Portal de Periódicos da CAPES

A Cluster of (Old) Portuguese Derivational Suffixes: -ece, -ice, -ez(a), Viewed in Relation to their Spanish Counterparts

1988; Liverpool University Press; Volume: 65; Issue: 1 Linguagem: Inglês

10.1080/1475382882000365001

1469-3550

Yakov Malkiel,

Linguistic Studies and Language Acquisition

Click to increase image sizeClick to decrease image sizeBSS Subject Index: PORTUGAL — PORTUGUESE LANGUAGE & ITS HISTORYSPAIN — LANGUAGES — SPANISH LANGUAGE & ITS HISTORY — MEDIEVAL PERIOD Notes 1. To cite only a couple of characteristic contributions, one of the first was: The Derivation of Hispanic ‘fealdad(e)’, ‘fieldad(e)’, and ‘frialdad(e)’, Univ. of California Publications in Linguistics, 1:5 (1945), 189–214; and one of the last (unfortunately, a victim of typographic carelessness): ‘Infinitive Endings, Conjugation Classes, Nominal Derivational Suffixes, and Vocalic Gamuts in Romance’, Acta Linguistica Hafniensia, XVII: 1 (1982), 15–48. 2. Some of these issues are discussed at length in a recent crop of papers: ‘Excessive Self-Assertion in Glottodiachrony: Portuguese sofrer’, to appear in Vol. 65 of Lingua; ‘A Spanish Conjugational Model Superimposed on Portuguese’, accepted for publication in Mediterranean Language Review; and ‘Divergent Developments of “Inchoatives” in Late Old Spanish and Old Portuguese: A Further Instance of Excessive Self-Assertion’, slated for appearance in the H. Blanc Memorial Volume. 3. See my two interconnected papers, ‘The Prelude to the Old French “Frequentative Action Nouns” in -ëiz’, in: Festschrift Kurt Baldinger zum 60. Geburtstag, eds. M. Höfler, H. Vernay, and L. Wolf (Tübingen: Max Niemeyer, 1979), 361-75; and: ‘The Old French Verbal Abstracts in -ëiz’ in the 1986 volume of ZRPh. 4. Useful, at most, as an alphabetic catalogue of older and modern Portuguese derivational suffixes has been Joseph H. D. Allen, Jr's University of Pennsylvania dissertation, Portuguese Word-Formation with Suffixes (Baltimore: ‘Language’ Diss. 33,1951); see in particular §§21,31,44,49,57–58, 62, 94–95. Allen's monograph must be judged in the light of the criticism (ranging from lukewarm to frankly negative) that was brought to bear on it by book reviewers and chroniclers of our disciplines, including K. Baldinger, ZRPh, LVII (1951), 458–61; M. Cohen, Bulletin de la Société de Linguistique de Paris, XLV (1949), 159; A. Kuhn, Romanische Philologie, 1: Die romanischen Sprachen (Bern: A. Francke, 1952), 427ff., with valuable cross-references but little originality in his verdicts; F. Lecoy, Ro, LXIX (1946–47), 120ff.; J. M. Piel, Biblos, XXII (1946), 372–74; L. Spitzer, RFH, III (1941), 393–95; R. S. Willis, Jr, RR, XXXIII (1942), 194–95; as well as the writer of these lines, in Language, XVIII (1942), 51–62, with relevant comments on -or (as in amargor), -eira (as in cegu-eira), also -ice and -agem (54), as well as -ez(a) and -idão (56–57). More important, on balance, than Piel's book review is, in retrospect, his slightly earlier article, ‘A formação dos substantivos abstractos em português’, Biblos, XVI (1940), 209–37, which, along with M. L. Wagner's almost simultaneous pieces, ‘Volkstümliche portugiesische Suffixe’, Volkstum und Kultur der Romanen, XIV (1941), 169–94, marks the peak of the pre-1950 phase of research and is yet to hold our attention. Representative of the latest stage of inquiries is Martha E. Schaffer's many-faceted Berkeley dissertation on -idão, of which two revised chapters have of late appeared in Romanische Philologie. The situation is rather different (yet not necessarily more advantageous) as regards studies conducted on Spanish suffixal derivation, where the starting point ordinarily chosen is a whole close-knit section in F. Hanssen's revised grammar, namely