My Interpretation of the Croatian Toponyms
The Salt Lake
on the Mljet Island.
It's sometimes suggested
that the islands
were once the places richest
in toponyms, because people
had to use
every single source of
fresh water and every single
piece of fertile land.
ATTENTION: Some of the opinions stated in the following text are contrary to the mainstream science. I will not advise you to read it if you don't have a substantial background in linguistics. I am not a conspiracy theorist who wants to bombard people with controversial statements they don't know how to evaluate. If you are ready to read it, click here.
SUMMARY: It will be argued that it is impossible that the Illyrian language died out in ancient times, as the mainstream historical linguistics claims, but that it is nearly certain that it, despite never being attested, died relatively recently and that many, if not most of the, Croatian toponyms come from it. The main argument for that is that, if the toponyms came from the Croatian language, most of them would make sense to those who speak Croatian, yet the vast majority of them don't. Alternative etymologies are proposed for 77 toponyms in or on the borders of Croatia. Some basic vocabulary of Illyrian is reconstructed, the most certain words are: *issa~iasa, meaning "source of a health-giving water", presumably coming from the Indo-European root meaning "boiling" or "spring", *karr~kurr, meaning "to flow", presumably from a Pre-Indo-European substrate language and cognate to the Late Latin fish names "carpa" and "carassius", *kull~kl meaning "mountain", presumably from the Indo-European root meaning "high", *mar~mur, meaning "marsh", almost certainly from the Indo-European root meaning "marsh" or "sea", *api, meaning "water", from the Indo-European root meaning "body of water", the word *pis~pik meaning "coniferous forest", almost certainly related to the Latin word "pix" (resin), and the word *melit~malit, meaning "sea island", probably of Pre-Indo-European origin. The evidence is consistent enough that some grammatical features of Illyrian can also be inferred, such as the /a/~/u/ ablaut. My conclusion is that Illyrian was a centum Indo-European language with some specific sound changes, such as *bh turning to /p/ in some conditions and into /β/ (voiced bilabial fricative) in others, *kj turning to /j/ (the semi-vowel) in front of a nasal, *gj turning into /x/ (velar fricative) in at least some conditions, single consonants being geminated after a short vowel (presumably because all the syllables had to be of equal length), the diphthongs *ew and *ey monophthongized into long /i/, syllabic consonants turning into /i+C/ and, most importantly, the initial laryngeals being preserved as /e/ (presumably because of a word-initial accent). The reconstructed grammar is then exemplified by translating the Lord's Prayer into Illyrian.
Croatian is a Slavic language. However,
I think there aren't actually that many Slavic toponyms in Croatia.
If there were, they would make sense to those who speak Croatian, and
the vast majority of them don't. Many of the toponyms commonly accepted
to be Croatian in origin don't even sound (phonotactically) Croatian,
such as Delnice (the cluster -eln-, in native Croatian words,
the 'l' in -iln- and -eln- gets vocalized into 'o',
such as in "čeoni", derived from "čelo", or "dionica", derived from
the verb "dijel-i-ti"). So, I argue that attempts to explain
the Croatian toponyms without an obvious etymology with Slavic roots
is, in fact, a very flawed method, and I attempt to reconstruct a
language of the ancient Croatian toponyms (the Illyrian language) and
explain the toponyms using it.
