. Beih. 2.39-47 (1906); on Occitanian reflexes of -usc(ul)a: A. Thomas, Étymologies françaises, Rom. 29.181 (1900); on -usc(ul)-a, -uscus in Gallo-Romance: id., Mélanges de philologie française ² 63, 109-10, 120-1, 128-9, 139-40 (Paris, 1927). J. Leite de Vasconcelos was apparently the first who inferred *-asculu from -acho; see RL 2.271-2 (1891) [indirect quotation] and—with explicit rebuttal of Meyer-Lübke (1894) and Gorra (1898)—Notas filológicas (9), RH 5.417-8 (1898), in regard to nev-asca ‘snow storm’, penh-asco ‘steep cliff’, verd-asco ‘tart wine’, and verd-asca ‘switch’ (on Gal. vardascazo see n. 29, above), also verr-asco (cf. JAOS 68.175-84 [1948]). Leite visualized an encroachment of -(c)ulu on -ascu and correspondingly associated Ptg. -echo, -icho, -ocho, -ucho with, on the one hand, pre-Lat. -esco, -isco, -usco and, on the other, Luso-Lat. -elho, -ilho, -olho, -ulho (on the last two see RPh. 1.212-3, 224-6 [1947-8]). A. R. Gonçalves Viana, Étymologies portugaises, RH 11.160-1 (1904), though aware of Leite's only safeguard being masculu > macho, accepted as indirect evidence amplu ‘spacious’ > ancho, inflāre ‘to puff up’ > inchar, sarculu ‘garden-hoe’ > sacho. He voiced greater assurance in his Apostilas 2.238 (Lisbon, 1906), apropos of Trasm. parr(ach)o ‘big drake’, likening it to adj. parracho ‘low, creeping’ (Supplement to C. de Figueiredo's dictionary); cf. the aforementioned names of cubs. A follower of Leite, J. J. Nunes, Compêndio de gramática histórica portuguesa ² 399 (Lisbon, 1930), drew no border line between autochthonous and imported formations (among the latter, note bonacho ‘good-natured’, penacho ‘panache’, vulgacho 'rabble') and lumped together suffixal and quasi-suffixal uses of -acho, -ucho, as in cap-acho, augm. -acha 'mat', synchronically a pseudo-derivative in the absence of a readily identifiable radical morpheme cap- (both objections apply to capucho 'capuchin'). His acceptable examples comprise fogacho 'little flame', riacho ‘brook’; ventrecha ‘slice of fish, cut off next to the head’; cornicho ‘small horn, antenna of an insect’, rabicho ‘crupper of a horse, pigtail’; gorducho ‘plump, chubby’ (from fogo, rio, ventre, corno, rabo, gordo, in this order). Allen, Portuguese word formation with suffixes 6, shares some weaknesses of Nunes' approach. G. Alessio, Problemi di etimologia romanza (II), RLiR 17.60 (1950), championed Leite's hypothesis with special reference to Aient, caspacho. On the ‘vocalic gamut’ holding together related suffixes see Michaëlis, Lições 74-6.

⁶⁸ This opinion, cautiously advanced by E. Gorra, Lingua e letteratura spagnuola delle origini 64 (Milan, 1898), rests on the assumed identity of Sp. cenacho ‘basket for fruits and greens’ and c(o)enāculum ‘dining room, garret, attic’. Yet these two words cannot be simply equated, even if one disregards the semantic hiatus : cenacho, channeled through Mozarabic, may have been influenced by capacho ‘hamper’ and, as Corominas further suggests, by Moz. qanâč < canistru or rather back-formation from dim. canistellum (DCELC 1.624-5, 756-7 [1954]). Only in an exiguous section of the Asturo-Leonese territory does one find [č] < [j] (including [j] < [λ]), e.g. enochar ‘enojar’, mechor ‘mejor’, mucher ‘mujer’ (see G. Sachs, ed. Libro de los caballos, p. xxiv and p. 54 [Madrid, 1936]).

⁶⁹ H. Keniston, Notas léxicas (4), RFH 4.69-70 (1942).

⁷⁰ Hanssen, Gramática histórica §310: -acho borrowed directly from Italian, -iche via a French detour.

⁷¹ L. Spitzer, Das Suffix -one im Romanischen, in Gamillscheg and Spitzer, Beiträge zur romanischen Wortbildungslehre 190 (BAR 2.2; Geneva, 1921): ‘... sp. -acho, das relativ jung ist und in geschichtlicher Zeit (aus ital. -acchio [sic]) sich konstituiert hat: sp. -acho ist teils augmentativ, teils diminutiv, ptg. -acho diminutiv, ptg. -echo, -icho verkleinernd ..., ebenso ... mall, -etxo verkleinernd; ital. -acchio ist verkleinernd’. Spitzer here seems ignorant of the use of -acho in Juan Ruiz and other medieval authors, solidly shielded from any Italian influence, and he speaks of -acchio [ak·jo] < -āculu where for the most part only -accio [ač·o] < -āceu may be seriously considered. On Mall, -etxo (viveretxo) and Cat. -atxo (amigatxo, ricatxo, viveratxo), possibly borrowed from Castilian, see Forteza y Cortés, Gramática de la lengua catalana 108. Bouzet, Grammaire espagnole 34, likewise attributes Sp. -acho to Italian, without bothering to draw any border line between picacho ‘summit‘—cf. the oronym Picacho de la Veleta in Sierra Nevada—and the triad poblacho, populacho, vulgacho. On the other hand, L. Spitzer, Hispanistische Wortmiszellen, NM 22.49 (1921), and Entgegnung, ibid. 23.33 (1922), was probably justified in advocating for Sp. malacho a native derivation from malo rather than descent from malehabitus (Meyer-Lübke) or local adaptation of It. malatto (J. Brüch, Etymologisches, ibid. 22.117).

⁷² Borrowing does not loom large in the pioneering works of Diez, Reinhardstoettner, and Förster. However, Meyer-Lübke, in 1894, traced to French or Italian mostacho and penacho (also capricho), if not hornacha, ricacho, verdacho. Gorra, in 1898, made no specific subtractions: ‘Dei sostantivi spagnuoli in -acho alcuni sono importazioni straniere’. In 1905 (Manual ²) Menéndez Pidal championed the native status of capacho and, less cogently, that of ricacho.

⁷³ This slightly agnostic conclusion G. Baist reached after sifting a sizable array of formations in -acho and congeneric suffixes (Gibt es ein Suffix -scl-?, ZRPh. 30.464-7 [1906]). See the following discussion.

⁷⁴ García de Diego, Elementos de gramática histórica española 188 (tentative formulation). But in Sp. hacha, Ptg. facha ‘ax’ the segment -acha, aside from being nonsuffixal, is nonindigenous, the source being Frk. hapja via OProv. apcha, OFr. hache. See E. Gamillscheg, Romania Germanica 1.174 (Berlin and Leipzig, 1934).

⁷⁵ Space is lacking here for a thorough discussion of this thorny side issue. The etyma conjectured include mocho ‘blunt, stub-pointed’ < mutilu (Covarrubias, Diez, Schuchardt in 1899, Meyer-Lübke, and many others), < *muttilu (Schuchardt in 1914), or derived from V.-Lat. (recorded) multilāre (Castro in 1914, 1916, and 1921); mozo, OSp. Ptg. moço ‘young’, itself of disputed parentage (Michaëlis in 1876, Schuchardt in 1881); moço in conjunction with Arag. mesacho ‘messenger-boy’ (Sainéan in 1905); Mūsculus ‘little mouse’ (A. Sperber in 1911; seconded, with certain reservations, by Spitzer in 1912, but later disavowed), and numerous other bases, incompletely surveyed by Corominas, DCELC 3.467-8 (1956).

⁷⁶ As a sample of gratuitous complication let me quote J. Brüch, Miscellanea linguistica Hugo Schuchardt 49-52 (BAR 2:3; Geneva, 1922). Affirming that masculu > maslo represents the older (vernacular) and masculu > macho a more recent stratum (popular adaptation of a learned form)—Baist and Meyer-Lübke in REW ¹ had previously classified macho as western—Brüch set out to explain Sp. hacha, Ptg. facha ‘torch’ < *fascla (an assumed blend of factjla ‘splinter used as a torch’ and fascis ‘fagot’: REW ¹ §3137) as follows: Ptg. acha, Cat.-Prov. ascla ‘torch’ (see Schuchardt, ZRPh. 33.82 [1909]) reflect a word which the cultured pronounced astula, through a compromise between assula ‘splinter, shaving chip’ and syncopated *astla (Meyer-Lübke preferred to speak of ‘Zerdehnung‘), cf. Sp. astilla, OProv. astela, while the illiterate used *ascla. Those striving after a pretentious pronunciation then adopted *fastula, modeled after astula, in preference to *fascla. This *fastula was finally absorbed into the lower stratum and reshaped as *fas-tla, -cla, which then yielded Ptg. acha. In this tortuous argument not a single statement is wholly unassailable (cf. Lg. 28.302-13 [1952]) ; it is quoted here extensively to exemplify the thick growth of loose lexical reasoning that the student of morphology, in his uneasy advance, is compelled to remove at every step.

⁷⁷ Among the cognate suffixes, Baist singled out Sp. pinocho ‘pine cone’, pinocha ‘pine needle’, and Ptg. medocha ‘small hayrick’ as the only authentic examples of indigenous -ocho, brushing off garl-, garr-ocha ‘goad stick’, car(r)-, quer-ocha ‘seminal substance of insects’, and birl-, mil-ocha ‘(paper) kite’ as obscure, Quevedo's pat-och-ada ‘blunder’ as only in part transparent, and mazocho (= It. mazzocchio ‘architectural ornament’, lit. 'tuft of hair'), galocha 'clog', and birl-, barr-ocho ‘two-seat light wagon without top’ as imported ('man könnte sich fragen, ob überhaupt ein Suffix vorliegt'). Conversely, Baist did not question the productivity of pejorative-diminutive -ucho, which he attributed squarely to the imitation of (a) It. -uccio: animalucho 'hideous animal', capucho ‘cowl, hood’, casucha ‘miserable hut’, etc.; (b) It. -occio: cartucho ‘cartridge’, noting its scarcity in Portuguese (pequerrucho 'tiny') and in Galician (platucha = Cast, platija 'flounder'; Pirucho = Cast. Pedrucho ‘Pete’ [add Perucho de Setién, name of a Basque youth ridiculed in an early play by Jaime de Huete]). The few traces of -echo known to him were by and large confined to the northwest. I plan to expatiate on this section of his note in a separate article.