Gramática histórica de la lengua castellana (Halle: Niemeyer, 1913; repr., Buenos Aires, 1946), Chap. 14 (‘La formación nominal’); see, in particular, §268 for the older bibliographic sources as well as §§301, 303,314 (on sord-era), 318 (on such out-and-out Latinisms as pingüedo), 336 (on -ura). Hanssen's arrangement of the material assembled was infelicitous and redolent of nineteenth-century preferences. The weightiest and most detailed appraisal of his book came from the pen of A. Castro: RFE, I (1914), 97–103,181–84. The rival undertaking, José Alemany Bolufer's Tratado de la formación de palabras en la lengua castellana: la derivación y la composición (Madrid: Victoriano Suárez, 1920) is actually, despite its deceptive hard covers, no book at all but a reprint of a lengthy article of somewhat earlier vintage whose instalments had been spread over several numbers of the BRAE; see, especially, §§87 and 88 on -ez(a); also 52, 55, 92-94; 123, 156,168 (the array of the items is, again, singularly unhelpful). R. Menéndez Pidal's thinking about suffixes must be painstakingly pieced together from (a) appropriate sections of the successive editions of his Gramática histórica … (§§82-84, in particular); (b) nearly parallel sections of the Grammar preceding his monumental edition (1908–11, 1944–46) of the Cid epic (§§69–70, with an outline of the problem of the patronymic suffix -ez); (c) a few short statements wedged into the author's Orígenes del español (rev. edn., 329–30); (d) some long-scattered inquiries—to the exclusion of the last crop— into substratal suffixes eventually assembled in the volume: Toponimia prerrománica hispana (Madrid: Gredos, 1952); (e) individual articles spanning more than a half-century, from the piece on ‘Sufijos átonos’ contributed to the Adolfo Mussafia Memorial (1905) to some of the master's latest writings, including his definitive attack on the extraction of the patronymic suffixes: ‘Los sufijos españoles en -z, y especialmente los patronímicos’, BRAE, XLII (1962), 371–460. On the inadequate attention younger critics have paid to this piece see my comment in Section 8 of my article, ‘La última fase (1939–1969) de la labor lingüística de Ramón Menéndez Pidal’, to appear in Mexico City's Anuario de Letras. On balance, the single best treatment of -ez, -eza, and learnèd -icia in Early Old Spanish (including Old Navarro-Aragonese, where it significantly deviates from Old Castilian) may well be Chap. X, brief but well-documented and balanced in its judgements, of D. G. Pattison's monograph: Early Spanish Suffixes; A Functional Study of the Principal Nominal Suffixes of Spanish up to 1300 (Publ. of the Philological Soc., XXVII; Oxford: Basil Blackwell, 1975), pp. 142–50. For all the merits of this venture, it is, at best, indirectly useful in the present context, on account of the author's scant concern with Old Galician-Portuguese. 5. Most of the glosses here used have been borrowed—sometimes with slight adaptations—from the excellent Portuguese-English Dictionary by James L. Taylor (Stanford U.P., 1958). For glosses of modern Spanish entries I have chiefly fallen back on Edwin B. Williams, Spanish and English DictionarylDiccionario inglés y español (New York: Holt, 1955). 