To give you an example, Issa, the ancient name for the island of Vis, isn't explicable by Latin or Greek, and is widely stated to have an unknown, perhaps Pre-Indo-European etymology. But here is a clue: there were some quite strong springs there in ancient times, that the Roman baths were getting water from. And wherever there were large Roman baths in Croatia, the same or similar element repeats. The ancient name for Daruvar was Balissa and the ancient name for Varaždinske Toplice was Iasa. There we have it: *issa~iasa meant "spas" (health-giving springs). And I believe I can trace that name back to Proto-Indo-European: it probably comes from *yos (spring, derived from *yes, "to boil") and the collective noun ending *eh2 (probably pronounced "ah"). The Bal- in Balissa can be explained as being related to the Latin word bullire (to boil), so that Balissa means "boiling health-giving springs". Using that knowledge, we can explain some modern Croatian toponyms without an obvious Croatian etymology, like Jozinci (assuming an ancient name such as *Iasona, meaning "on the health-giving spring", in the sense "on the source of salt water", since salt was believed to have health-giving properties by people of ancient Slavonia). Namely, Jozinci is a small village near Donji Miholjac. It's known that, in ancient times, there was a stream of brackish water flowing through Donji Miholjac, but it's not certain where its source was. It might have very well been in Jozinci (show/hide details).
Of course, things are a lot more complicated than the above paragraph might lead you to believe. Here is what Dubravka Ivšić, perhaps the most prominent Croatian etymologist (supporting the mainstream interpretations of the Croatian toponyms), said about it when I asked her via e-mail: Naselje "Issa" su osnovali Grci, a ne Iliri. Rimske terme su bile jednostavno kupališta, nisu nužno bile ljekovite. Puno je vjerojatnije da je "-issa" u toponimima poput "Balissa" i "Certissa" sufiks nego da su to složenice (uspoređuju se s keltskim toponimima u kojima je "-issa" sufiks). Drugim riječima, po svemu sudeći, toponim "Issa" morfološki nije usporediv s "-issa" u Balissa, a i čini se da ti toponimi pripadaju različitim slojevima, tj. različitim jezicima (treba voditi računa i o geografskoj udaljenosti). "Iasa" u "Aquae Iasae" se i inače izvodi od indoeuropskog *yes 'vreti', i moglo bi pripadati istom jeziku kao "Balissa", ali ponovno nije morfološki usporedivo. Ako i pretpostavimo da "Issa" i "Iasa" pripadaju istom jeziku (što je malo vjerojatno), trebalo bi objasniti razliku u obliku (ako ciljate na indoeuropski ablaut, treba objasniti kakav je to ablaut i zašto se razlikuju stupnjevi).
Dubravka If you are not able to understand it, not even using Google Translate, which usually does a good job translating Croatian to English (though not the other way around), perhaps you shouldn't proceed reading.
Anyway, I find the arguments rather weak, primarily because the mainstream linguistics considers the name "Issa" to be related to the toponym Antissa on the Greek island of Lesbos, and Vis and Lesbos are even further away from each other than Vis and Daruvar are. As for the Celtic suffix "-issa", it probably doesn't go back to Proto-Indo-European for it to be relevant here, since Proto-Indo-European phonotactics generally avoided geminates (show/hide details). There does indeed appear to be a suffix "-issia" in Illyrian. As for the differences in the grades of ablaut, that's not so surprising if you take into account that there were two words for spring in Proto-Slavic that were different only by a grade of ablaut, *virŭ and *vorŭ. Again, I can't promise you (and probably nobody can) to know all the relevant technical details.
Here is how Dubravka Ivšić responded to the question "Why would you assume that the suffix -issa, found in Celtic languages, also existed in Illyrian, when it obviously couldn't have existed in Proto-Indo-European (whose phonotactics didn't allow geminate consonants)?": Ako je sufiks -issa postojao u ilirskom, onda je to bio ilirski sufiks, kao što je svaki sufiks kojim se tvore hrvatske riječi hrvatski sufiks.
Međutim, postavlja se pitanje što je to ilirski. Po onome što se danas zna, čini se da se naziv "ilirski" koristi za domorodačke indoeuropske jezike na našem tlu (i šire) koji su se govorili prije latinskog i grčkog (neko vrijeme su supostojali), i čije su jedine potvrde toponimi i antroponimi zapisani latinskim i grčkim slovima. Vjerojatno je bilo više "ilirskih" jezika (npr. liburnski, delmatski, histarski) i nije jasno koliko su bili srodni. U novije doba govori se i o jednom takvom "panonskom" jeziku, kojem bi pripadali i (neki) predslavenski toponimi u sjevernoj Hrvatskoj. U sjevernoj Hrvatskoj još postoji i teoretska mogućnost keltskih toponima (iako je samo Carrodunum sigurno keltski), pa ako je Certissa keltskog podrijetla, onda je u njemu sufiks -issa keltski. Uglavnom, to je metodološko pitanje - kako ćemo što definirati, a o svemu skupa ima premalo podataka da bismo išta sigurno tvrdili.
