STOP Early Journal Content on JSTOR, Free to Anyone in the World This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in the world by JSTOR. Known as the Early Journal Content, this set of works include research articles, news, letters, and other writings published in more than 200 of the oldest leading academic journals. The works date from the mid-seventeenth to the early twentieth centuries. We encourage people to read and share the Early Journal Content openly and to tell others that this resource exists. People may post this content online or redistribute in any way for non-commercial purposes. Read more about Early Journal Content at http://about.jstor.org/participate-jstor/individuals/early- journal-content . JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people discover, use, and build upon a wide range of content through a powerful research and teaching platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please contact support@jstor.org. THE LATIN CONFIXES -EDON-, -EDNO-, 'EATING' By Edwin W. Fat 1. The reproach of not being able to see the forest for the trees was perhaps first addressed to a scholar. But the opposite vice is also to be shunned, of not examining the trees for looking at the forest. In linguistics this fault seems to me not infrequently caused by collecting a large statistic for a suffix without a close analysis of the meaning of individual words exhibiting the suffix — to say nothing of the not infrequent treatment of a suffix with entire neglect of chronology. I propose to reverse this proceeding by the analysis of some of the more characteristic words in -do, -dinis, and I shall start with uredo, a passing glance at which set me to studying the group to which it belongs. uredo 2. Used by Cicero (N.D. 3. 86) of 'blast, blight,' by Pliny (N.H. 9. 147) of a ' burning itch.' I see here a compound, with prius ur- : writ 'burns,' posterius -eden- : edit 'eats' (cf. far-edo, of a [corn-like?] abscess). The meaning of -eden- may be illustrated by Eng. fret 'herpes, tetter,' Germ, fresser "name einer krebsartigen baum- krankheit," cf. also Eng. mange from Lat. manducata and scurf : OE. sceorfan 'rodere.' As to formation, the pair eden- : edit may be matched with Skr. rdjdn- 'auctoritas' : Lat. regit 'rules,' cf. also offendo 'offensio' : offendit. To be sure, we may be dealing with a root vowel originally long, cf. inedia, Lith. edmi 'edo,' and the schwa- vowel in d>fXrdBio<; 'raw-eating' ( : ad-ep-s 'fat,' with ep- : opimus 'fat' ?).* In Plautine absumedo, even though it be a merely momen- taneous invention of humorous intent, -eden- 'eating' is clear beyond doubt, cf. edo | comedo (gen. -onis) 'eater, devourer.' 1 With ad-eps also compare Skr. dpsas-, name of a part of the body which a chaste woman hides (RV, 1, 124, 7; 5, 80, 6), quasi 'abdomen.' In RV, 8, 45, 5, Indra's mother proudly likens his enemies afight with him to " a < cloud- > belly at war with a mountain" (girav dpso nd = ' in-montem abdomen velut ). Perhaps abdomen is also tautological, ab- : the posterius in ad-eps, + -domen \ -dumen : 8rm6s ' fat ' : the root daw- ' urere ' (so Prellwitz) . [Classical Philology VI, July, 1911] 315 316 Edwin W. Fay dulcedo 3. That dulcedo is a compound of -eden- 'taste' has, I doubt not, been long suspected. At any rate, though without remarking on the derivation, Lewis and Short have defined it as literally 'a sweet taste'; cf. the late forms acredo 'a sharp or pungent taste,' salsedo ' a salt taste.' The force of -eden- is also sensibly retained in putredo (Apuleius) ' festering,' but for a word so late to emerge analysis may well be waived. In mulcedo (Gellius) we see an instance of irradiation from dulcedo. gravedo 4. I am not immediately sensible of the force of -eden- 'eating' in gravedo 'catarrh,' but the names of diseases, passions, pains (cf. curae edaces) so commonly end in -eScov -v&mp in Greek that -edSn- may be said to have come to signify 'disease' (cf. OHG. maga- bizzado 'magen-krampf,' -bizzado : Eng. bites). I note the following from Brugmann's list (Grund. 