6. Ptg. -iedade corresponds to Sp. -iedad where the Latin model already displayed the sequence -ietāte; this applies, e.g., to Ptg. propriedade/Sp. propiedad (anciently propriedad). Secondary -ie- before -dad(e)—produced, e.g., through loss of intervocalic -d-—is more likely to be tolerated in Spanish than in Portuguese, with OSp. suziedad ∼ suziedumhre ‘dirt’ (mod. suciedad) vs. Ptg. sujidade (orig. çugidade); the point, of course, is that, at the level of the underlying adjectives, suzio contrasted with çujo. For the same reason, Ptg. limpeza ‘neatness, cleanliness, cleaning’ fails to agree with Sp. limpieza, and Ptg. antiguidade ‘ancient times, antiquity, seniority’ swerves from the path of Sp. antigüedad. (Scattered traces of the by-form antigü idade, if at all authentic, betoken Spanish influence.) Completely alien to Portuguese have been such elaborations on learnèd or vernacular -iedad as -iadad (OSp. piadad ‘piety, pity, mercy’; later replaced by piedad, but cf. piadoso ‘merciful, pitiful, devout’, apiadar ‘to move to pity’, refl. ‘to have pity on’) and esp. -aldad (as in fealdad ‘ugliness’, from feo, and in frialdad ‘coldness’, from frío; Portuguese ordinarily uses frieza instead). 7. For details of the coalescence of -én 1 < -ēdine (as in OPtg. amarguém ‘bitterness’, which calls to mind Class. dulcēdine ‘sweetness’) and -én 2 < -agine (as in Sp. sartén ‘pan’ < sartāgine and llantén < plantāgine ‘plantain, ribwort’) see my forthcoming paper, ‘Las fuentes del sufijo luso-hispánico -én: -AGINE y -EDINE’, to appear in Vol. 2 (1985) of Homenaje a Manuel Alvar (Madrid: Gredos). 8. For a full-fledged history of the rise and spread of the -āticu suffix from its Gallo-Romance focus we now have at our disposal the exhaustive monograph by Suzanne Fleischman; several shorter companion studies from her pen; plus a number of substantial critical reactions. On the superimposition of learnèd -agem (as in imagem) on imported -age—a process which entailed a striking change of gender—see my note in RPh, XXVIII:·I (l974), 20–27. Note that the derivatives in -agem tend to acquire collective rather than abstract overtones; thus, miudagem is tantamount to ‘odds and ends’, ‘leftover goods’, ‘small livestock’, ‘small fry’. No truly adequate account of the split of inherited -āgine, -īgine, -ūgine into a whole gamut of suffixes and suffixoids (-ãe, -ã, -anha, -agem, etc.) is available at this writing. 9. This is, clearly, not the place for presenting a circumstantial account of the congealment and subsequent partial recession of Sp. -eza; some—possibly still useful—ideas and data can be culled from a piece pertaining to my juvenilia (it was completed in 1939), but published after the war: ‘Probleme des spanischen Adjektivabstrak-tums’, Neuphilologische Mitteilungen (Helsinki), XLVI (1945), 172–91; XLVII (1946), 13–45. To my knowledge, no new derivative in -eza has been minted in Spanish after 1700 on either side of the Atlantic. In the Middle Ages, Sp. -eza was still productive, and its ranks swelled through absorption of an occasional trans-Pyrenean word, e.g., avinenteza ‘attractiveness, prettiness’. In the highly stylized literary language, new formations in -eza continued to be coined every once in a while, from newly acclimatized cultismos or in imitation of foreign models (witness Garcilaso's successful experiment with delicadeza ‘delicacy’, which echoed the delicatezza of his Italian model text). In the spoken language, -eza is apt to have come to a standstill at a distinctly earlier date, so that it should cause no shock to learn that the immediate source of Sp. menudencia ‘smallness, minuteness, meticulousness, trifle’ (for which Corominas could not cite any source anteceding Nebrixa)—an abnormally structured derivative—was Ptg. miudeza (from miudo < minūtu), orally transmitted and adjusted somewhere in the fifteenth century. Since Sp. -eza was approaching congealment and -encia appeared to flourish in all sorts of neologisms, it could have been pressed into service vicariously for the occasion. In the course of the Golden Age, trisyllabic adjectival bases, including paroxytones, exchanged -eza for -ez; thus, vernacular escas(s)eza ‘scarceness, stinginess’ and redondeza ‘roundness’ as well as hybrid robusteza ‘strength’ yielded ground