To što u praindoeuropskom nisu postojale geminate ne znači da se one nisu mogle razviti u pojedinim jezicima. Her reply doesn't make a lot of sense, does it?
By far the most common counter-argument I get is that, if a toponym doesn't come from Croatian, perhaps it comes from some other attested language, the most commonly proposed being Latin. Most of the such proposed etymologies can be easily refuted, click here to see (or hide) how.
Many, if not most of the, supposed Latin etymologies of the toponyms of modern-day Croatia can be refuted simply by pointing to the basic facts of the historical phonology, that I think everyone who studies Croatian toponyms knows about, yet many people choose to ignore them. Here are the ones that I consider to be most important. Classical Latin long /i/, in Vulgar Latin, turned into a sound similar or identical to the Old Croatian front yer, that later turned into 'a' (or 'e' in the Kaikavian dialect) or simply disappeared. A beautiful example of that is the toponym Cavtat (folk etymologies often connect it to the Croatian word cvat, meaning flower), derived from Latin Cīvitāte (Vetere) (in the old city). Similarly, the ancient name for Caska was Cīssa (probably an Illyrian word related to the Greek name for the ivy plant, κισσος). Classical Latin short /i/, in Vulgar Latin, turned into a sound similar or identical to Old Croatian yat, which, in standard Croatian, turned into the diphthong /ie/ (spelled 'je' or 'ije'). A beautiful example of that is the toponym Srijem, from the Latin name Sĭrmĭum (the name is obviously Illyrian in origin, from the Indo-European root meaning "flow", as there are many rivers there, the /r/ and the yat switched places regularly due to the metathesis of the liquids, operating in Croatian until the 11th century). The apparent counter-examples (Silba<sĭlva, Piškera<pĭscis...) are confined to the dialects in which the Old Croatian yat regularly turned to /i/. The same is true for the Classical Latin short 'e', which is illustrated by the Latin name for Mljet, Melita (which is probably a Pre-Indo-European word for sea island, compare the toponyms Molat and Malta). Classical Latin long /u/ turned into a sound similar or identical to the Old Croatian back yer, which behaved just like the front yer, except that it didn't participate in the Second Slavic Palatalization (still productive in modern Croatian). For example, the Latin name for the island of Krk was Cūricum (perhaps from an Illyrian word meaning "north", related to Lithuanian "šiaure" and Latin "caurus"). Vulgar Latin short 'u' turned into 'o'. For instance, the Latin name for Rovinj was Ruinium. The only counter-example I know is the name Trogir, derived from the Latin name Tragurium, but that's actually a name (apparently) derived from a Greek word (τραγος, goat), and the 'u' may have been pronounced /y/. Vulgar Latin long /o/ usually gets borrowed as /u/ in Old Croatian, because, quite simply, Old Croatian didn't have long /o/. An example of that is the toponym Pula (from the Latin name Pōla, probably from an Illyrian word related to the Greek word πολις). Apparent counter-examples (Labin<Albona, Solin<Salona...) are explicable as Croats changing the suffix to fit their declension patterns. And, of course, the most obvious rule, yet apparently sometimes ignored, the Classical Latin long and short 'a' came to be pronounced as a sound that was almost always borrowed as 'o' in Old Croatian (Trogir<Tragurion, Solin<Salona, Mosor<Massarum...). Almost all the apparent counter-examples (Labin<Albona, Raša<Arsia, Skradin<Scardona...) are explicable as a result of the metathesis of liquids, which regularly turned *orC and *olC into /raC/ and /laC/. Other apparent counter-examples can be explained as borrowing directly from Illyrian (Drava, Sava...) or, coincidentally, unknown phonotactical details of Old Croatian (the Latin names for Drava and Sava were Draus and Saus, but Proto-Slavic didn't allow