⁷⁸ The functional shift -āceu (adj.) > Ptg. -acho (name of a young animal) here assumed is strikingly paralleled by the development of -īnus, cf. Sp. palomino ‘young stock dove’ (from paloma, -o), lit. ‘belonging to the dove’.

⁷⁹ On the concept of catalyst, applied to diachronic phonology but applicable, by extension, to morphology as well, see A. Martinet, Économie des changements phonétiques §3.35 (Berne, 1955), and my comment in RPh. 10.362 n. 18 (1956-7).

⁸⁰ See C. Fontecha (1940) on corbacho, despach-ar ('comer') and -o ('manera'), dia-cho (-nche), empach-ar(se) and -o, gabacho, gar- and guar-nacha, matachín, moharracho, and tacha; J. H. Terlingen (1943) on barrachel and vars., fachada (Fray José de Sigüenza; it preceded by a few decades facha [Calderón], which Corominas labels as 'italianismo náutico y militar'), matachín, regacho, and verdacho, also on puchas 'gachas'; F. Rodríguez Marín (1922) on regach(uel)o. Special studies or sources of information, only in part tapped in DCELC, are available on cacho ~ quejo (queixo) : J. Corominas, RFH 6.33-4 (1944); F. Lecoy, Rom. 68.1-6 (1944-5); Y. Malkiel, Lg. 21.142-83 (1945), and V. García de Diego, BRAE 37.13-24 (1957); des-, em-pachar: L. Feiler [Sachs], RFE 23.190-2 (1936); Arag. fardacho: H. Schuchardt, ZRPh. 41.701 (1922); J. Corominas (in criticism of W. D. Elcock), RFH 5.1-20 (1943) ; M. Sanchis Guarner, RFE 33.61 (1949) ; M. L. Wagner, ZRPh. 69.370 (1953); Gal. pen-acho ‘últimas hebras del copo de lana’, -echo ‘vellón’: D. Catalán Menéndez-Pidal, RFE 39.422-3 (1955), who regards penna as indigenous to the west. Mandracho is a Neapolitanism, according to Corominas (DCELC 3.223). The transformation of didblo into diacho was caused by superstition (cf. noramala > noramaza) ; the parallel dichotomy diabo(o) ~ diacho in Portuguese was not clearly grasped by F. A. Coelho, Formes divergentes portugaises, Rom. 2.289-90 (1873).

⁸¹ Corominas lists capacha merely as a by-form of capacho (around which a small family has crystallized: capach-ada, -ero; en-capach-ar, -adura), tracing the feminine variant to M. Alemán, Guzmán de Alfarache 2.107, 4.117 (ed. Gili Gaya). Add Quevedo's usage: estudiante de la capacha (Buscón 189, ed. Castro [1927]), los de la capacha ‘los de la orden de San Juan de Dios’ (Sueños 2.27, ed. Cejador); the Academy Dictionary comments: ‘... cuyos religiosos en un principio recogían en capachas la limosna que pedían para los pobres’. Though twin designations of containers are common in Romance (H. and R. Kahane, RPh. 2.135-75 [1948-9]), the margin of preference for the feminine in referring to the monks may be due to secondary association with capa.

⁸² Fontecha documents borrachesca, capacha, capacho (Cervantes; also s.v. berza), cobachuela, i.e. cov- (Ramón de la Cruz; cf. s.v. papelista), guachapear, hornacha; Rodríguez Marín, grosacho, lapachar (s.v. blandizal), largacho. For examples of pre-Cl. beçacho, bocacha (but It. Boccaccio), corp-acho (var. -ancho), pezacho, quartacho see J. E. Gillet's ed. of Torres Naharro's Propalladia 2.85 (acos beçachos torrados, / como terrón quebrajados), 1.262 (aquessa bocacha quiçá te besasse) and 2.463 (bocacha de oreja a oreja), 2.464 (corpacho mal empleado), 2.462 (que pezachos, pernejones), and 2.464 (quartachos de requesón); add corpacho from D. Sánchez de Badajoz, Recopilación 2.171. On cenacho see DCELC l.xxii-xxiii; on coiracho and Aient, sombracho (Morais Silva, Dicionârio ¹⁰ 10.360b: 'pequena armaçâo feita de ramos de árvores ... para abrigar dos raios do sol') : Leite de Vasconcelos, Lições ² 399 (Lisbon, 1926), and on coracha: my article Español antiguo euer y coraçón (in BH 60 [1958]); on covach(uel)a: Terlingen (s.v. espilocho), DCELC 1.858b (s.v. haba), P. Henríquez Ureña, El español en Santo Domingo 181 (Buenos Aires, 1940), who characterized it as not necessarily depreciative, in contrast to Castilian usage exemplified by B. Pérez Galdós, Misericordia, Ch. 13: No iban a su covacha más que a robarle. Baist (1906) was doubly wrong in tracing hilacha to It. filaccia and in declaring hilaracha patterned on moharracho. Llobo < lupu clearly underlies the configuration of llob-acho (La Cabrera), though rivalry between -m- and western -mb- < -mb- and the ensuing insecurity with regard to that medial cluster may have helped llobo 'wolf' overlay llombo < lumbu ‘loin’ (on this controversial point, raised by Casado Lobato's field data, contrast my statements in Lg. 25.439 [1949] and RPh. 9.52 [1955-6] with M. L. Wagner's in ZRPh. 69.352 [1953] and RF 68.443-50 [1956]). Through folk etymology or deliberate pun, perdigacho (Pardo Asso), not unlike perdigón, developed a secondary meaning: ‘perdicón, calavera, perdis’. Pezacho combines a monophthong in the stem syllable (cf. corpacho, poblacho, corpachón, bonachón) with an atypical use of the interfix -(e)z-, cf. UCPL 11.111-2 (1954). Pic-acho and peñasco are near-synonymous: cf. R. Menéndez Pidal, Toponimia prerrománica hispana 270 (Madrid, 1952) and H. Meier, RF 68.7 (1956). Pouc-ach-inho, with an obligatory sequence -achinho reminiscent of the occasional coalescence of Sp. -achón > -anchón: Leite de Vasconcelos, RL 2.237-72 (inaccessible to me). For an example of the continued use of Ptg. riacho (A. Ribeiro) see H. Sten, BF 13.123 (1952) ; on the Mozarabic channel of transmission see DCELC 1.858b (s.v. haba).

⁸³ Fontecha provides clues to the classical use of camar- and caram-anchón (Cervantes), hobachón (Espinel; see also Rodríguez Marín, s.v. bracitendido), montancha, and villanchón. The metathesis characterizing the first of these four words reappears in the case of carap-, capar-, garap-acho ‘shell’ (J. Hubschmid, Die Stämme *kar(r)- und *kurr-, to appear in RPh. 13:1).

⁸⁴ There is no dearth of border-line and dubious cases. The contrast between pataje and velacho is highlighted by H. and R. Kahane and A. Tietze, The lingua franca in the Levant 483, 702 (Urbana, 1958) ; Corominas sizes up the situation quite differently. Despite It. covaccio ‘lair’ (from cubáre), dial, coasso, Med. Lat. copacium (Battisti and Alessio, DEI; G. Serra, Lineamenti di una storia linguistica dell'Italia medievale 1.195 [Naples, 1954]), Sp. covach(uel)a may rate as native on the strength of meaning (‘tiendecilla en el sótano de una iglesia’, pointing in the direction of ca-, *co-ua) and form (old diminutive suffix). Cl.-Sp. most-acho (Góngora), like Fr. moustache, reflects Gr.-It. most-accio, a variant of -acchio; the selection of an uncommon, but more readily assimilable suffix variant is as noteworthy as are the fleeting emergence of mostazo (C. de las Casas)—abandoned on account of the disturbing proximity of mostaza ‘mustard’, from mosto < mustu—and the transformation, correctly posited by Corominas, of It. mostacciuolo into mostachón (Espinel) ‘macaroon’, in harmony with the spread of -achón. Ricacho (Guzmán de Alfarache) may, but need not, be classed as an Italianism (Baist regarded it as imported; Corominas eschews the issue and García de Diego, DEEH, omits the word). Baist and Corominas almost agree on the prototype of tablachina (It. tavol-accio vs. -accio, *-acc-ina; why not assume the shift to -ina within Spanish, under the pressure of tabla?). Corominas omits tristacho and valentacho, no doubt ephemeral formations, possibly nonces, which Baist termed adaptations or imitations of Italian models (tristaccio: 15th century, DEI). Among Italian topo- and oronyms conceivably denotative in Spanish observe Testacho ‘Monte Testaccio’ (mentioned in Cervantes' Novelas exemplares). All told, It. -accio developed much more vigorously than Hisp. -acho; a Hispanist would have a hard time matching the playfulness of the Italian saying: Garb-ino è morto, c'è rimasto Garb-accio.

⁸⁵ Corominas omits amigacho, refers the reader s.v. dicharacho to decir where he neglects to discuss the erratic derivative, disregards flaires gaspachos (Fontecha), lists but fails to date terminacho and velacho, offers no source other than the original Academy Dictionary for moracho and vulgacho.

⁸⁶ Much of the following sketchy discussion is based on F. T. Cooper, Word formation in the Roman sermo plebeius 113-4, 118-20 (New York, 1895), and—through Cooper—on C. von Paucker, with collateral reference to Ernout-Meillet. According to Cooper's count only 10 formations are ante-Augustan (gallīnāceus is Plautine and Ciceronian; Cicero further launched mustāceus and uīnāceus; also, let me add on my own, maluāceus); Pliny used (and in many instances introduced) 25, about two thirds of all post-Augustan forms. Cooper sums up his view of the situation thus: ‘The forms in -āceus are by far the most foreign to classic Latin’. The weak representation in the highly formalized epigraphic style of -āceus, -ācius, as compared to -ācius (this was the favored graph), appears from G. N. Olcott, Studies in word formation of the Latin inscriptions: substantives and adjectives 215-20 (New York, 1898) : only 8 pertinent words (2 of them new) had by then been discovered all over the Empire: gallīnācius, hederācetts, membrānāceus, mustāceum, rosāceus, sōlācius ‘sympathetic’, testāceus, and tjernācius ‘former home-born slave’ (from uerna), as against 43 derivatives in -īcius. Large-scale confusion of -ciu and -tiu, long known from the transmission of sōlācium ‘solace’ (Ford, Old Spanish sibilants 16), was confirmed by A. Carnoy, Le latin d'Espagne d'après les inscriptions 147-9 (Louvain, 1903); 2d ed. 141-3 (Brussels, 1906); cf. here testuātium and M.-Lat. creatius (?) mentioned in n. 10 to Hubschmid's forthcoming article on *kar(r)-, *kurr-. For armoniacia as a gloss of chrysomila ‘golden apple’ (CGL 3.185.31) read Armeniaca (B. Hasselrot, L'abricot, SN 13.227 n.2 [1940-1]).