to escasez, redondez, and robustez; then, too, -ez became mandatory in proparoxytonic -ido formations, such as avidez from ávido ‘greedy’. The reduction of delgadeza ‘thinness’ to delgadez carried with it the advantage of neater demarcation vis-à-vis delicadeza. Other lexical items in -eza were allowed to die out without any replacement, e.g., on condition that the underlying qualifier likewise went out of use, as happened with avoleza ‘craftiness, wickedness’, from ával < habile. In individual cases OSp. -eza lost out to rival suffixes (as when fineza became overshadowed by finura ‘nicety, courtesy, subtlety’), acquired a quaint flavour {proeza ‘prowess’), was dislodged by its learnèd counterpart (avareza > avaricia), or lost its abstract character (maleza ‘underbrush, thicket, weed’) ; naturaleza, orig. ‘indigenousness’, underwent a radical change of meaning, and fortaleza was split into two semantic nuclei, one abstract, the other concrete: (a) ‘(moral) strength’, (b) ‘fortress’. The last instance of keen rivalry between -eza and -ez was altivez(a) ‘haughtiness’, whereas alt-eza ‘highness (as a title)’ or ‘loftiness’ and alt-ura ‘height’ have each gone out of the other's trajectory. 10. Sceptics may hesitate to draw any sweeping genetic conclusion from the observable distribution of -ez and -eza in modern Portuguese, arguing that other suffixes pressed into service to produce abstracts show a similar degree of overlap. True, one comes across situations such as fresc-or, -ura / fresqu-ice ‘freshness, coolness’; cox-eira ∼ -eadura (the latter, strictly, a verbal abstract) ‘lameness’; mouqu-ice ∼ -idão ‘hardness of hearing’; rouqu-ice ∼ -idão ‘hoarseness’; vadi-ice ∼ -agem ‘vagrancy’, etc., testifying, above all, to the inroads of -ice and -idão, the two characteristic aggressors. But there is scarcely an instance of such recurrent fierce rivalry as with -ez and -eza. Competition with -idão, incidentally, has occasionally produced, by way of compromise, a by-form -idez entirely unknown to Spanish, e.g., amarel-idez ‘yellowness, paleness’ (beside more familiar amarel-idão); one also encounters crendidice beside more readily understandable crendice ‘credulity, gullibility’, produced with some help from the ger. crendo. In the former instance, pre-existence of the series estupidez, limpidez, palidez, etc. may easily have acted as an opening wedge. 11. Latinists and Indo-Europeanists are divided on the wisdom of declaring *-yē- or *-yā- the actual trail-blazer. The former conjecture was advocated, with a side-glance at the Baltic evidence, by F. Sommer, Handbuch der lateinischen Laut- und Formenlehre, 2nd and 3rd edns. (Heidelberg: C. Winter, 1914), §236; Id., Kritische Erläuterungen zur lateinischen Laut- und Formenlehre (Heidelberg: C. Winter, 1914), pp. 112ff.; the latter hypothesis was spearheaded by A. Ernout, Morphologie historique du latin, rev. 3rd edn. (Paris: Klincksieck, 1953), §89. Some additional data, but few truly enlightening insights, can be culled from Etienne-Abel Juret, Formation des noms et des verbes en latin et en grec (Paris: Les Belles Lettres, 1937), pp. 38ff. All investigators are agreed that the total contingent of -ēs words is composed of heterogeneous elements, and that the abstracts in -iēs represent its core, whereas lexical items such as diēs ‘day’, rēs ‘thing’, and spēs ‘hope’ involve accretions. While at a later stage, of particular relevance to Romanists, the wavering between the 1st and the 5th declension (e.g., the transition from glaciēs to * glacia, which underlies Fr. glace and It. ghiaccia) became crucial, students of the earlier phases focus on the hybrid paradigms—namely compromises between the 3rd and the 5th declension—of famēs ‘hunger’, fides ‘faith’, plēb(ē)s ‘common