wovel sequences within a morpheme). Of course, sometimes (usually quite obvious) folk-etymologies change the accents (and therefore, by affecting the later sound-changes, the vowels), such as in the case of the island name Lastovo (the ancient name was Ladesta, perhaps an Illyrian word derived from the Indo-European root meaning "water", but the modern name looks as if it was from the Croatian word for the swallow bird, "lasta"), however those things happen rather rarely. Another rule, maybe slightly less obvious, but still hard to miss if you study Croatian toponyms, the Classical Latin /v/ turned into /β/ in Vulgar Latin, which was regularly borrowed as /b/ in Old Croatian (Silba<silva, Bribir<Varvaria...). Therefore, the toponyms starting with a 'v' can't be Latin in origin. Those rules are especially accurate if you look at the names in the local dialects. For instance, the name for Brač (ancient name Brattia) in the local dialect is Broc, the name for Hvar (ancient name Pharos) in the local dialect is For, and the name for Šolta (ancient name Solenta) in the local dialect is Sulet (probably because the exonyms weren't actually borrowed from Vulgar Latin).
Yet, those things are often ignored by the etymologists. For instance, the mainstream etymology considers the toponym Viljevo to be derived from the Latin word vīlla. The simple truth is, vīlla, borrowed into Old Croatian, would give something like *Bal (genitive singular *Bla) in Modern Croatian (in my opinion, the name Viljevo is indeed related to the homonymous dialectal Croatian word meaning oak forest, which is perhaps derived from a phrase meaning "in the Saint Elias'es forest"). The mainstream etymology considers the toponym Batina to be derived from Latin *Batinus (dedicated to Bato the Illyrian). Well, first of all, the grammatical way for naming a city after a person named Bato would be *Batinianus. Second, *Batinus, borrowed into Old Croatian, would give *Boten or *Botan in modern Croatian. So much about it. Some people (though it's not really mainstream linguistics) say that the name of the river Una derives from the Latin word ūna (one) supposedly in the sense "the most beautiful". Well, first, if you were to name a river as "the most beautiful", you would use the Latin word primum (flumen), and certainly not una. Second, ūna, borrowed into Old Croatian, would give *Van in modern Croatian (the prothetic 'v' would have to be added in front of the back yer). I hope you see where I am coming from. The phonology of Vulgar Latin is known, and it doesn't explain most of the modern Croatian toponyms. The phonology of Illyrian is not known right now, but it can hopefully be reconstructed from the toponyms.
And the same goes for some other languages that are suggested to be the source of modern-day Croatian toponyms. Mainstream linguistics connects the ancient name for Split, Spalatum, with the Greek word for spiny broom (a flower), ασπαλαθος. Yet, the modern name Split points to it being borrowed as *Speltos, with an alternation between 'a' in Spalatum and 'e' in *Speltos. This alternation would have been irregular in Ancient Greek, but it might have very well been regular in Illyrian (since, for instance, the name Dalmatia was also attested in antiquity as "Delmatia", and the ancient name for the island of Brač, "Brattia", was also attested as "Brettia", and those names are without a doubt Illyrian). In addition to that, a placename being apparently derived from a plant name or an animal name without there being a suffix (as in the case of Spalatum being apparently derived from aspalathos) is a typical sign of a folk-etymology.
Here is a list of the etymologies of the Croatian toponyms I've supported on the "Croatian Toponyms" thread linked on the left:
Aenona (Nin) < *h2ekj-mon (where a lot of stones are)
Albona (Labin) < *h2elbh (white). It's unknown what the precise pronunciation of 'b' was, some sources spell it as 'v'. It might have been /β/.
