Eric P. Hamp CDU 811.163.6: 811.18 Department of Linguistics, University of Chicago 
THREE BRIEF STUDIES: SLOVENE AND ALBANIAN 
I. Slovene enajst 
Bernard Comrie, in his contribution on the Slavic languages to the volume edit­ed by J. Gvozdanovicl (p. 766), reports a fact of particular historical interest. In pre­senting the forms and behavior of the 'teens amongst the numerals' Comrie states that two accentual patterns apply, an oxytone and a recessive, i.e. -najst and VNIT­najst. However, he points out that exceptionally in the case of '11' only enajst the oxy­tone occurs. 
In the same volume, as it happens, I have mentioned (919-20) my observations2 ofthe IE rule deleting 'l' in complex phrases. It would seem that we may reconstruct enajst as *na(d)es(?)t by this rule, and then only later a »normalizing« e(n)-would have been added. An interesting archaism matching OLith. liekas. 
Since I am no Anglist I do not know what has been said on this, but contrary to German elf, zwolf (and apparently Speyer Gothic ainlif) English has e/even but twe/ve3. This should point to a post-Verner *lif(V)n, but *tw6lifunder the same rule, preserved in this marginal insular dialect. 
Jadranka Gvozdanovic ed., lndo-European Numerals. Berlin: Mouton 1992. 
2 
To the references cited add 'first', ZCP 45, 1992, 85-6. To D. Greene's report (502-4) 1 have some further comment to offer in the pages of Eriu. 
3 
1 notice the following, which seem surprisingiy inadequate and inattentive. Chambers 1993 (normally compact, sound, and independent) OE en(d)Ie(o)fan: Go ainiif perh. (10 and) one left: Lat. linguere. Merriam-Webster Collegiate 10 (a popuiar reputabie standard) ME enieven < adj. < OE < end--iin + -Ieofan : Ieon 'Iend' : Ioan <ME Ion< (OE Iren =) ON Iiin 
Random House Unabriged2 (1 must accept Iapsed responsibility) ME elleven(e) -<ME Ien(d)en < OE Irenan < OE ellefne, endieofan: OHG einlif> eif, ON ellifu: LEAVE < Ieven < Irefan (caus.): Iiif'remainder': OHG Ieiban'--+ bieiben, ON Ieifa: Go. -Iaibjan. The mention Iii refers one to Iave2. 
Of course, in such dictionaries much must be omitted, and not only rich detail; but not the main features of shape and content of the word, while introducing dubious reievance. 
From the forms cited we may recover the English series: end'Ieofan > en'Ieven (with removaI or reduction of whatever seemed to be end-) > e!'Iefne (by phonetic simplification) > e'Ieven (further reduc­tion) > [J"!svQ.] (modem vaiues). 
The following reconstructions are then imposed by known comparanda: *en-'Iefan-< *aina-'Iiq,ana (Old English post-Verner intervocalic spirant voicing) ~*aina+'liq,a-na-(Germanic initiaI accent) ~*'Iiq,a­(na-)(+'tehun) (loss of deietion ruie for phantom 10, and survivaI of entai!ed inflexion or suffix) or *'lihJa­: OLith. Iiekas; versus *NumeraI [du.jpl.] + 'lihJ/kwa-(+'10) < *(1-deietion) + Numeral + I(e)ikw-o-+X. The last stage belonged to a more general PIE rule for NPs than just Numeral sequences; 11 is a precious relic. 
II. Albanian kercas and kercej 
L V. Orel in treating Alb. kercas 'shout, resound' (1998: 180) subscribes to a rea­sonable conclusion, that we have an early borrowing from Slavic, but his account of the detail is inexact. The affricate c is not the direct reflex of Slavic č. The primary modem result is Alb. s, e.g. porosis <poročiti 'order' (correct at Orel 338). Another instructive example is sorre 'crow' < IE *kwersnii, where Romanian cioarii with its sub­stratum borrowing as *bril in the Roman period shows us a relevant chronological benchmark; but this tirne Orel 399, with many conflicting but unannotated refer­ences, essentially reproduces my argument but fails to say that we have here an ini­tial labio-velar before the front vowel *e. This Albanian evidence has consequences for Slavic črno as well as Baltic and Indic. Demiraj 355 will give much ofthe missing exact content, although both fail to report that the Dehnstufe (not just »a different ablaut«) of Albanian matches precisely the accent contrast of Slavic (=Baltic) v6rna against v3rn:Js. 1 add this to the interesting features linking Proto-Albanian to Balto-Slavic. 