folk’, quiēs ‘quiet, rest, calm, lull’ alongside requiēs ‘relief, relaxation’, and t?bēs ‘melting, dwindling, rot’. Some items bear only deceptive resemblance to the patterns here studied, or are of dubious descent, and thus invite microscopic inspection. See, e.g., A. Ernout and À. Meillet, Dictionnaire étymologique de la langue latine, rev. 4th edn. (Paris: Klincksieck, 1959-60), pp. 84b-85a: caesar-iēs ‘long, abundant hair growth’ (secondary association with caedere, caesus?); p. 140b: contumēl-ia ‘affront, insult’; p. 191b: ēsur-iês (beside -ĭtiō in Martial) ‘hunger’, from êsus (< *ēssus) ‘food’; p. 316a: industr-ia ‘zeal, activity’, pl. ‘efforts’, and their relation to industrius, orig. -uus ‘active’; p. 375b: mēcer-ia, rarely -iēs ‘affliction’ (a meaning uniquely attested), ‘brick wall, enclosure’ (akin to mācerāre ‘to knead, soften’ [cf. Sp. maznar] rather than to mācer ‘lean’); and pp. 131b, 168ab, 347a, where a tricky case of homonymy is threshed out: (a) dēlicia 1 (Paulus ex Festo), -iquia (Vitruvius), far from functioning as an abstract, designates a sort of concave brick and belongs with (equally rare) coll-iciae, -iquiae (pl.), whereas the vastly more familiar dēliciae 2 ‘allurements’ flanks dēlicium ‘seduction, perversion’, both being akin to lax ‘ruse’, laciō, -ěre ‘to induce’, and—by far best known—lacessō, -ěre ‘to provoke’. Heavier bibliographic underpinning of the recalcitrant word histories is provided by J. B. Hofmann's revision of A. Walde's Lateinisches etymologisches Wörterbuch (3 vols.; Heidelberg: C. Winter, 1938-54), s. vv., with repeated references to J. B. Hofmann's updating (1928) of F. Stolz's Lateinische Grammatik: Laut- und Formenlehre. On the likelihood of a blend of săniēs with īnsānia ‘madness, frenzy’ (from adversative in- plus sānus ‘healthy’) see my earlier piece: ‘Primary, Secondary, and Tertiary Etymologies: The Three Lexical Kernels of Hispanic saña, ensañar, sañudo’, HR, XLII (1974), 1-32. 12. As a dependable critical guide to the older relevant literature on general Romance and Gallo-Romance, including the latter's influence on Middle English and Middle High German, one can use the excellent article by Margaret Sinclair Breslin, ‘The Old French Abstract Suffix -ise: Studies in its Rise, Internal Diffusion, External Spread, and Retrenchment’, RPh, XXII:4 (1969), 408-20. The author surveys and assesses pertinent writings spanning fully 130 years of animated scholarly debate; heeds in particular the opinions of (in alphabetical order) : G. Alessio, F. Diez, G. Cohn, A. Darmesteter, J. Dubois, E. Gamillscheg, H. Kurath, H. Lewicka, W. Meyer-Lübke, K. Nyrop, E. Öhmann, É. Pichon, and E. Schwan; ably summarizes and evaluates some of my own earlier thinking on OFr. -ece and OSp. -ez(a), as presented in: ‘Genetic Analysis of Word Formation’, in: Current Trends in Linguistics, ed. T. A. Sebeok, Vol. III: Theoretical Foundations (The Hague: Mouton, 1966), 305–64, at 341–45, and elsewhere. Particularly relevant are her remarks on the (semi)-learned transmission of -itia and on the sporadic penetration of the OFr. variant -ise into Old Spanish, as in riqu-iza (vv. 481,1269, 2663 of the Cid epic, ed. R. Menéndez Pidal). 13. See Max Leopold Wagner, Historische Wortbildungslehre des Sardischen (Rom. Helv., XXXIX; [Bern: A. Francke, 1952]), §95. Of heightened interest to those aware of the pivotal role played by the evidence of per-eza in Spanish and of pregu-iça in Portuguese—two formations bereft, we recall, of their respective primitives—is the similar fate that befell pigritia on Sardinian soil: Nuor. príθja, Logud. preítta, Campid. preíttsa. 