2 II, 1, §363), viz.: rvKeBdov 'consump- tion,' aKrjX^on' 'cura,' avireSdiv 'putredo' (cf. aye8aipa 'krebs- artiges geschwur,' tautological), irprjScop 'inflammation' (i.e., 'uredo,' §2), iSnBwv 'greediness,' 1 neXrjScov 'cura' (cf. fiekeSmvr) and, for the sense, Vergil's dolor edit), axQw&d>v 'pain' (cf. dolor: dolat 'hacks'), aXjrjBdov 'pain' and the counterterm x^PV^v 'joy.' 2 5. With these Greek names of diseases, disease-like passions, pains before our eyes, gravedo becomes clearer, for gravedo belongs to the group of "coughs, colds and consumption," just as does rv/ceSmv, which contains in its prius the semantic note of phthisis, viz., emacia- tion, while -eS(ov tells how the disease feeds upon its victim. By 1 This reduplicated form is to be compared with iSr/Stis (pf. ptc), if we consider it solely as to formation. Incidentally it reveals how, in Sanskrit, a -van- stem became contaminated with the -wis- stem in the pf. ptc. suffix -vans- (cf. Fay, AJP, 31, 425). 2 Brugmann's remaining examples less obviously exhibit -eden-. They are TV^>eSiii> ' vapor ' (? cf. Aen. 5, 682 "lentusque carinas | est vapor") , K\ei)Sdp ' omen, fama ' ( ? from k\c[-/:o] -\- ftiSiSiv 'singing': dfr)Sd>v 'luscinia'), Tr(fipi)Sil)v 'wasp' (-riSilsv 'singing' if, as Brugmann thinks, the wasp was named from his humming; v 'bee' = flower-eating ?>), x e ^$tiiv 'swallow' (with -Swv by irradiation from &ri8 may mean 'cup-eater,' not a bad description of the octopus's suction-feet (cf. Germ, saugnapf). There remain ffxaSdv 'larva' (rvaSdiv 'tear, cramp' which I feel to be genuinely conglutinata, with "determinative" S +the suffix -ev-, cf. Hesychian K\a8ibi> ' bough,' blended perhaps from /cXctSos and k\ pride of birth, flf we read imperatival meminerint the sense is "let men show their consciousness of birth by dressing within their means."] 318 Edwin W. Fay c) unedo 'arbute shrub/ derived by Pliny (N.H. 15. 98) from unum(?) and edendi. I do not understand Pliny's derivation, and have none to offer. d) teredo, a gnawing worm, contains in the Greek original, Tep-qZatv, the root ter- 'to bore' -r-eden-. e) torpedo — if the original sense is 'numbness,' note the propor- tion gravidus : gravedo=torpidus : torpedo (see §4). But if the original sense was 'giving-numbness,' then the division torpe-do commends itself. /) hirundo ' swallow.' Here -do can hardly have come by irradia- tion from arjBmv. But we may take (h)iro(m) as imitative of the bird note (cf. §4, fn. on ^e\iSa>z/) -\-na x den- 'canens' (cf. on arundo). g) alcedo 'kingfisher'; either with suffix from *ficedo \ ficedula (v. Niedermann, ap. Walde; Martial, 13, 49), or with -edo from ar)8d>p. h) testudo ' turtle '=schal-wasser-thier, from testu ' shell' -\-uden- : vSpa 'wasser-thier.' i) hirudo 'leech' may also contain in -udo a word for 'wasser- thier,' for at least one variety was Xifivala (cf. the glosses). 1 Perhaps in Mr-, to the root gher-, we have the sense of 'grasping'; then Mrudo is from *herddo, with pretonic vowel weakening. j) unguedo (Apul.) can hardly be anything more than a pre- tentious expansion of unguen. k) capedo= patera quae capit edulia(l). I) capudo= patera quae capit uvida(7). m) intercapedo 'intervallum' (cf. capedo 'spatium inter parietes'), perhaps from -ca[pi]-pedo = spatium quod capit pedes ( = fundamenta) muri. cuppedo; the abstracts in -ido 8. The analysis of cuppes, quasi 'gourmand, gourmet,' as a com- pound whose posterius is -ed- 'eating' and its prius cupa, quasi 'trencher' (lit. 'trough'), proceeds from Walde, and is undoubtedly 1 The Sanskrit word for ' leech,' jaluka \ jalaukd \ jalaKkas-, with perhaps a dozen other variations with jal-, meant, according to indigenous interpretation, ' aquam- habitans.' But because of Ir. gel, Cornish ghel (glossed by sanguisuga) , the indigenous interpretation has been challenged (e.g., by Brugmann, ALL, 15, 3 fn.). There is a root gu>el- 'consumere' (see cognates in Fick-Stokes, Wtbch.* II, p. 112, s.v. geld; cf. also Lith. gelti ' stechen'), and consumens is no bad description of a ' leech.' Still, a Latin Confixes -edon-, -edno- 'Eating' 319 correct in principle. It may be asked, however, whether cup(a) is not rather a dialectic form for copa ( : caupo) 'hostess.' Then Cupedo, the proper name, was derisive (cf . on Dossennus, § 20) = qui copam comedit, unless cupa = caupona 'cook-shop.' But for cuppedo 'tidbit' we have to ask what was its relation to kottcL&iov C?icoir-a&iov = ' morsel-bite'; cf. co/iaSto? 