The reason, then, that ki!rcas has a c is simple, and recent. In position after r or l Alb. s opening a stressed syllable routinely affricated; thus (conventional orthogra­phy) sjell 'bring' : pi!rcjell 'accompany; swallow' (also perrlam 'ivy' <ser. bršljan), pi!l­cas 'burst' but preterite plasa (not, with Orel 318, »archaic -c-« and never resulting from *-tj-which too gives a primary s). 
Therefore, in southernmost Albania <;am preterite 3 sg. kritsi, reported by Peder­sen 1895, p. 16, but not noted by Orel, must have the ts (=c) from the present stem, and therefore indirectly documents a preterite *kris-i (3 sg.) etc. We now have a per­fect match for Slavic *kričati. And the stressed vocalism of the present ki!rcas < *krisdtj -or better *kricdc-V, 3 sg. ki!rcet < *kricat-i-t fits the stem of *kriča-ti exactly. 
Orel is concerned that Serbo-Croatian now has only 'warn' as the meaning of this etymon, and that Slovene could not be the source -1 presume he considers it too far away. 1 think the pertinent conclusion must be that Slovene, a classic Bartoli lateral area, once again shows us an instructive conservatism. The semantics of our early Albanian borrowing are right. 
2. kercej is glossed by Orel (and others) as 'dance, jump'. He thinks the primary meaning is dance, and therefore traces it to ritual dance and badnjak. That may be, and 1 am no specialist in these matters. But Pedersen's <;am attests the alternant spelt ki!cej by Orel but ki!tsej by Pedersen; it means, as 1 have heard it in other Arberesh and Arvanitika dialects, just 'jump'. Scutari in his Lessico for S. Costantino Albanese (of which he is a very talented and precise native) lists kcenj as 'saltare' and 'montare a cavallo'; 'ballare, giocare' is litanj, the two senses in which 1 have heard it in many Arberesh dialects. 
Now ifwe consider kapercej 'jump over', 1 find on consulting B. Demiraj 212 that he segments (without any result) as a derivative formation on the word kap '(an)­fassen, erreichen'. I think this is clearly mistaken. When we recall, as I believe I have shown, that an unstressed efollowed by a stressed e results by contraction in a, we can segment kapercej (or Scutari's kapircenj 'oltrepassare') as *ke-eper-cej. Here ke-< *k'YIJ-(= Germanic ga-) I have claimed to be a preverb, often perfective. Now *ke-eper­cej looks surely like a haplology form *ke-eper-kercenj. The element eper (also eperm) 'upper' has been well discussed by Demiraj 168 (and deficient1y by Orel 89 with the wrong result), who accepts my 1981 explanation. Orel's error and Demiraj's mention of my erroneous 1965 comparison rest on their ignorance of my paper Acta Neophilo­logica (Ljubljana) (dedicated to Janez Stanonik) 25, 9-11, where I think I corrected and improved my (and our) old position.4 
The result is twofold: 1) Our compound means simply (perfectively) 'over-jump'; there is no trace of 'dance'. 2) The main verb and this compound are guaranteed by 
S. Costantino Albanese to the 15th century in age by their attestation. This probably really means closer to a millenium, since we must count Arberesh back to the tirne when Arvanitika separated from Epiros, ifwe regard Arberesh largely as an Arvaniti­ka colonization by emigration. 
The only history I see clearly for kercenj is 'jump', and nothing sacral. If the verb is related at ali to kerc 'stump', and this to 'leg', 'shin', 'log, branch', it might seem closer to find a connexion with the matters ofjumping. 