14. Unfortunately, the available sources are little more than bare catalogues. Tomás Forteza y Cortés, Gramática de la lengua catalana (n.pl. [Palma de Mallorca?]: Escuela Tipográfica Provincial, 1915), 118, distinguishes among three sub-classes: (a) derivatives inherited from Latin, either (α) vernacular: dur-esa, just-esa, mal-esa, or (β) learnèd: just-icia, mal-icia; (b) derivatives newly-formed in Catalan: bax-esa, besti-esa, rar-esa; (c) a string of almost uniformly antiquated forms in -ea: bon-ea, dur-ea, fort-al-ea, llibr-ea, and mal-ea. In this simplified schema no place has been reserved either for Occitan or for Castilian pressure on Catalan. Equally, if not more, disappointing is the treatment of the set of problems at issue in the 1943 Zürich dissertation by Orion Bernhard, one of A. Steiger's intellectual protégés: La formación de nombres por sufijos en catalán. In his controversial book, Das Katalanische: seine Stellung zum Spanischen und Provenzalischen . . . (Heidelberg: C. Winter, 1925), §87, W. Meyer-Liibke, falling back on his own earlier analysis (Romanische Formenlehre, §480), singles out the absence from Catalan of any such contrast as beod-ez, calv-ez, grav-ez vs. agud-eza, brav-eza, certeza in Spanish, as another feature reminiscent of Occitan (Cat. -esa = OProv. -eza). Concentrating on the contemporary language, J. Huber, Katalanische Grammatik: Laut- und Formenlehre, Syntax, Wortbildung (Heidelberg: C. Winter, 1929), §364, cites self-explanatory agud-esa, arid-esa, bell-esa, brav-esa, cert-esa, destresa, dur-esa, embriagu-esa, nu-esa, pur-esa (from agut, àrid, bell, brau, cert, destre, dur, embriac, nu, and pur, respectively) without broaching the issue of any infiltration from Castilian. 15. G. Rohlfs, Historische Grammatik der italienischen Sprache und ihrer Mundarten (3 vols.; Bern: Francke, 1949-54), §1153. Unfortunately, the author falls short of making it clear just how the local representatives of -itie relate to those of -itia. In an attempt to disentangle the imbroglio, I have investigated part of the dialect area at issue with the help of Rohlfs’ own Dizionario dialettale delle tre Calabrie, Vol. III:2 (Parte italiano-calabra; Halle: Niemeyer, & Milano: Hoepli, 1939). The result was less than entirely satisfactory. A few scattered traces of -itia were indeed detected, e.g., prijizza beside scializza (s.vv. allegrezza [gioia]), a(r)tizza (s.v. altezza), capizza/capizzune (s.v. cavezza), munizza (s.v. spazzatura), apparently to the virtual exclusion of -izza. But the locally by far most characteristic suffix that could be peeled off from adjectival abstracts was -ía; cf. largasía (s.v. larghezza), lurdía (s.v. lordura), gappería (s.v. prodezza), druraría (s.v. pulitezza), mazzía, also papalía and targía (s.v. sonnolenza), cifía (s.v. stanchezza), feturía/fitunzía (s.v. sudiciume); in addition, completely different devices are being put to use, without any benefit to this whole cluster of suffixes, in an effort to render into racy dialect speech, say, Tusc, caligine, calvizia, dolciume, molestia, moltitudine, raucedine, and ruvidezza. In default of any exhaustive inquiry one can venture the guess that the direct progeny of -itiēs, once peculiar to the entire South (Sicily included), has been on the wane for centuries, barely surviving in a scattering of privileged spots. 16. The geographic pattern emerging from the distribution of vernacular outgrowths of -itiēs (Portugal, Spain, South Italy) is strikingly similar to that of the habitat of certain characteristically innovative lexical units (e.g., of afflāre, lit. ‘to sniff out’, through subsequent generalization: ‘to find’, which a half-century ago loomed large in the thinking of Matteo G. Bartoli). This areal configuration must be carefully distinguished from the aggregate of territories pieced together, on phonological evidence, by the advocates of the ‘Oscan substratum theory’ (including R. Menéndez Pidal), who operate with Upper Aragonese (better still, with dialects occupying both slopes of the Pyrenees), rather than with the Luso-Castilian block, viewed in conjunction with Southern Italian. 