⁸⁷ Vermin-āceus ‘wormy’, like its more frequent synonym in -ōsus, is a satellite of uermina (n. pl.), presupposing *uermen (cf. sanguen beside sanguis 'blood'), rather than of standard uermis -is (m.). Pulleiāceus 'black', which Suetonius attributed to Augustus, has traditionally been linked to pullus ‘dark-brown, dusky, blackish-gray’ (Ernout and others), a classification satisfactory on semantic grounds. Yet its erratic form prompts one to assume concomitant influence by pūleiātus ‘seasoned with pennyroyal’ (Vegetius), subst. -um ‘pennyroyal wine’ (Lampridius, Elagabali Vita), from pūleium (in glosses also -egium, -eius) ‘fleawort, pennyroyal’, the name of an aromatic plant used as a spice. The vitality of the latter word (and the replacement of ū by *ŭ, on the evidence of the reflexes—an indirect confirmation of the assumed associative process) is testified to by the near-consensus of the Romance languages and by OHG polaia.

⁸⁸ Such points of contact and transition do not rule out occasional cases of noncoincidence. Rūt-āceus ‘made from rue’ clashes with -ātus ‘flavored, garnished with rue’ (Pliny, Martial); mini-āceus ‘cinnabar-red’ stands apart from -ānus ‘painted with red lead’ (Cicero) ; argill-āceus means ‘clayey’, -ōsus suggests abundance of clay, much as oner-(āc)eus, -īcius ‘ashy, ash-colored’ differ from -ōsus ‘full of ashes’. All told, -āceus is closer on the semantic scale to purely relational -eus and -ālis (-āris) than to either -ātus, expressive of the actual presence of a feature, or the twin suffixes -ōsus, -ulentus, indicative of its copiousness, especially in the shape of sediments: capill-āceus ‘hairlike, made of hair’, -āris ‘of hair’ vs. -ātus ‘hairy’, -ōsus ‘very hairy’. As for -ārius, it rarely intervenes as a rival (despite relevant derivatives from aes, crēta, faex, and frūmentum) on account of its almost complete shift to the class of substantival names of agents (‘manufacturer, vendor of ...‘).

⁸⁹ Grundzüge der galloromanischen Wortbildung 37-47 (§20). The author's triple merit consists in having brought to light a thick slice of dialectal material, etymologized the formations with a reasonably slim margin of error, and effectively organized the smaller semantic nuclei. Not a few of Gamillscheg's statements are too sweeping, and some of his explanations lack conviction. W. von Wartburg's review, partially eulogistic (ZRPh. 43.110-1 [1923-4]), admits greater variety of cause-and-effect relationships between the extinction of the adjectival function of certain Latin suffixes and the serial crystallization of corresponding substantives ; in my own opinion the overlapping of the two related functions was distinctly wider. Wartburg is noncommittal with regard to Gamillscheg's unwarranted assumption that the collective meaning of -āceu may be due to association with n. pl. (> f. sing.) -a (actually, the unstressed final vowel has never exerted any semantic and only seldom a more than vestigial morphological influence in Romance derivation) ; he heavily stresses the inadequate treatment of the contrast between pejorative-augmentative suffixes in French, where they are congealed, and Provençal, where they are as freely movable as in Italian. Gamillscheg, let me add on my own, should also have guarded against the repeated interference of German speech patterns with his analysis (37-8 et passim) ; he might have improved his results by enumerating at the start those recorded Latin formations that have surely or probably survived in Romance, by separating, at least at the initial stage of his analysis, matters of form (e.g., prevalence of -ace, -ache over -az) from matters of meaning, by taking into account the evidence of Spanish, and by stratifying Italian data (to which he makes desultory references) along the chronological and the geographic axes. In particular, the transition from collective to pejorative-augmentative is not forcefully enough explained via Prov. estopa(s), erm(as), i.e. via formations in which the suffix is merely hypercharacterizing a word inherently collective (39).

⁹⁰ Regrettably, K. Baldinger, Kollektivsuffixe und Kollektivbegriff (Berlin, 1950), relied so heavily on the excellence of Gamillscheg's treatment as to have refrained from a detailed discussion of -āceu in Gallo-Romance (124). His monograph remains a valuable source on account of incidental comments, especially in the concluding portion (227-9: on dialectal counterparts of Fr. feuillage ‘foliage’; 232-4: on fim-us and *-ora ‘dung’ plus -āceu; 237: on filic- ‘fern’ plus -āria plus -āceu; 250 n. 2: on -esse in La Bresse, a genetically dubious suffix) and of important methodological insights gained through the study of functionally related formatives.

⁹¹ See L. Spitzer, Das Suffix -ōne im Romanischen, in Gamillscheg and Spitzer, Beiträge zur romanischen Wortbildungslehre 183-205, 210-1, and the various reactions to it.

⁹² Much of my information on Old French color names is based on A. G. Ott's unexcelled Zurich dissertation Études sur les couleurs en vieux français 14, 25, 74, 76, 86, 89, 90, 116, 120, 122, 140 (Paris, 1899). For Italian dialects I have drawn freely on G. Rohlfs's historical grammar (§1037) ; the author offers a less intricate explanation of the sense development than does Gamillscheg, deriving Type C directly from Type A: ‘Aus dem Gedanken der Ähnlichkeit ... konnte sich der Begriff des geringeren Wertes entwickeln’ (3.254). For Sardinian my almost exclusive source has been Wagner, Historische Wortbildungslehre des Bardischen §22 (Berne, 1952). By rescuing from oblivion Latin words overlooked in Romance research, Wagner, an expert practitioner, here deftly carries out operations whose theoretical significance he has deemed it necessary to belittle in another context (see RPh. 9.50-68).

⁹³ Meyer-Lübke failed to list separately the first three bases and hid away s.v. pāuō OFr. peonace and its Hispanic counterparts, which he neglected to identify as Italianisms. OFr. rosaz ‘pink’ may but need not perpetuate rosāceu. Is there a point in reconstructing provincial *plumbāceu from OFr. plombace ‘leaden’ (later plommas 'leaden object, part of the armor') with the aid of Sp. plomizo 'leaden, lead-colored'? (But Sp. plomazón 'gilding cushion' essentially involves plūmāceu, while plomazo ‘hailshot’, which cannot antedate modern techniques of bird-hunting, is composed of plomo and -azo ₂).

⁹⁴ Blancazo underlies mod. blanqu-ec-ino, arrived at through contamination with inchoative -ec-er and with the progressively common interfix -ec- (-ez-); see QIA 18.89-92 (1956).

⁹⁵ In all likelihood Sp. picaza dial, and obsol. pig-, peg-aza ‘magpie’ (a bird more commonly known as pega or urraca; cf. picaza chillona or manchada ‘shrike’, picaza marina 'flamingo'), picazo 'young magpie', fig. 'piebald (horse)' must be linked to this group. The woodpecker (pīcus) and the magpie (pīca) have both struck folk imagination by the motley color pattern of their plumages (G. Bunt- beside Grün-, Schwarz-specht; Spanish proverb: Dame pega sin mancha, darle he moza sin tacha; graphic description extracted from an encyclopedia article: 'Tiene la cola con reflejos rojo y verde, hombros y vientre blancos, parte inferior del dorso de un gris blanquecino, las grandes remeras con las barbas externas blancas'). Pigaço also served as a nickname: Abomalique (Yo ten Cate, ed. 1956, reads Abomelique) el Pigaço (Alfonso XI, 893c, speaking of Albohaçén's son); the Navarrese chronicler Eugui (ca. 1390) preferred the var. Picaço. See D. Catalán Menéndez-Pidal, Poema de Alfonso XI 132 (Madrid, 1953). Don Juan Manuel used the variant pigaça (Caballero y escludero, fol. 22r° I; see S. Gräfenberg's ed.), not pegaza, as Corominas assumes in the wake of P. de Gayangos. Note in particular the signaling of a glittering effect, closely paralleling the function of the suffix in Sardinian.

The much-discussed rivalry of the radicals pic-, pig-, peg- is inconsequential in this context; cf. García de Diego, Contribución §455 (Madrid, 1923). What matters is the wide use of the suffixed variants: picaça occurs in Ruiz, quatr. 504d, 881a, 920a, 924b; in two late-14th-century glossaries (E 1627, T 901); in the 1499 version of Calisto y Melibea (ed. Foulché-Delbosc, p. 108.22); and in Fray Pedro de Valderrama (Cejador y Frauca, VMC; add from Corominas the references to P. Guillen de Segovia, A. de Palencia, A. de Nebrixa, and H. Núñez, also to OArag. picaraza [a.d. 1379] ‘speck, pockmark’, with the differentiating interfix -ar-). Picazo ‘baldpie’ is clearly of later coinage: the witnesses include G. Morillo (ca. 1600) and L. Vélez de Guevara (Rodríguez Marín, Dos mil quinientas voces 289-90), also Lope de Vega (Fontecha). The semantic bridge to Ast. apigazarse ‘to be sleepy’ (B. Vigón), which Corominas recognizes as a cognate, remains to be built.

⁹⁶ This nonce contains an interfix -el- reminiscent of ros-el-enc ‘of gentle pink hue’, rouss-el-ot ‘slightly red-haired’, rouv-el-ain (-ant) ‘reddish’; cf. fauv-el-in (fig.) ‘deceitful’, (lit.) *‘fawn-colored’.

⁹⁷ Bloch-Wartburg, DÉLF ² 133a, graft Fr. cocasse ‘stupid, vain’ (1750; anthroponym: 1650) onto OFr. coquart 'id.', from coq. Characteristic of French is the reference to such qualities as 'soft, weak, lukewarm, effete, barren' or, by way of polarization, to toughness and tenacity. Gamillscheg, loc. cit., examines MFr. vainasse ‘effeminate’, Wall, [widriás] ‘soft’ beside [stokás] ‘tough’, from [stok] ‘stump’; B. Hasselrot, Le rôle de la formation diminutive en français, SMS 18.49 (1953), comments on tiédasse ‘rather tepid’ and the Provençalism savantas(se) ‘pedant’: 'Ces formations ont toujours brillé par leur rareté'. W. Wandruszka, Haltung und Gebärde der Romanen 19 (Tübingen, 1954; ZRPh., Beih. 96) quotes MFr. hommace 'mannish', calling to mind Sp. hombruna (RPh. 4.32 [1950-1]).