'roh-essend,' §2). The forms cupp-edia and cupp-edium leave no doubt as to their posterius. 9. In cuppedo I see the source of cupldo, and it is curious that Lucretius uses cuppedo for cupldo. This was certainly either by way of refreshment of the etymology as Lucretius conceived it, or by way of restoration of a genuine old form known to him, though among his predecessors from Plautus on the form cupido is alone in use. 10. Given a name Cuppedo felt as "Mr. Greedy" and cuppedo as 'greediness' — a form which, under the influence of the "law of mam(m)illa," would tend to cUpedo — the proportion cupedo : cupidus =gravedo : gravidus could hardly have failed to suggest itself. By the influence of cupitum, so far as a dialectic interchange of e 1 I be not reckoned with, the shift of cupedo to cupido was fostered. Besides, Etruscan influence apart, the proper name Cupiennius, if from *cupi-ednios l 'qui cupit ehava,' would warrant a form cupi-eden-, contracted to cupiden-. primitive compound *g w eluka, from g">elu- 'water' (in Olr. oil . i . uisge) + uka 'wohnend' (: Goth, bi-uhts 'gewohnt'), after coming to be felt as 'leech,' was liable to conception as a diminutive, with subsequent reduction to a non-diminutive g^eKfi)-. 1 The d of *cupi-ednios is not a mere shot in the air but is to a certain extent warranted by Oscan Perkedne[is 'Percennii,' which I have before now adduced by way of confirming my explanation of Plautine soci-ennius as 'qui socium edit' (see AJP, 28, 417), a definition certainly apt for the interpretation of the passage in which it occurs (Aul. 659). The definition of Siio-o-iros might also hold good. In a Greek vase painting reproduced by Gerhard Auserles. Vasenb., pi. CLV (known to me from Duruy's History of Greece, Eng. transl., Boston ed., I, 416), the flute-player in an altar-scene is named 2, and Fr. frayeur =Lat. fragorem (see AJP, 26, 180 and fnn.). Accordingly I would analyze for-mido as a compound, with (s)miden- : Eng. smites (primary sense 'strikes,' and not 'smears' as Skeat has it; cf. Germ. streicht = [1] 'strikes,' [2] 'smears'), unless the sense was 'shrinking' ( : Germ. meiden, root mei-t- | mei-dh-). Two explanations suggest themselves for the prius, for-: (1) for- : ferit 'strikes,' which would make for- mido a tautological compound, 'strike-smiting,' (2) for- (? fort- /ford-) to the root of Eng. dread (OE. on-drcedon 'timere,' OSax. ant-drddari), which would make for-mido mean 'dread-smiting.' An ultimate cognate perhaps in Oprja/ceia 'religio' (= religious practice, while dread = religious awe). Also cf. dpdaaei 'vexat' and, for the sense, Eng. awe 'dread' : Goth, agis 'anguish.' Dossennus 12. The common interpretation of the name of the farce character Dossennus is briefly stated in the lexicon of Stowasser as follows: "der Bucklige, eine stehende Maske der italischen Posse (Atellane), Typus der Pfiffigkeit, verbunden mit dem der Gefrassigkeit." But so far as the material at my disposal permits me to judge, the notion that the Dossennus was a hunchback lacks Latin literary evidence, and is a mere inference due to the etymological connection of the word with dorsum 'back.' Given a name Dossennus of other pro- venience and popular etymology would connect it up with do(r)sus. The best support of this interpretation lies in the proper names Dorsus, Dos(s)us, and Kiessling in his note on Horace Epist. 2. 1. 173 mentions a family of Fabii Dorsuones to which the Fabius Dossenus of Pliny, N.H. 14, 92, may have belonged. 13 a) The literary evidence for the gluttony of the Atellan Dossennus is very clear: 1) Varro, 1. 1. 7, 95, dictum mandier a mandendo, unde manducari a quo in Atellanis Dossenum (MS ad obscenum) vocant Manducum, cf . Latin Confixes -edon-, -edno- 'Eating' 321 2) Plautus, Rud., 535-36, quid si aliquo ad ludos me pro manduco locem ? : : qua propter ? : : quia pol clare crepito dentibus. 3) Horace, Epist. 2. 1. 173, (adspice Plautum) quantus sit Dossennus edacibus in parasitis. This means, I take it, in the form of a statement: Plautus Dossennus est Dossennorum=parasitus est parasitorum. 4) Pomponius ap. Non. 513. 11, publicitus | dantor Dosseno et fullonibus cibaria. 5) Novius, Fullones Feriati, ap. Non. 81. 