Alma (Fruška Gora) < *h2elm (fertile)
Almissia (Omiš) < *h2elm-yess (on a fertile ground)
Andautonia (Zagreb) < *h2en-dheh2-ont-om (near that which flows). The name Zagreb itself probably means "the God's hill", from the Illyrian word Dzis (God), mentioned on the Messapian inscriptions, and a word related to Proto-Slavic *gora and Sanskrit giri. So, the Illyrian name might have been *Dzigurevos, borrowed into Old Croatian via Vulgar Latin.
Apševci < from an Illyrian name (borrowed directly from Illyrian, not via Vulgar Latin) such as *Apissios, meaning "river island", derived from the Indo-European root *h2ep (water), for the semantics compare the Croatian toponym Otok near Vinkovci. Similar names occur repeatedly in microtoponymy, for instance, near Donji Miholjac there is a field called Apšaldol.
Arsia (Raša river) < *h3rews (to flow fast)
Asinium (Sinj) < related to Proto-Slavic *asenŭ (ash tree). Or, perhaps even more likely, related to the Hittite word haš (stone), since the same element repeats in the names of other stony places. For instance, the ancient name for Osor was Assarum.
Balissa (Daruvar) < *boll-yos-eh2 (where a lot of health-giving boiling springs are)
Baranja < some sort of derivation from *mory (marsh), compare also Greek βορβορος and the ancient (almost certainly Illyrian) name for the river Bojana, "Barbanna". The Illyrian name might have been *Barrania, with the first 'a' stressed and long, compare the ancient name for Dinara, "Dindaria" (while the modern name looks like as if it was from *Dinnaria), for the consonant cluster, in this case -rb- to -rr-, simplification.
Biđ (the river) < From an Illyrian name such as *Vōdios, from Proto-Indo-European *wed (wet), borrowed into Old Croatian from Vulgar Latin (therefore, 'v' regularly being borrowed as *b, long 'o' as *y, turning into 'i' in Modern Croatian, and 'di' regularly yotated into 'đ').
Bosut < *bel-sewh1-nt (strong waterer), see "Sutla" and "Sunja" for the sound changes.
Cersia (Cres) < *(s)kwer (cliff)
Certissia (Đakovo) < *(s)ker-yess (on a sharp, infertile ground)
Cibalae (Vinkovci) < *kjey-bel (strong house)
Colapis (Kupa) < *kwol-h2ep ("wondering water", river with many meanders)
Dalj < *dhel (milkmen, the Illyran name might have been *Dillioi)
Daljok (Duboševica) < *dhel (milkmen)
Daurum (Dubrovnik) < *deh2w (fire, in the sense "a place that was built where the forest has been burnt down"). The toponym *Daurum wasn't attested in antiquity per se, but the ancient name for Cavtat (some 10 kilometers from Dubrovnik) was Epi-Daurum ("epi" almost certainly being the Greek prefix meaning "beside"). *Daur- borrowed into Old Croatian would regularly give *Dubr in modern Croatian, for the sound changes compare the ancient name for Bribir, Varuaria.
Dilj (the mountain) < *delh1 (wide, the Illyrian name might have been *Dilios)
Dindaria (Dinara mountain) < *dent (tooth)
Drava (the river) < *drew (to pour)
Dunav (Danube river) < *danu (river, It's unknown what was the first vowel in Illyrian, because both nasalized 'a', nasalized 'o' and nasalized 'u' turn into 'u' in Croatian)
Eira (perhaps the river Mura) < *h1er (to flow)
Ervenica (a stream in Slavonia) < *h3rews (to flow fast)
Esseg (Osijek) < *h1es-seg (healthy field)
Histria (the Istria peninsula) < related to Proto-Slavic *ostrovŭ (island), maybe in the sense "people from the island".