⁹⁸ As for the specific semantic overtone of these color names, reminiscent of Lat. -eus (russeus), Fr. -âtre, It. -astro, Rum. -astru, E. -ish, G. -lich ('indistinct, transitional, approximate shade'), it is tempting to attribute it to the leader words cinerāceus 'gray' and pāuōnāceus ‘purple’, which denoted colors inherently so placed within the total spectrum.

Our suffix is used seldom in zoonyms. Muraçico (Caballero y esondero, fol. 22 r° I; ed. Gräfenberg, RF 7.502 [1893]) probably referred to a long-legged bird similar to a crow (F. Huerta Tejadas, BRAE 34.451 [1954]). Lym-azo ‘snail’ (Semejança del mundo, B §20 [eds. Bull and Williams]) merely represents an elaboration on a prototype in -āx. Cagachín ‘tiny multicolored bird, small reddish gnat’ (Acad.) bears no resemblance to Ptg. -acho as found in names of young animals; without knowing the habits of bird and insect, any rapprochement with cag-azas (see n. 119, below), *-achas ‘droppings’ remains uncertain; but cf. agu-achales ~ -açales (Librode los caballos 7, ed. G. Sachs), Arag. morc-acho (= Sp. -ajo) ‘mixture of wheat and rye’ ~ Ast. morg-azu (DCELC 1.195b), Aient, lap-ach-êro ~ Trasm. chap-ac-eiro ‘quagmire’ (see n. 132, below). On Fr. carabasse (1552) < Prov. carabasso ‘crab’, from cārabu, see P. Barbier, Noms de poissons, RLaR 52.103-4 (1909) and postscript in RDR 2.155 (1910); A. Thomas, Rom. 38.620 n. (1909). The family name Gamazo may be an ancient zoonym (gamo ‘buck‘).

⁹⁹ Meyer-Lübke was correct in translating the word by ‘Haarsieb’, while Gamillscheg (Grundzüge 39) indefensibly spoke of ‘Seidenstoff’ (under the influence of sae-, sē-ta > Fr. soie, etc. 'silk'?). But Tusc. st- dramatically shows the blurring of the derivational pattern, independently pointed out by García de Diego, Elementos de gramática histórica castellana §199. Moreover, sēta only gradually sloughed off its primary meaning, witness OSp. seda (de cavallo) 'horse's hair' (Tratado de las enfermedades §12.5, ed. Maler).

¹⁰⁰ Cedaço was the spelling favored, in Nebrixa's wake, by A. de Corro (1560) and J. de la Cuesta (1584), see A. Alonso, De la pronunciación medieval a la moderna 1.200, 283, 299, and before him A. Horning (93); Ruiz used çedaçuelo (S 919b). But the Escoriai Glossary, failing to discriminate between ç and z, offers cedazo (548, 815, 1861), cedazero (623), and even zedazuelo (2042). Modern Aragonese shows the expected erosion of secondary -d-: (Titaguas) ceazo (Torres Fornés, Voces 186), UArag. [θjáo, θeáθo] (Kuhn, RLiR 11.232 [1935]). J. de Guzmán (1586) used cedacillo (BRAE 33.385 [1953]). The shift s- > ç- is due to anticipatory assimilation over a long distance, as in assechança > acechança, later acecho ‘ambush’ (HR 17.183-232 [1949]).

¹⁰¹ G. S. Colin, Étymologies maġribines, Hesp. 6.65-6 (1926); Wagner, Pro Domo, RF 64.420 (1952).

¹⁰² Add B. Vigón, Juegos infantiles 146. The shift an- > en-, under pressure of the common prefix, as in OSp. entenado, Ptg. enteado ‘stepson’ < antenātu; -ar- is an interfix characteristic of Ibero-Romance.

¹⁰³ Cap-azo, -aza: DH 2.661; W.-Ast. cabazo = C.-Ast. panera ‘granary built on pillars’ (Acevedo and Fernández) ; W.-Cat. cabassa: M. Sanchis Guarner, RFE 33.42 (1949) ; C.-Arg. capacho ‘leathern or wooden container for the transportation of gold’ (B. E. Vidal de Battini, El léxico de los buscadores de oro, Homenaje a F. Krüger 1.321 [Mendoza, 1952]). The older derivation of cap- from *capp- (Menéndez Pidal, Manual ² §53.4) is still championed by V. Bertoldi, La parola quale testimone della storia 75-6 (Naples, 1945), with a reference to It. cap(p)-agno. On peaça, Gal. peaza and apeazar 'to fasten, tie (an animal)', see further UCPL 11.19 (1954) and D. Catalán Menéndez-Pidal, RFE 39.441 (1955). The coexistence of Fornāx and furnāceus (beside fornix -cis 'brothel', surviving in hornecino ‘bastard’, in rhyme with mortecino ‘dying away, pale’ < morticīnu) paved the way for the frequent transition -az > -azo; see also Colin, Hesp. 6.73. For illustration see Vidal Mayor, ed. Tilander (4.12.21, 6.28.76: forna(t)z), Castro's medieval glossaries (E 522, 858, T 392: fornazo; E 981: ornazo), Ruiz (768b: fornacho); Rodríguez Marín, Dos mil quinientas voces castizas 178, 207 (fornaz(a), hornaz); Fontecha, Glosario 166, 194 (fornaza, hornazo); Gillet, ed. Propalladia 3.767 (hornazos pascuales). In formations originally traceable to -z, a consonant whose voicing in final position was unsteady (RPh. 3.32 [1949-50]), one is not surprised to discover wavering between ç and z : forna-z ~ -ça ~ -za, morda-ça ~ -za (General estoria 2.42a, 250b). A propos of -acho, I follow A. Alonso (RFH 8.11-76 [1946]) and Lapesa (Historia ³ 122) in taking the label Mozarabic in a broad sense. The alternation of -ice and -ace (subject to further hypercharacterization of gender) is further illustrated by Sp. romaza ‘dock, sorrel’ < (m., f.) rumice (Corominas, Indianoromanica, RFH 6.152 [1944]); note also the changed quality of the second vowel and the resultant stress shift.

¹⁰⁴ Gamillscheg, Grundzüge 39-44, operated with *cuneācea, *ferrācea, admissible if suggestive of an archaic layer, not so if connotative of pan-Romanic scope (most of these formations were regional). By way of alternative, he proposed for bann-, bonn-asse extraction, via suffix change, from dial, banasta < benna × can-istrum, (Hesychius) kánastron (REW ³ §1035). Corominas pleads cogently for the acceptance of a back-formation from canistellum (DCELC 1.624b-625a).

¹⁰⁵ DH 2.964b: ‘parte de la copela que se arranca con los pallones de piata’ (Gómez Ortega, Arte de ensayar)

¹⁰⁶ See DCELC 1.970b and, in greater detail, my study of OSp. euer ~ coraçón in BE 60.180-207, 327-63. The older classification of cuirasse as an Italianism (REW ³ §2233) is untenable. Old Spanish wavers here between -aça (from Ruiz 924c, via Calisto y Melibea [1499] 136, 143, ed. Foulché-Delbosc, to R. Percyvall [1591]; see Alonso, Pronunciación 1.99, 242) and -aza (St. Catherine's Epistles [tr. 1512], see RFE 34.130 [1950]).

¹⁰⁷ Fuero de Ledesma §102 and §254, with an oscillation between pretonic -aç- and -eç-familiar from blancazo ~ blanquecino ‘whitish’ (QIA 18.89-92 [1956]). The immediately underlying form is in all likelihood fabāceu.

¹⁰⁸ Judging from the gender, hardly a qualifier of traditionally masc. pānis ‘bread,‘ despite statements to the contrary by several scholars.

¹⁰⁹ Ernout-Meillet and Meyer-Lübke point to OHG fohanza and to noteworthy prongs, across language borders, into southeastern Europe.

¹¹⁰ For an attempt to reconstruct *mīluāceus from māluus ‘kite’ see G. Alessio, Problemi di etimologia romanza, RLiR 17.72-4 (1950).

¹¹¹ The itinerary proposed in REW ³ §5482 fails to carry conviction. Rather than analyze Sp. melaza as a Gallicism why not surmise that the replacement of -ll- by -l- may be due to the pressure of the coexistent primitive miel < mel?

¹¹² P. Aebischer, La literie et l'histoire du matelas, ZRPh. 66.304-12 (1950) : the word was replaced, in this order, by cocedra, almadraque, colchón; A. Steiger, Zur Sprache der Mozaraber, Festschrift Jud 647 n. 7 (RH 20 [1943]). For derivatives see further REW ³ §6611; to plumass-eau add -ier (1480), -erie (1505), which were directly related to plume after the extinction of short-lived plumas, evicted by plumet (but cf. Norm, plumas ‘tuft of feathers’, Anj. plumasse ‘peels of fruits’ [deverbal?]). On the whole, in Gaul plūmācium failed to strike roots as deep as it did elsewhere. It was in this particular sense that plūma was absorbed into Celtic and into Germanic (Gamillscheg, Romania Germanica 1.11).

¹¹³ C. Michaelis, Lições de filologia portuguesa 58; W. Giese, Miscelânea ... Leite de Vasconcelos 73 (Coimbra, 1934); cf. Ptg. W.-Ast. chu-, Minh. cha-maceira, Sp. (< Ptg.) chumacera ‘pillow-block of an axle’ = Sp. cojinete (Wagner, RPh. 6.326 [1952-3]), W.-Ast. enchumazarse ‘to get soaked’.

¹¹⁴ R. Lapesa, RFE 18.57 (1931) ; also lomaço (Ruiz, G 744d; but the passage may have been corrupted).

¹¹⁵ Cl.-Sp. pomazo (Pineda, Agricultura 2.2) lends no support, since it means ‘throw with an apple or any hard object’. The verb apomazar seems to reflect pómez < pū-, *pō-mex ‘spongy substance formed of hardened lava’ (DCELC 3.841a).