25, vortit se in omnis bestias; comest quicquid tetigit tantum. 6) Festus, 536. 1 (de Pon.), Novius in duobus Dossenis, sequimini preminatef, sequere temeti timor. 6a)There is less reason for supposing that a Dossenus, qua glutton, is depicted in the following, Pomponius ap. Charisium 1, p. 101, P.: "nam cibaria | vicem duorum me comesse condecet | solum." 14 /8) From the following passages we may infer the cunning (Pfiffigkeit) or, as I rather suspect, the malice (note indica) of the Dossennus: 7) Novius, ap. Non. 514. 23. ergo, mi Dossenne, cum istaec memore meministi, indica qui illud aurum abstulisti ? : : non didici hariolari gratiis. In Seneca (Ep. 89. 7), there is no certainty that the reference is to a typical Dossennus, rather than to an individual of that name : 8) hoc verbo (sc. ia) Romani quoque utebantur, sicut philosophia nunc quoque utuntur. quod et togatae tibi antiquae probabunt et inscriptus Dossenni monumento titulus, hospes resiste et sophian Dossenni lege. 15 7) In another group of passages the Dossennus is character- ized by lewdness, viz., in 9) Pomponius ap. Non. 516. 24. praeteriens vidit Dossennum in ludo reverecunditer non docentem condiscipulum, verum scalpentem natis. 1 1 In the light of this passage I would emend the Novius line cited in (6) above as follows: sequimini premi nateseque [re] temeti timor. I interpret praemi nateseque and temeti timor as derisive nicknames of the two Dossenni. The complex praemi nateseque is a parody of something like virtutis pedisequae (Auc. ad Her.) or eloquentiae pedisequa (Cicero). Should one question the propriety of halving a proceleusmatic word between the tread (thesis) and the take-up (arsis) of a dactyl followed by an anapaest in the senarius he should not forget that the compound nate-seque has all the prosodic privileges of two pyrrhic words, and that words of the type """" w — at least in Plautus — were pronounced with accent on the initial syllable: thus ndteseque exhibits conformity of accent and ictus. 322 Edwin W. Fay A similar note perhaps in 10) Pomponius ap. Non. 515. 5, . . . . nolo parciter tamquam frater mihi sis; medium abdomen tecum dividam, though abdomen may here mean pinguedo carnis and refer to gluttony. In the following a combination of glutton and rake : 11) Pomponius ap. Non. 477. 5, - nescioqui, asellam urget quasi asinus, uxorem tuam ita opertis oculis simitu manducatur ac molit. 1 16. At the end our net gain concerning the Dossennus is small: (a) He was certainly a glutton. (/3) He was perhaps a malicious informer or a prater of a "philosophical" turn of mind (cf. the char- acter of Castell in ValdeV El Capitan Ribot, which combines "philos- ophy" and malice in a high degree). (7) He was sometimes, at least, a lewd libertine. So much for the literary evidence for the character of the Dossennus. We have nothing further in the way of evidence for fixing the significance of this name. All the evidence we have is compatible with the following derivation, from dorso- ' back '-{-ennos (from -ednos) ' edax.' This definition is applicable (a) to the glutton, he who actually feeds upon the back; (/3) to the malicious informer, or 'back-biter'; (7) to the libertine, nates scalpens (cf. no. 9, above). 17 a) The glutton. The back was an epic delicacy, awarded as a mark of distinction to Homeric heroes (H 321, f 437), and fed upon by Aeneas and his soldier boys (8, 183, "vescitur . . . . | perpetui tergo bovis . . . ."). In Plautus the back, usually in the form tegora, 2 was one of the delicates; cf. Eng. bacon, ultimately the same with back. I find no literary instance of dorsus in the sense of 'bacon,' but we have the gloss dossum 'lumba, la-x^ov' (=ham), and dorsus is often glossed by vmrot (v&tov), which Homer used, 1 In the equivoques here manducatur may be taken literally of a glutton, the ass with blinkers, managing to get something to eat while grinding, but the suggestion of 'fumbles' (manu ducit) may lurk in the equivoque, and the equivocal sense of molere, permolere is well attested in erotic poetry. There is also an equivoque in asellam ( — jwSXi) 'the nether millstone') roughly fashioned as a counterterm to asinus =6vos ' the upper millstone.' But those who bethink themselves of the Pompeian millstones will note the queer functional inversion of the — asella and asinus. Is manducatur allusive to the love bite (see Fay TAP A 37, 20) ? 