17. Edwin B. Williams, From Latin to Portuguese: Historical Phonology and Morphology of the Portuguese Language (Philadelphia: Univ. of Pennsylvania Press, 1938, 19622), rashly mistakes -itie > -ez for a characteristic native development, then uses it, in turn, as a norm in accepting OPtg. az < acie, faz < facie as a valid alternative to face §§46:2E, 89:2, and §98:3. A fine-tuned separate study may some day determine whether az and faz are ancient dialectal words or plain Castilianisms temporarily absorbed into the standard. Cf. cīvitāte ‘city, town’ > Ptg. cidade vs. OSp. cibdad. 18. Written ca. 1904; published in Revista Lusitana, XXIII (1920 [-22]), v-xii, 1–95. In this section of the present paper the use of Modern Portuguese glosses is deliberate, serving as it does to cast into bold relief the evolutionary trends of lexis and affixation. 19. On all three word-families there has accumulated a considerable etymological literature over the last eighty years. Some of it has been digested and supplemented in papers of my own; e.g., those that have touched upon the provenience and vicissitudes of sandeu/sandío and ledo. See RPh, XXIX:4 (1976), 435–500, at 479–81; and La Corónica, IX:2 (1980), 95–106. I doubt that Jerry R. Craddock's comments on the former item (in Festschrift für Johannes Hubschmid [Bern & München: Francke, 1982], 955–59) and those by Geoffrey Stagg on the latter (La Coránica, XI: 1 [1982], 46–48) have invalidated my statements. Craddock returned to the jousts in BHS, LX (1983), 1–14. 20. In this weighty review article (‘Zu Text und Interpretation des Cancioneiro da Ajuda’), which appeared in RF, XXII (1907 = Mélanges Camille Chabaneau), 339–85, I find references solely to doudice (363) and to ensandecer (372). Has there occurred a confusion with some other piece from Nobiling's pen? (From M. L. Wagner's remark in VRR, XIV [1941], 170, I gather that Nobiling's edition of Garcia de Guilhade may be involved.) 21. (Lisboa: A. M. Teixeira, 1922). The segment scrutinized runs actually from p. 59 to p. 99; but one must make generous allowance for several by no means concise editorial comments. 22. The passages included in this more leisurely examination had been culled from such late-medieval texts as Corte Imperial, Cronica dos Frades Menores, Boosco delleytoso solitario, Chronica breve do Archivo Nacional, at one end of the line, and, at the opposite end, Estoria de Vespesiano, Poesia religiosa (Livraria dos Monges de Alcobaça), and Tratado de cozinha (Bibl. di Napoli). 23. The absence of any derivatives in -ez (also of such in -ece and -ice) is the more noteworthy as most of the other relevant suffixal patterns are represented in this heterogeneous sample of texts, including -(or)ia: benfeitoria (-uria), melhoria, sabedo-ria; -idõe (the forerunner of -idão): fresqu-idõe and (hybrid) mult-idõe; -ura: curd-ura (in lieu of expected cord-ura), fremos-ura, verd-ura; plus, above all, -(i)dade in practically all its varieties: bondade, er-dade, ver-dade alongside von-tade (which lacks a primitive); the numerous items in -idade, ranging from vernacular to learnèd: cast-idade, clar-idade, etc. all the way to va-idade; plus the crass Latinism propriedade. 24. Cantigas d'escarnho e de mal dizer, dos cancioneiros medievais galego-portugueses, rev. 2nd edn. (Vigo: Editorial Galaxia, 1970), ‘Vocabulário galego-português’ (separate pagination, with subscript page numbers). 