¹¹⁶ Through contamination with vigna < uīnea ‘vine’, cf. Sp. Viñaza, viñedo (an oblique reflex of uīnētum 'vineyard'), liñaza, a compromise between lino 'flax' < līnu and OSp. liña ‘fiber’ < līnea.

¹¹⁷ Among those missing note cinerāceus, coriācea, hordeāceus, miliāceus, mustā-ceum, oleāceum, pirācium, Pōmāceum.

¹¹⁸ Partially contaminated by aquātiō ‘fetching of water’ (RPh. 2.63-82 [1948-9]).

¹¹⁹ In the case of highly technical or obscene words, avoided in conventional literary and notarial style, the date of first appearance is quite inconclusive.

¹²⁰ In REW ³ §1599 (but not yet in Romanische Formenlehre §414) Meyer-Lübke subordinated Fr. canevas, Sp. cañamazo ‘coarse canvas’ (-aza 'id.'; 'tow, burlap'), Ptg. canhamaço to Nit. canavaccio, but granted equality of status to Prov. canabas 'canvas' and to Surs. [kanavíč] ‘hemp stalk’, which involves a dissimilatory shift to *-īceu. Conversely Bloch and Wartburg (1950) trace canevas to a cross between OPic. canevach and OFr. chenevas, the latter based on chaneve. Note coll. Ptg. calhamaço ‘thick ancient book’, (slang) ‘fat, ugly woman’.

¹²¹ Beside carniza (with vowel dissimilation), from which a lexical subfamily has branched off.

¹²² Meyer-Lübke, RFE 8.238, 244-5 (1921) ; A. Castro, Unos aranceles de aduana del siglo XIII, ibid. 9.266. Filaza may have influenced filacista (P 77), which Castro, Glosarios 221a, does not fully explain.

¹²³ Widely borrowed in the figurative sense: Sp. pall-, pay-aso, G. Pój-az, -as beside infrequent Bajazzo, Russ. poj-ác, -ás, flanked by the verb pojásničat' 'to act like a buffoon', etc.

¹²⁴ Ruiz, S 752d: -aça (mutilated in G); Berceo, Milagros, I 593d: -aza (mutilated in A).

¹²⁵ For illustrations see C. Fontecha, who also documents pre-Cl. empelazgado ‘plunged in a fight’, Cl.-Sp. -ar tr. 'to start (a feud)', refl, 'to feud' (Quevedo), the former involving contamination with -azgo < -āticu, a suffix originally suggestive of rights, privileges, dues, tolls, and the like. A sporadic shift in the opposite direction (Ast. mayorazu beside Sp. mayorazgo ‘primogeniture, heir to an estate’, fem, -azga 'heiress') is left unexplained in DCELC 3.302b. The related case of hartazgo beside hartazón, Cl.-Sp. also hartazga 'fill, bellyful' will be presented later.

¹²⁶ Milagros, AI 532d (in rhyme); Semejança del mundo, A §16-§17 (= B §19-§20); Ruiz 243b, 1404a; Miseria 66b; Escoriai Glossary 1033, 1995; on modern use see W. Beinhauer, RF 55.15 (1941). OSp. espinaza (San Laurencio 87b, in rhyme [see RPh. 3.69]; Ruiz 1091d) probably meant ‘fish bone’, hardly ‘spinach’, as was rashly conjectured by Cejador y Frauca (VMC 191a) and M. Alonso (Enciclopedia del idioma 2.1857b) ; Corominas ignores this by-form s.vv. espina and espinaca. Spīna ‘dorsal vertebra’ is classical, but the derivative is not: Macrobius referred to medulla spīnālis ‘marrow’. Spanish has rival-words for ‘backbone’, e.g. entrecuesta (Lazarillo 176, ed. Cejador; cf. C.-Ast. costazu ‘shoulder’ [Canellada], UArag. [kostaláθo] ‘slope’ [Kuhn, RLiR 11.232]). OArag. esquinazo ‘backbone’ (Los fueros de Aragón 396, ed. Tilander) and UArag. esquin-, osquil-azo (RLiR 11.233) represent departures conceivably caused by contamination with the same Germanic word that underlies It. schiena.

¹²⁷ Mariscal Pero García, Cancionero de Baena, No. 423.

¹²⁸ More than sixty years ago Meyer-Lübke here suspected inter-Romanic diffusion at work (Formenlehre §414); the exact channels remain to be identified. As terms of architecture and landscaping Fr. terrace and Sp. terraza may have been imported; but why suspect Ptg. terraço of borrowing from Gallo-Romance (K. Baidinger, ZRPh. 67.459 [1951])? The facts point in the direction of multifarious native development partially blurred by late medieval diffusion from a culturally privileged focus. Adjectivally the word is restricted to Old French: terrace ‘earth-colored’ (Type A); originally m. *-az, f. -ace?

¹²⁹ Traced to pre-Lat. *tšar- by Hubschmid, RF 65.296 (1953).

¹³⁰ A lexical tangle involving burra ‘shearings’ via Ptg. bôrra ‘dregs’ and Gr.-Lat. boreās 'north(wind)' (REW ³ §1219); see Hubschmid, RF 65.292-4 (1953). H. Meier's earlier attempt to reconstruct *moll(in)āceu (ibid. 63.333) was doomed to failure. The culinary term emborrazar ('poner albardilla al ave para asarla') is a 19th-century import from Cat. borrassa 'coarse woollen blanket' (Sanchis Guarner, RFE 33.42; DCELC 1.494a), but the var. aborrazar was used as early as E. de Villena, ca. 1420 (DH).

¹³¹ Hubschmid, RF 64.53-4 (1952); Menéndez Pidal, Toponimia prerrománica hispana 99-100 (Madrid, 1953), with references to Simonet and Meyer-Lübke; DCELC 3.19.

¹³² Two etymologies are on record: Sp. Ptg. lapa ‘swamp’ < Celt. *lappa, with onomatopoeic var. chap- (Hubschmid, RF 64.53-4 [1952]) vs. straight onomatopoeia (DCELC 3.32a, s.v. lapa ₃). Here and in the preceding case (except for lamach) -aç- occurs only as nonfinal suffix (interfix); similarly lodacero = lodoso ‘muddy’ (Autos 4.486, ed. L. Rouanet), cf. the saying salir de lodazales y entrar en cenagales (J. Morawski, RFE 14.119 n. [1927]), above-mentioned OLeon. ervaçal and fenaçal, C.-Ric. polvazal ‘dust cloud’ beside topon. Polvacera (Canary Islands) and Castillo de los Polvazares (J. G. C. Herculano de Carvalho, Coisas e palavras 56, 76, 170 [Coimbra, 1953]).

¹³³ Menéndez Pidal, Toponimia prerománica 273; Hubschmid, RPh. 6.190-8 (1952-3); DCELC 1.973a-975a. UArag. cotazo ‘half-baked’ (Kuhn), based on coctu, echoes focāceu.

¹³⁴ In Catalan-Provençal garg- combines with a chain of descendants of -āceu and -ōne (DCELC 2.681b).

¹³⁵ J. M. Piel, RPF 2.263 (1948); DCELC 3.1101b.

¹³⁶ Here -azo ₁ borders on -azo ₂ (UArag. peñazo, cantalazo, suggestive of 'stonethrow') and -asco (Sp. peñasco 'large rock', see n. 191, below); cf. And. peñ-asc-azo ‘hurling a rock’. On Ptg. dial. (Baixo-Minhoto) pedraço 'hail [of stones]' see M. T. de M. Lino Neto, RPF 2.161 (1948). Add OSp. eriazo beside erial 'unplowed land' (Calila e Dimna 153, 220, ed. C. G. Allen), also ería (Ruiz), poorly dated but correctly etymologized by Corominas (area, as against Allen's eremu and the Academy's aridu) ; cf. REW ³ §627.

¹³⁷ Fr. bagasse is a Hispanism. See DH 2.16; REW ³ §859; DCELC 1.364; Wagner, RPh. 6.323, noncommittal with regard to Trasm. cang-aço, -(anh)o, Ptg. engaço ‘(empty) grapestalk’.

¹³⁸ Cognate to des-boronar, -moronar ‘to crumble‘? See PMLA 63.785-802 (1948).

¹³⁹ Livro de cetraria (ed. G. Tilander), lines 129, 172-3, 753.

¹⁴⁰ Corominas credits linaza to Nebrixa, yet Alonso, Pronunciación 1.449 (n. 331), attributes linaça to Fray Luis de Granada. Gloss 417 of the Toledo Glossary (-aza) may be spurious (A. Castro). On Ptg. bater a linhaça see Carvalho, op. cit. 125.

¹⁴¹ DCELC 1.195b-196a, s.v. amurca ‘orujo de aceituna’.

¹⁴² Hubschmid's etymon (pre-IE *murro- ‘snout’, see RF 65.293 [1953]) seems far-fetched; cf. BH 57.84-128 (1955).

¹⁴³ The alternate spelling s.v. belortu. Corominas omits the word; is the etymon neruu?

¹⁴⁴ The adjectival stage, blurred in Gallo-Romance, has, we recall, been vestigially preserved elsewhere (Sp. cebadazo).

¹⁴⁵ DH 2.199b-200a. Corominas analyzes berra, but inexplicably neglects its satellites (berro, berrera, berraza).

¹⁴⁶ See Ruiz 175b; P.-López de Ayala, Aves de caça 45 (ed. P. de Gayangos); La Celestina 2.189; Alemán, Guzmán de Alfarache 3.11. For a rebuttal of the latest conjecture see RPh. 11.387 (1957-8).

¹⁴⁷ J. Mir y Noguera, Frases de los autores clásicos 322a, s.v. ennegrecer, quotes Fray Pedro de Vega, Declaración de los siete salmos penitenciales 4.8.2 (1606).

¹⁴⁸ Produced as a practical joke by pranksters (Alemán, Quevedo); see C. Fontecha, Glosario.

¹⁴⁹ Corominas toys with extracting pe(l)maço directly from pēgmátion (cf. D. Alonso, BRAE 33.385 [1953]), admitting derivation in early Romance as an alternative; F. Lecoy, Recherches sur le Libro de buen amor 99-100 (Paris, 1938), favors the latter solution. A few Greek diminutives are known to have survived independently into Romance (cf. Sp. cazo dipper, ladle, founder's scoop'; pedazo ‘piece‘); in this particular instance -átion was undoubtedly equated with -āceum [átju] on account of semantic affinity. At no time have pelmazo and (p)lumazo (and vars.) been in actual contact (A. Morel-Fatio, Recherches sur le... Libro de Alexandre, Rom. 4.47-8 [1875]).