2 Capt. 902, 915 tegoribus (dat.), Ps. 108, tegoribus(a.bl.) . This is an interesting instance of the allocation of a special meaning to a special flexion (es-stem), reinforced by a dissimilation whereby the first r of terges- is dropped. Further citations are Ovid Met. 8. 649 (tergore, of a chine of bacon), Phaedrus 2. 1. 9, diviso tergore (of the cutting up of a bullock). Latin Confixes -edon-, -edno- 'Eating' 323 both in the singular and plural, to describe the best pieces of meat at a feast. So buffalo-rump was the delicacy of our plains hunters. But in the compound Dossennus the stem dorso- may well exhibit a sense lost to the simplex. The case is as though we had a name *Arm-ennus without the Horace and Ovid instances of armi leporis. 18 /3) The malicious informer. The literary evidence for this definition is not strong, but so far as it goes it is bound to suggest to English folk our word backbiter, and the specific character of Sir Benjamin Backbite in Sheridan's School for Scandal. If the qualification of hunchback be added — and a literary type is extremely tenacious of life, extremely prolific — we may think of the treacherous and malicious hunchback of Boker's Francesca da Rimini, or of that pander and libertine, the hunch-backed father of Gilda in Verdi's Rigoletto. Such parallels imply the survival through the centuries of the literary tradition of the Dossennus. But if Dossennus started life as ' back-eater,' the qualification of the bent back must have been added subsequently by a popular etymology. 19 7) The libertine. In no. 9 above scalpentem natis may be a translation, as it were, of the foregoing Dossennum. For other tautological definitions of this kind cf . the Homeric examples I have collected in AJP 25, 172. 20. Doss-ennus is precisely parallel in its makeup with compounds like Skr. Uksdnna-s 'ox-eating' (cf. Wackernagel, ai. Gram. II, §22, p. 53) and vacanna- 'cow-eater,' 1 and with Plautine pulti-fagus 'pulse- eating.' See further instances above (§ 10, n.), where I have ex- plained soci-ennius. 21. For composition with the root ed- 'edere' I also cite Benr- vr/a-To? and So/jtt^o-to'? 'mealtime.' Indeed, a rather close relation may exist between SopTr-no-Tv; and Doss-ennus, for the comparison of Zopirov 'supper' with Alban. dorke justifies a base dork w - (cf. G. Meyer Alb. Woert., p. 61), doubtless an extension by a root deter- minative of the root der- as it appears in Germ, ver-zehr-en. It is 'The Roman names Vaccaena (CIL vi, 13858) < Vac >aenia, and Vacaenius (x, 3034) may also be derived from *wakd ' cow ' + Sdno- 'eating.' It has lately been pointed out that Lat. vacca : Skr. vaca owes its cc to use in a call (cf. our cow call "Bossy"). — The rule should be stated as applicable to cries and calls, which means that in most cases consonant gemination is due to the vocatival initial accent, as to which in Latin cf. Nigidius, ap. Gell. 13. 26. 1. 324 Edwin W. Fay permissible to derive Doss- from dorqu-(e)s-, and then Doss-ennus will mean 'supper-eater.' 22. In proper names -ennus is a common Latin ending, often of Etruscan derivation, and was perhaps felt as a sort of patronymic, in token whereof Schulze (Lat. Eigennamen, p. 283) has brilliantly and definitively accounted for levenna (Laberius, ap. Gell. 16. 7. 11) as ' levis 6/c rpiyoviai.'' But the Oscan name Perkedne [is ' Pescennii ' cited above in § 10, n. seems a very valid reason for challenging the Etruscan origin of every name in -enn(i)us. Schulze's index (op. cit.) gives the following in -ennus: Bag(i)enni, Dercennus, Passennus, Percenn-, Peucennus, Sebennus, all of which, save the first, seem to consist of names of things edible (Derc- : Sopirov, see above) followed by -ennus. Examples were given above of nicknames that character- ized individuals or groups by their food, and to these I add the name Eskimo, derived from Wiyaskimo-wote 'raw- (flesh-) eaters' (cf. a>fir)ffTrp, Vedic amad, with the same sense), a derisive name bestowed upon the Innuit people by their Cree neighbors. To the class of derisive nicknames Dossennus 'back-eater' (or 'supper-eater') will belong. Austin, Texas