25. I can only hint at the rival devices available to that coterie of satirical and epigrammatic poets: (a) unstressed -ia, inherited from Latin: sobêrv-ia ‘haughtiness’ (from the primitive sobervo) = mod. soberba; (b) -don: semel-don ‘semelhança, amizade’; (c) -dade after consonant: bel-dade, mal-dade, and, in their wake, livel-dade ‘frouxidão de carácter’ (lit. ‘liberty’?) as well as podes-tade, again bereft of a primitive; a handsome contingent of items in -idade: car-idade (only in formulaic exclamations), castidade, escass-idade ‘mesquinhez, avareza’, salv-idade ‘inocência, pureza’, so-idade ‘pena’, torp-idade ‘ignorância, estupidez’; (d) a smaller molecule of formations in -ura: brav-ura ‘aspereza, rispidez, selvajaria’, cord-ura ‘juízo, prudência’, also self-explanatory louc-ura; plus (e) a long phalanx of items either in -ia (including -esia and -oria), or in expanded aria: agoir-ia (garbled by the copyist) ‘arte de agoiros’, bavequ-ia ‘tolice, parvoice’, benfeitor-ia, cleri-, crere-zia ‘profissão de clérigo’ (with velar/sibilant alternation), cont-ia ‘quantia’, cortes-ia ‘dignidade, decência cortesa’, fol-ia ‘loucura’, jograr-ia ‘profissão de jogral’, maestr-ia ‘conhecimento’, maloat-ia ‘doença’, melhor-ia ‘superioridade’, peior-ia ‘má situação’, preites-ia ‘ajuste, acordo’, put-aria, sabedor-ia ‘conhecimento’, taful-aria ‘jogo, tavolagem’, valentia ‘esforço, coragem’, val-ia ‘valor’, vilan-ia ‘desonra, injúria’. 26. Cantigas d'amigo dos trovadores galego-portugueses, 3 vols. (Coimbra: Imprensa da Universidade). 27. The remaining adjectival abstracts do not rise above the level of one's expectations: sobérv-ia ‘soberba’; maestria, (cantar de), and sabedor-ia ‘manha’; car-idade ‘esmola’ and torp-idade ‘torpeza’; soi-dadelsui-dade. One is left wondering whether some speakers tended to associate vertude (cf. Fr. vertu) with the family of verdade, before this isolate almost reverted, through Latinization, to its pristine form (virtude). 28. Alfonso el Sabio, Cantigas de Santa María, 2 vols. (Madrid: Real Academia Española, 1889). 29. For the two latest pronouncements on this topic, both highly competent indeed, see Thomas J. Walsh, ‘Two Problems in Gallo-Romance Etymology, II: Affixation as a Clue to Etymology: the Case of Old Prov. faduc, Old Fr. (mal-, dur-) feu’, RPh, XXXV:1 (1981), 99–104, and Otto Gsell's succinct and imaginative elaboration thereupon, ibid., XXXVI:3 (1983), 391–93. 30. The relevant suffixes bracketed by their dependence on nuclear ? (namely -ūra, -tūte, -itūdine, and -ūmen) are all four marginally represented: by cordura alongside laidura, vertude, limpidõe, and queixume, in this order. There are some notable instances of -ia having been so pressed into service (arlot-ia ‘picardía’, bavequ-ia ‘bobería’, malau-, malou-tia ‘enfermidade’, malhoria ‘superioridade’), but -ia and, especially, -erial-aria perform many other functions as well, as is observable in baylia, crerizia, companhia beside compánía, diabria, felonia, messageria, and mongia, preitesáa, tafuraria. One encounters the usual subclasses of the progeny of parental -tāte: (a) magestade (in reference to an effigy of the Virgin); (b) Crys-, Chris-chayndade ‘Christianity’; (c) (par) caridade, saydade ‘health’, soy-/suy-dade ‘consuming nostalgia’; (d) me(y)adade ‘half. Occasionally, -ança or -ença intrudes upon the domain of adjectival abstracts, e.g., iguança ‘igualdad’, sabença ‘sabiduria’, viltança ‘vileza’, to some extent also femença, a blend of ‘fervour’ and ‘vehemence’. 31. As regards grãado, there is no assurance that it was actually preceded by a finite verb; it could have spun off from the noun grão ‘grain, kernel’, potentially paving the way for a verb. On such tricky situations see my two mutually complementary papers, ‘The amulatado Type in Spanish’, RR, 32 (1941), 278–95, and ‘A Lexicographic Mirage’ [Pseudo-Participles Unaccompanied by Finite V