¹⁵⁰ Enzina, Teatro 414; cf. Canellada and, earlier, García de Diego, Elementos de gramática histórica castellana 198-9.

¹⁵¹ Occasionally vidraça refers to a material other—though no less precious—than glass, e.g. fine silver; see the folk-song quoted by A. C. Pires de Lima in Homenaje a F. Krüger 1.191.

¹⁵² Again I owe many raw data to Gamillscheg, but frequently swerve from his interpretation. His mention of foliāceus and uīnāceus is welcome, but his reconstruction of *blatāceu, *burrāceu, *cotoneāceu, *fēnācea, [*]nigellāceu, *pisāceu, *terrācea is subject to caution (*picāceu is reasonably certain). Why not operate instead with attested fabāceus, hordeāceus, miliāceus? Gamillscheg voices doubt as to the relevancy of the characteristically western series Poit. boit-r-asse ‘limping woman’, fillasse ‘coarse (a.d. 1587), big girl’, mouffiasse ‘snot-nosed girl’ (cross-connection with moufl-ard ‘chubby’, -u 'puffed out'?), Anj. pétasse 'chatterbox' (péter), Ven. pét(r)asse ‘girl of ill repute’ (péter × dial, put-asse), Anj. pouffiasse ‘evil woman’, poupiasse ‘whore’, Poit. ponill-, pougn-asse ‘little girl’, pouffiasse ‘dirty girl’ (note the permutation of initial labials in mouff- ~ pouff-). The alternative would be to separate these mostly vituperative words from -ācea, grafting them onto OFr. boi-, bai-asse ‘maid-servant, girl’ (Norm, basse, Berrich. bogniasse), Prov. bagassa > It. bagascia ‘whore’ (REW ³ §861: *bac- or *bag-assa < ? Corominas, DCELC 1.365, is equally uncertain except about rejecting Carnoy's Late Greek etymon). Strong arguments in favor of the alternative are (a) the lack of a masculine counterpart (Gamillscheg errs in translating OProv. merdas by ‘Schmierfink’ and, once a captive of his own rendition, in drawing bold conclusions from the supposedly new function of -as-, merdas may very well have meant ‘dung-hill, muck-heap’, opprobriously applied to a person); (b) the striking regional limitation: if bagassa is of Oriental background, it would fit perfectly into a unique historical situation, namely the presence of captured Moorish slave-girls at the ducal court of Poitou.

¹⁵³ Of unknown ancestry; see REW ³ §3796. One is tempted to class with this group Fr. fatras ‘jumble, medley, balderdash’, also name of an ancient literary genre (P. Le Gentil, RPh. 12.26 [1958-9]), OFr. fastras [1327] preceded by fatrasie, from fa(r)sūra ‘stuffing full’; the -t- is a buffer consonant.

¹⁵⁴ Here the verb fer-rer, dial, -ter ‘to card (with an iron tool)‘ rather than the plantname provides the starting point: inferentially a late variant of the pattern.

¹⁵⁵ From patte beside poe ‘paw, leg’ (Illyrian word transmitted by Franks?).

¹⁵⁶ Almost same sequence of morphemes as in Poit. poum-er-asse.

¹⁵⁷ Through cross with the congener brouiller ‘to jumble up, shuffle’; OFr. broillas flanked by brouillasser (1624), brouillasse (1863); suffix change leads to MFr. brouillard, cf. foillas > feuillard, above.

¹⁵⁸ Not ‘vorhistorisch erweitert’ (Gamillscheg), but from dial. frime, OFr. frume (1289) < Frk. *hrīm, later overlaid by southern givre.

¹⁵⁹ See Rohlfs, Historische Grammatik der italienischen Sprache §1037 (3.253-4). Two points in Rohlfs's presentation seem unconvincing: the antiquated insistence on ellipsis prompting him to supply the substantive purportedly omitted (why class vinaccia as a former qualifier of massa?) and the inclusion in this group of OIt. mogliazzo ‘wedding ceremony’ which in fact pertains to the andazzo, codazzo, guazzo cluster reflecting, as Rohlfs admits (3.255), Lat. -ātiō.

¹⁶⁰ Wagner, who supplies examples and interpretation (Historische Wortbildungslehre des Sardischen §23), fails to discriminate between unquestionable Ibero-Romance reflexes of -āceu from such formations (Sp. nevazo, salivazo) as contain either -azo ₁ or -azo ₂.

¹⁶¹ H. B. Richardson upholds this derivation in his etymological vocabulary (1930), but Corominas concerns himself only with Ptg. caraça, Val. carassa ‘mask’ (the latter with prongs into Sardinian) and is silent about Ruiz (DCELC 1.662a).

¹⁶² UArag. caloraza ‘oppressive heat’ (from color [f.]), casaza ‘bad house’, cas-ol-azo ‘rickety house’ (cf. Sp. cas-ucho beside -ucha, -uca 'shack'), catenazo 'ugly, boorish person' (DCELC 1.974b, s.v. cueto), marguinazo ‘big landmark’ (omitted by DCELC 3.266b, which likewise fails to account for the phonology of marguín 'margin'), micazo, -a 'big crumb', narizaza ‘oversized nose’, tripaza ‘large intestine’, adj. guapaza ‘very pretty’ (f.)—all extracted from Kuhn; Sal. peñazo ‘wooden jaunt of a mill-wheel’ (Correas, also Lamano y Beneite; see Wagner, RPh. 6.332, s.v. pana, and my study of pecten, BF 14.76 [1953], notes 211-2); Dom. ladronazo ‘big-time thief’, onzaza ‘the largest of all ounces’ (in a fairy tale), adj. vistonaza ‘very shapely’ (f.), all recorded by Henríquez Ureña.

¹⁶³ Illustrations from DH: bragazas ‘wide breeches’ > ‘henpecked man’ (J. I. González del Castillo, D. Ramón de la Cruz, M. Bretón de los Herreros), bribonazo ‘great cheat, big rascal’ (González del Castillo, Bretón de los Herreros), bromazo ‘heavy joke’ (Bretón de los Herreros, M. Tamayo y Baus; the use of the masculine form of the suffix here suggests contamination by -azo ₂), calmazo ‘dead calm’ (Dicc. marít. esp., 1831; preference given to -azo over -aza on account of association with winds, through lexical polarization?), calzonazos ‘big breeches’ > ‘weakling’ (M. J. de Larra; add B. Pérez Galdós; see also W. Beinhauer, Beiträge zu einer spanischen Metaphorik, RF 55.50 [1941], and cf. G. Pantoffelheld). Add coll. talentazo ‘unusual measure of ability’ (Pérez Galdós, Doña Perfecta, Ch. 9). The beggar-women from Madrid's slum district depicted in the same novelist's Misericordia (Chs. 2-3) use such uncomplimentary expressions as escorpionaza, ricos sinvergonzazos, esos hormigonazos, unas grandísimas pendonazas. Gamillscheg's assumption of the obsolescence of augmentative -azo in Spanish thus turns out to be entirely unfounded (Zur Frage der Auswahl bei der suffixalen Ableitung, Behrens Festschrift 57 n. 1 [Jena and Leipzig, 1929]).

¹⁶⁴ Full documentation of the formations here adduced is found in the Diccionario histórico, Fontecha's Glosario, and Rodríguez Marín's Dos mil quinientas voces.

¹⁶⁵ A case apart is the occasional substitution of -aza for another final word segment under the pressure of taboo: noramaza ‘unluckily’ (Autos, Quevedo) beside more common noramala, i.e. enhoramala. Cf. Ptg. diacho ‘devil’ for diabo(o).

¹⁶⁶ Cf. the proverb A cada ollaza ('big pot') su coberteraza ('big lid').

¹⁶⁷ From gorra ‘cap’; A. Castro's definition (ed. El buscón, 1927): ‘estudiante pobre, de capa y gorra’.

¹⁶⁸ Menéndez Pidal, in J. Leite de Vasconcelos' wake, so explains topon. Barbacena (Toponimia prerrománica 118).

¹⁶⁹ Beinhauer, RF 55.43, contrasts bocaza ‘Schwatz-, Klatsch-base’ with bocón ‘Grossmaul, -sprecher’.

¹⁷⁰ The interfix -ar- sets it apart from lenguaz ‘foul-mouthed’ (B. de Villalba, Quevedo) < linguāce ‘loquaceous’ and from hypercharacterized coll. lenguaza (f.) ‘gossip, user of bad language’ (Misericordia, Ch. 2).

¹⁷¹ B. Gracián, El criticón 2.25 (ed. Romera Navarro), wittily opposes ojazos to ojetes ‘eyelets’ (also used as a term of sewing, mod. ojales).

¹⁷² A case apart is gruazo ‘male crane’ (Lope de Rueda) vs. gru(ll)a ‘[female] crane‘—one of the rare instances in Spanish of the derivation of a male's name from that of the corresponding female.

¹⁷³ Apparently not directly linked to *bouācea ‘cow dung’ (REW ³ §1244), chiefly inferred from Italian, Raeto-Romance, and Sardinian reflexes.

¹⁷⁴ Recopilación en metro 2.171-8, ed. V. Barrantes; in rhyme with bellacazo, bobazo, Davidazo, gigantazo, all semantically transparent. The DH 2.7b errs in attributing to bocazo the divergent meaning ‘violent contusion from a fall’ (= Sp. bacada, batacazo) in Viérades para reír / dar con ellos los bacazos (174) ; the only authentic trace in this passage of -azo ₂ is a chuzazos ‘with pike stabs’. From here only one step separates us from cornudazo ‘big (or old) cuckold’ (Alonso de Salaya, Farsa, ed. Gillet, PMLA 52.24.55 [1937]) and F. de Suva's rufianazo ‘big scoundrel’ (lit. ‘pimp‘); see Gillet, RPh. 10.150 (1956-7).

¹⁷⁵ On loiraça see above; penteaço, a term of carpentry, is based on pente ‘comb’; mulheraça suggests tallness and strength, coll. matronaça merely stoutness in a woman (D. Maçâs, Os animais na linguagem portuguesa 347 [Lisbon, 1951] quotes the latter from A. Ribeiro).

¹⁷⁶ Modern examples: gatàs ‘big cat’ (gat), peuàs ‘big foot’ (peu), donassa ‘big woman’ (dona), manassa ‘big hand’ (mà).