DEMIRAJ, Bardhyl 1997. Albanische Etymologien. Amsterdam. 
OREL, Vladimir 1998. Albanian Etymological Dictionary. Leiden. 
PEDERSEN, Holger 1895. Albanesische Texte mit Glossar. Leipzig. 
ScurARI, Pasquale 2002. II lessico de/la par/ata arbifreshe di San Costantino Albanese. Rende (CS): UniversitiL della 

Calabria. 
III. Cesta and Language survival 
Cesta, near Trubar's birthplace, is the site of an old Roman road. It is known that people moved there to settle after the burning of their village by Turks. Perhaps, in fact, they settled there because of the availability of stone for building on the site. The presence of Slavs here would date to a period long after the Romans. 
The name Cesta 'street' appears to point to an old road, i.e. *strata, as in Strasse in the Ost-Tyrol Drau (Drava) Puster Tal valley or in Stratford or Stretton in Britain, where old stone ruins have been found. Cf. British Celtic Magnis in England (: Welsh mein 'stones', plural of maen), and Stanford, Stansted and Staines (near Heathrow), all to OE stan 'stone'. 
My friend Janez has always pushed me to improvement. My life is a catalog of rejected positions relieved by victories. Many victories were inspired by friends. 
To convey cesta 'strata' to the Slavic presence we are led to an earlier Dolenska term. It is not probable that this was German speech here. It seems, then, that Do­lenska Celtic may have been the intermediary. We thus conjecture that the Slovenes heard 'strata' still spoken Celtic much as in Resia (Friuli) a surviving *oseako was heard oso'ane <[oso-'iane] and understood.5 
lfwe try to conjecture what such Celts actually said, it seems likely that the word was not *kammi/en-'a walk way, run way, path',6 for that might have given *ponti-vel sim. Since cesta could carry the notation of čist-'clean, clear', 7 we might think of Welsh sarn 'highway' <'strewn, spread' <*tsar-no-j-a <*str-n-H-,8 which of course is ultimately cognate with strata. 
Povzetek 
TRI KRATKE ŠTUDIJE -SLOVENŠČINA IN ALBANŠČINA 
I. Slovensko enajst 
Avtor izvaja slovensko enajst iz *na(d)es(<;)t in predpostavlja izbris števnika '1' že v indoevropskem prajeziku (izpričano tudi v starolitovskem liekas) in poznejšo »normalizacijo« z dodatkom e(n)-. 
II. Albansko kiircas in kercej Avtor vidi kot Orel v albanskem kercas 'vpiti, razlegati se' zgodnji prevzetek iz slovanščine. V zlit­niku c vidi mlad razvoj starejšega s v položaju za r in/. Tako bi 3. edn. kercet izhajala iz *kiirset, to pa iz *krisat-<*kricat-, kjer slednje popolnoma ustreza osnovi v slovanskem *kriča-ti. Stari pomen 'vpiti' te slovanske osnove je ohranjen v slovenščini kot obrobnem arealu. Avtor izvaja kapercej 'preskočiti' iz *ke-eper-cej, kjer je *ke-predpona, često za dovršnost (ger­
mansko ga-), eper pa pomeni 'zgoraj'. Rekonstruirano besedo dalje izvaja s haplologijo iz *ke-eper­kiircenj, kjer kercenj pomeni 'skakati' (ne tudi 'plesati') in je morda povezano s kerc 'štor'. 
III. Cesta in preživetje jezika Avtor razmišlja o možnem keltskem viru krajevnega imena Cesta. 
5 Hamp, Eric P„ I nomi di Resia. Ce Fastu? Sotla Nape 33, 1981, 11-16; also review ofG. Frau, Linguistica 28, 1988, 141-7. 
6 Hamp, Eric P„ On *srn in Gaulish and on the prehistory of *camminu. Romance Philology 28, 1974, 17-20. 
7 Hamp, Eric P„ Srbohrvaško cesta, slovensko cefsta, Slavistična revija 31, 1983, 60-1. 
8 Hamp, Eric P„ Varia, Celtica 2, 1991, 33-47 § 3.