¹⁷⁷ In Italian dialects (Tuscany, Corsica, northeastern areas, Trentino) -azzo (-accio), at the terminal point of a radically different evolution, denotes the inhabitant of a town or a valley. Rohlfs cites and carefully localizes numerous examples: Tusc. Melazzo (Lucca), Cors. Aleriaccio, Basteligaccio, Bastiaccio, Pianaccio, Urbalagunaccio; Trent. Fiemazzo. No Hispanic equivalents come to mind, and the tone of facetiousness and abuse clearly dominant in the older Spanish texts seems to be altogether absent.

¹⁷⁸ See La subagrupación románica del catalán (1926), reprinted in Estudios lingüísticos: temas españoles 11-100 (Madrid, 1951).

¹⁷⁹ The inclusion of nonsuffixal formations is a precautionary measure, taken on the dual postulate that a few such formations may conceal a blurred suffix and that in diachronic analysis one may reckon with a tenuous connection between suffixal and nonsuffixal final word-segments (for illustration see UCPL 4: 3.117-9, 149-57 [1951]).

¹⁸⁰ Best presentations: Wartburg, ZRPh. 68.28, 39-40 (1952), revised in the miscellany Von Sprache und Mensch (1956), cf. id., ZRPh. 72.293 n. 2 (1956) : pitt- in Italy, along the Mediterranean coast west of the Rhône estuary, and in the hinterland between Garonne and Loire, pit- in the Iberian peninsula, in Gascony, and in far-off southeastern France; Corominas, DCELC 3.713 : copious inventory of Late Latin prototypes, from among which pitaccium is chosen as the closest model; and Hubschmid, ZRPh. 72.289-94 (1956) : refutation of C. H. Livingston's and G. Tilander's anti-Celticist views. The rival derivation from *ped(it)ātium (a reconstruction based, in the last analysis, on pede 'foot') must be written off as infelicitous (see G. Tilander, SN 27.31-42 [1955]); for other, equally conjectural, views favoring -āceum cf. A. Zauner, Altspanisches Elementarbuch §146 (also 2d ed. [1921] §140); J. Huber, Altportugiesisches Elementarbuch §435; G. B. Pellegrini, Grammatica storica spagnola §171; note M. Roques' indecision in Rom. 76.277 (1955). The spelling with -z- is found only in archaic documents, where it is inconclusive (P. Aebischer, Estudios de toponimia y lexicografía románica 108 [Barcelona, 1948]) or in late medieval texts tending to confuse the two affricates, e.g. the Glossaries (E 1961, 2336, 2495, 2580, 2734 vs. 1156; T 1617 vs. 1616, 1644; but P 100, 452: pedaçón). Pedaçar: Rimado, E 1114d; pedaçuelo: Tratado de las enfermedades §35.1 (ed. Maler); espedaçar: Poema de Alfonso XI 730b (des-: 1792d); apedazar: A. de Ercilla (Fontecha).

¹⁸¹ Corominas adopts this view—not shared by Hubschmid (see RPh. 13:1)—on the assumption that the incipient meaning of carrazo was ‘haz de troncos que se transportan en forma de almadía o de rastra’ (DCELC 1.705b-706a, 4.961). He might have strengthened his case by referring to Fr. échalas ‘vine-prop’ (12th century; deriv. échalasser, a.d. 1396) < échelle ‘ladder’ × *charas, inferrible from Berrich. charasson andGr.-Lat. carratium (Lex Longobardorum). This lexical type was blocked in part of the Gallo-Romance territory by the descendant of paxillus ‘small stake, plug, peg’; see the discussion by Bloch and Wartburg. Corominas traces OArag.. (al)car(r)açón ‘scales, iron weight’, found in inventories (B. Pottier, Étude ..., VR 10.125 [1948-9]), to Gr. kharistíōn via Ar. qarastûn. The topon. Carriazo (Santander) stands apart; cf. Diego de Carriazo, a native of Valladolid portrayed in La ilustre fregona. Italian counterpart: (ag)g-, aggh-iaccio ‘tiller’ < Byz. oiākion (Kahane and Tietze, Lingua franca §22).

¹⁸² Most of the words here examined occur in Ruiz, some also in the medieval glossaries (collaço, haça, laço beside enlaçar, maça and maço, plaça vs. plazo beside enplazar 'to summon to court'). Biz- in bizazas (unlike in biznieto 'great-grandson') shows assimilation of s to the following affricate; see Studies... J. Whatmough 165-71 (The Hague, 1957). On desembrazar 'to fling' (M. Alemán), ‘to take away’ (Lope de Rueda) see Fontecha, who further documents the occasional contact between abraçarse ‘to embrace’ and abrasarse 'to burn (amorously)'. Corominas defends collacteus against collātiō 'levy', the rival etymon proposed for collaço (DCELC 1.861b-862a). Some of the reasons for the withdrawal of faz, (h)az are stated RPh. 3.27-72 (1949-50); deshazado occurs in L. Rouanet's collection of farces, more advanced desazado in Cervantes' plays, of correspondingly later date. Laço and enlaçar enjoyed exceptional currency on the figurative level, with -z- making inroads in the east (Vidal mayor offers an ensemble of lazo ‘lasso’, enlazar, and laçaduero '[viga] que arranca de la lomera'); Gal. lazo 'ice, hoarfrost' mirrors glaciēs or represents the corresponding postverbal. Eastern menaza and menazar (Vidal mayor 3.73.3, 5.20.14, and 8.3.6 [intr.], 5.20.14, 8.3.2 [tr.]; (a)menazar, menaza also in MSS N P of the Rimado) contrasts with western amenaç-ar, -aua in Ruiz (GS 425b, 964b; other passages afford inconclusive graphies with σ, ṣ) ; the Poema de Alfonso XI (ed. ten Cate) shows the expected inconsistencies (f in 60a; σ in 1067c, 1798c, 1902c). The eastern meaning of plaça dominates in the Fueros de la Novenera and in Vidal mayor (1.59.51, 8.4.5). In Aragon, where placitum tended to yield plazto rather than -zdo (Fueros de la Novener a §297; Vidal mayor 1.56.12), one encounters occasionally plaço (Vidal mayor 1.11.10) beside vastly more common plazo.

¹⁸³ Not from Fr. chasser, as Corominas argues (DCELC 2.40a). On the two-pronged link to Minh. chaço ‘iron piece forming part of a cart’ and to Chil. chaza ‘metal tool used by coopers’ see Wagner, RPh. 6.326.

¹⁸⁴ On esquirazo and estraza consult Terlingen, Italianismos 244-5, 297-8 (conversely, Meyer-Lübke, REW ³ §2962, posited *distractiāre as the common base of stracciare and estrazar, ruling out territorial shifts). On paço ‘buffoon’ (< 'lunatic') in Torres Naharro see A. Alonso, Pronunciación 1.187; also Propalladia 3.102, with a side-reference to paça 'female clown' (J. de Timoneda). Kahane and Tietze, Lingua franca, trace to Italian Sp. bancaza ‘seat-locker’ (§61), Sp. contrabraza ‘preventer brace’ (§221), Ptg. filaça ‘rope yarn’ (§280). For related, but somewhat differently slanted cases see ibid. §§103, 359, 486.

¹⁸⁵ A. Steiger, Aufmarschstrassen des morgenländischen Sprachgutes, VR 10.19 (1948-9), traces albar(r)-aço, -açado to Ar. baršā', fem. of 'abraš 'marbled'. Corominas links Hisp.-Ar. al-baráṣ ‘spotted leprosy’ to Ptg. alvaraço ‘kind of grape’, thinking of white specks (from alvo < albu). However, at least part of his argument against Eguílaz y Yanguas, Steiger, and the Academy collapses under the semantic weight of the American-Spanish offshoot. On baço see AGI 39.166-87 (1954). Explaining bad-, bi-aza as mere variants of bizaza due to consonant dissimilation borders on oversimplification (DCELC 1.362b; one welcomes Corominas' reference to Cat. botafió, OCat. badafió). Calabaza, the center of a luxuriant wordfamily (-ada, -ate, -ero, -ino, E.-Ast. -ón), has been varyingly analyzed as foreign (Meyer-Lübke, Rom. Gramm. 2.460), presumably Oriental (REW ³ §1623), traceable to Ptg. cabaça, -o ‘frail, hamper’ (L. Sainéan, Sources 2.116), reflecting through anaptyxis (cucurbita) caluācea (G. Alessio, RLiR 17.55-6 [1950]), or nondescript pre-Latin (DCELC 1.579b-581b). On calaça see the fleeting remarks by Spitzer, ZRPh. 53.287 (1933), and Wagner, VR 10.326-7 (1948-9). Of carazo one can scarcely say more than that it rhymes with its local synonym lazo ‘frost’ (Hubschmid, RPh. 13:1). Corominas' analysis of enga(r)zar: Moz. *engaçrar < *incastrāre (DCELC 2.279a-280b) is unconvincing; the actual possibilities open to the unstable clusters -çr-, -zr- are discussed NRFH 6.209-76 (1952) and MLR 49.322-30 (1954). The coincidence between Ptg. eng-aço ‘grape-stalk’ and bag-, cang-aço is hardly fortuitous. One detects no link from UArag. esturazo to dial, as-, es-turar ‘to burn’ (DCELC 1.309b). Holgazán, rivaled by holg(ach)ón, flanked by holgazanería, and sharing the final segment of charlatán (RPh. 11.39-40) and trufán (Duelo 39a), is viewed by A. Steiger as an Arabism grafted onto folgar (Aufmarschstrassen ..., VR 10.22-4), whereas Corominas, rejecting Steiger's conjecture (unfortunately, without taking notice of J. Oliver Asín's pertinent comment on OSp. alaroza ~ Garoza, BRAE 30.409-10 [1950]), associates the segment -án with haragán ‘idler’ and the interfix -az- with (en)aguazar, deslavazar, despicarazar, apuñazar, though the parallelism is specious, since the formations underlying these four verbs (aguaza, lavazas, etc.) already contain -azo or -aza. Despite the existence of a copious literature on raça ‘lightly woven stripe in fabrics’, the possibility of decomposition: hilaza ~ (hila)raça (cf. hilacha ~ hilaracha) has not yet, to my knowledge, been pointed out; for another example of -ar- preceding the suffix cf. Ptg. topon. Cadraço < Cadarazo (a.d. 1142; see J. M. Piel, RPF 2.263 [1948], in criticism of J. da Silveira). After attaching recazo to OArag. caço, Corominas admits the uncertain origin of the latter (DCELC 3.1036b-1038b). To account for regaço, which Old Portuguese (Maria Egípcia, fol. 51r°, ed. Cornu) shared with Old Spanish (Glossaries, E 1233, T 1240; Mena[?], Tratado del amor, fol. 75v°, ed. Street), Corominas justifiably champions the traditional solution regaçar < *re-captiāre ‘to catch, pick up again, lift’; Rodríguez Marín's collection of lexical oddities includes regacear (F. de Silva); on arregaçar in Pero Guillén de Segovia see J. O. Tallgren, NM 18.141 (1917). The unexplained loss of -r- weakens Corominas' derivation of retaçar from *retraçar which, if correct, would call up inopportune memories of atarazar ‘to bite’ < *tractiāre (DCELC 1.314a-315b). Sciaberraça causes concern on account of the equally archaic by-form (faulty graphy?) in -aca (Menéndez Pidal, Toponimia prerrománica 12-3). On viaraça (Cancionero de Baena, fols. 34v° I, 161v° I) W. Schmid's glossary sheds scant light; Corominas, after some noncommittal flourishes, leaves the problem unsolved (DCELC 4.724b-725b).

¹⁸⁶ See the concluding pages (esp. 386) of D. Alonso's entertaining article Español lata, latazo, BRAE 33.354-88 (1953), with mention of such pairs—incongruent as to gender—as plancha ~ planchazo, (obsc.) coño ~ coñazo. In an effort to explain dar (la, una) lata, dar (el, un) latazo ‘to bore’, Alonso starts out from lata ‘log, batten, lath, stick’ rather than ‘tin (plate), tin can’ (as did most of his predecessors), then posits the sequence dar un latazo 'to give one a heavy blow (with a stick)', (fig.) 'to bore' > dar una lata 'id.'. He feels that, despite the discrepancy in gender, the average modern speaker takes latazo to be a plain augmentative (385). This subfamily (including such offshoots as adj. lat-ero, -oso 'boring', Am.-Sp. latear ‘to talk excessively’, Chil. -se ‘to feel bored‘) is noteworthy on account of its observably recent emergence (Valera, Unamuno) from low-class parlance.

¹⁸⁷ On the situation in Old Leonese, as reflected in Castilianized texts, see K. Pietsch's meticulous researches culminating in his inquiry into the dialectal idiosyncrasies of the Grail Fragments (1925). On the earliest oozing of Old Castilian sound features into Old Aragonese see my critique of M. Gorosch's ed. of the Fuero de Teruel, Lg. 31.261-91 (1955), esp. 270-5.

¹⁸⁸ See Leomarte, Sumas de historia troyana 160 (ed. Rey); var. vançinateo, with anticipation of the nasal. At the other extreme one finds the unalloyed Latinism incrusted, as it were, in a vernacular context (Ruiz 8c: salutacio; Barlán y Josaphat [ed. Moldenhauer], fol. 127 r°: fornicacio; Old Murcian documents, a.d. 1256, 1274: donacio beside a.d. 1274, 1356: donadío < dōnātīuu [J. García Soriano, Vocabulario del dialecto murciano 193]; F. López de Villalobos, Algunas obras 359 [Madrid, 1886]: la minoracio de la leche). The chief difficulty which the preservation of abstracts created on this level of learning was the dilemma between the tendential change of gender and the hypercharacterization of the ending; the latter solution was widely practiced in the vernacular layer, e.g. Mall, na-, ne-scia < nātiō ‘birth, breed’ (REW ³ §5848a), Logud. krentia ‘speck’ beside It. screzio ‘motley’ < discrētiō ‘separation, distinction’.

¹⁸⁹ On cillérveda and vars. see Meyer-Lübke, RFE 7.369-70 (1920), in criticism of Menéndez Pidal, ibid. 7.4-6; on kida, G. Serra, RPF 4.1-19 (1951) and Lineamenti 1.105-9.

¹⁹⁰ I disregard OArag. judizio, OCast. juýzio, OGal.-Ptg. juýzo < iūdiciu ‘judgment, trial’ on account of the almost consistent use of the voiced sibilant (the learned spelling -icio in some fueros [Teruel, La Novenera] is unlikely to reflect a variant pronunciation); and novicio ‘young, inexperienced’, quicio ‘pivot hole’ as semantically irrelevant. Conversely negocio ‘affair’ < negōtiu seems pertinent; a tenuous connection may exist between -icio and -icia < -itia, cf. OSp. laz-erio, -eria ‘misery’. I expect to present full documentation in a sequel to my article on cansar, canso, cansancio (NRFH 9.225-76 [1955]).

¹⁹¹ Friasco is listed by Hanssen, Gramática §359. Ventisca is erroneously analyzed as a postverbal by Alemany Bolufer (De la derivación ... §116); in reality, the verbs ventis-car, -quear ‘to snow and blow, to drift’ are as clearly subordinated to the substantive as are ventis-coso ‘snowy and stormy’, -quero ‘blizzard, snowdrift; snow-capped mountain’; cf. pedris-cal ‘stony ground’, -quero ‘hailstorm’. Pedrisco, in turn, is connected with peñasco ‘spire of rock, pinnacle, crag’ beside -cal ‘rocky country’, -coso ‘craggy’. The derivational pattern is obscured by vowel variation (cf. OSp. pinesc-, penisc-al in the Alexandre) and by the collateral pressure of substantivized adjectives like pedriza ‘stony spot’ (cf. paliza ‘[beating with a] wooden stick’, granizo '[grain-like] hail'). The group is closely connected with (Gr.-Lat.) borrasca 'storm, tempest, hazard, setback'. Sp. lloviznar, N.-Mex. ñoviznar may presuppose *llovizón (HR 12.57-65 [1944]). On the vestigial oscillation between -aza and -azga, -azo and -azgo see n. 125, above; P. M. deUrrea, Penitencia de amor 64, offers a late example of hallazo ‘(reward for a) find’.

¹⁹² In all likelihood this triad of suffixes exemplifies polygenesis, as does -(i)ego (UCPL 4:3.111-214 [1951]); the statements by Hanssen and Alemany Bolufer fail to do justice to this complexity of the situation. Aside from proper names of the type Velasco, attributed to Ligurian, one may set off (a) non-Latin phytonyms like carr-asca (var. -asco) ‘kermes oak’, flanked by the compound pincarr-asco (var. -asca) ‘Aleppo pine’; (b) an occasional Latin phytonym such as verb-, barb-asco ‘great mullein’ (= gordolobo) ; (c) designations of young animals, e.g. ternasco (from tierno < teneru) Ar. ‘new-born lamb’, Nav. ‘kid’, C.-Ast. poll-ascu ‘young rooster’, -asca ‘young hen’ (Rato y Hevia), with clearly discernible ties to -ast(r)o similarly used in dialects (cf. C.-Ast. pollascrón) ; (d) designations of piles of (dry, dead) leaves and small branches: chamar(asc)a, ch(av)asca, hojarasca—with a possible link to ‘snowdrift’ etc.—and, less frequently, of green twigs: C.-Ast. ramascu, Sp. verd-, vard-asca. The plausible relation to marisco ‘shell fish’, mariscar ‘to gather seafood’ poses yet other problems.

¹⁹³ Contrast Sp. aparicio in folklore (W. Beinhauer, Beiträge zu einer spanischen Metaphorik, RF 55.28 [1941]), heraldry and genealogy (cf. V. Dávila Jalón's monograph [1953] on the provinces of Vizcaya, Soria, and Burgos), and anthroponymy (Ángel Aparicio Laurencio, name of a contemporary Spanish jurist), with OPtg. apariço (J. de S. R. de Viterbo, Elucidárlo ² s.v., with J. P. Ribeiro's footnote) < appāritiō. Sp. desperdicio ‘waste’, pl. ‘by-products, rubbish’, from post-Cl. perditiō ‘ruin, perdition’ (glosses, Vulgate, Lactantius, Alcimus Avitus; vestigially contaminated with dispertītiō 'dispersal') has penetrated into Portuguese, but on the level of the corresponding verbs (Sp. desperdiciar: Ptg. desperdiçar 'to squander, fail to take advantage of') the initial contrast has been preserved. OCat. paricia 'epiphany' owes its ending to hypercharacterization (Meyer-Lübke proffers a less simple explanation in REW ³ §538); cf. OSp. ambicia < ambītiō in Ruiz.

¹⁹⁴ The Diccionario histórico illustrates carpazo ‘scratch’ with a passage from Gil Vicente (2.765a). Carpazo (patterned on some such near-synonym as arañazo?) is deverbal but, significantly, displays the transfer of newly activated -azo to the -ir verbs.

¹⁹⁵ Material has been extracted from the Diccionario histórico, Rodríguez Marín's wordlist, and Terlingen's monograph. Corominas' dictionary has proved disappointing, omitting as it does several formations and mostly failing to date the few that the author bothers to mention. The reason for the avoidance of azadon-azo, a, situation which has redounded to the benefit of -ada, is not euphony (note the distance between the two z's and Lope de Vega's uninhibited use of azotazo), but rather the lack, in typical references to the hoe, of any humorous overtones.

¹⁹⁶ Hence such heavily suffixed derivatives as encontronazo ‘sudden encounter’, in which -azo is analyzable either as a straight augmentative, i.e. -azo _1c, attached to encontrón so as to add the graphic connotation of ‘big bump, splash, bang’, or merely as an emphatic, clarifying restatement of the idea of blow, i.e. -azo ₂. There has been scarce, if any, affinity between -ada and -ón-, -ada may precede a diminutive suffix: ordinarily it attracts -ela in Portuguese (two of the rare Spanish counterparts involving -illa are pesadilla ‘nightmare’ and zancadilla 'sudden catch to trip one') and -ita in some varieties of modern Spanish (gleanings from a novel by Pérez Galdós include patadita 'friendly kick'; Flórez, Lengua española 229, quotes Col. echar una asomadita or llamadita ‘to drop in‘).

¹⁹⁷ To judge from D. Alonso's incidental remark in his study of latazo (386), speakers, at least in Spain, have already started seizing upon this latent possibility of clear-cut segregation.

Among the many side-issues not brought up in this paper are the relation between -az and -azo, -iz and -izo; the exact delimitation of OSp. -azo, -izo from -aço, -iço, on the basis of paleographie evidence and against the background of the checkered record of Ibero-Romance sibilants; and the history of learned -áceo, as in Sp. arenáceo, cretáceo, farináceo, herbáceo, which is intertwined with the history of scientific nomenclature.