Hrvat (Croat) < *ser-h2ekw-ot (one who lives near a stream), from some language related to Armenian (where PIE *s gives 'h')
Incerum (Požega) < *h1eyn-kjer (the heart of the valley)
Iasa (Varaždinske Toplice) < *yos-eh2 (where a lot of health-giving springs are)
Iovallum (Valpovo) < *yow-h2elom (magical, health-giving herb)
Issa (Vis) < *yos-eh2 (where a lot of health-giving springs are)
Kalnik < *kelh3 (high, the wovel in "Kal-" must have originally been short 'u', reguarly changing to back yer and then to 'a' in Croatian, and there must have been either short 'i' or short 'u' after l, disappearing due to the Havlik's law. So, the Illyrian form was either *Kullunnicos or *Kullinnicos. The toponym was borrowed from Illyrian into Old Croatian after the Third Slavic Palatalization, but possibly before the Second Slavic Palatalization.)
Karašica < from a verb *karr~kurr meaning "to flow", perhaps of Pre-Indo-European origin
Kentina (Sinjsko Polje) < *kjem-t-h1eyn (the valley of horses)
Klek Mountain < *kelh3 (high, the change from *kelk to *klek would be regular in Croatian, because the methatesis of the liquids ended way after the first Slavic palatalization.)
Korana < from a verb *karr~kurr meaning "to flow", perhaps of Pre-Indo-European origin
Krapina < from a verb *karr~kurr meaning "to flow", perhaps of Pre-Indo-European origin, -ap- is probably from the Indo-European root *h2ep (water)
Krbava (likely originally a hydronym) < from a verb *karr~kurr meaning "to flow", perhaps of Pre-Indo-European origin
Krka < from a verb *karr~kurr meaning "to flow", perhaps of Pre-Indo-European origin
Krndija < *(s)ker-nt (cutting, steep). The Illyrian name might have been *Curindea (compare the ancient name for Krk, Curicum, for the sound changes from Old Croatian to Modern Croatian). The change from Proto-Indo-European *-ent- to Illyrian -ind- is hard to explain, but it's not isolated (if Dindaria comes from *dent).
Lika (likely originally a hydronym, attested in the ancient toponym "Epi-licum") < *lewk (clear)
Mantouna (Motovun) < *men (mountain)
Mariniana (Donji Miholjac) < *mory-h1eyn (marshy valley)
Marsonia (Slavonski Brod) < *mory-h2en (a town on a marsh)
Medvednica (the mountain) < from an Illyrian name such as *Menduenda, from the Indo-European root *men (to be high), perhaps exactly cognate to the Latin word mons (genitive singular montis), compare Dindaria for the change from -nt- to -nd-.
Moslavina (perhaps originally an oronym) < *megjs (large)
Mosor (the mountain) < *megjs (large)
Mursa (Osijek) < *mory (marsh)
Neretva < *ner (canyon)
Odra < *wodr (water) The Illyrian name might have been *Udra, coming into Croatian via Vulgar Latin, where short 'u' merges with 'o'.
Orljava < *h1or (spring) The Illyrian name might have been *Arrulia (a river with many springs), coming into Old Croatian via Vulgar Latin, whose 'a' was regularly borrowed as 'o' in Old Croatian and whose long 'u' was regularly borrowed as back yer.
Osseratis (unidentified city near the confluence of the river Vrbas into Sava) < *h1en-ser-ot (where one [river] flows into [the other])
Padoma (an ancient place in the highlands on Vis) < *bheh2gjos (beech tree)
Paklenica < *pikj (resin, the change from short 'i' to short 'a' is regular in Croatian when it's affected by the Havlik's law, so the Illyrian name might have been *Picculla)
Pannonia < *pen (marshland)
Papuk (originally a hydronym spelled "Papugh") < *bhogj (to flow), the Illyrian name might have been *Pippuhos (that which flows and flows), borrowed into Old Croatian directly from Illyrian, compare the Proto-Indo-European *kwel (to turn)>*kwekwlos (wheel) for morphology.
Parentium (Poreč) < *por-h1ey-nt-y-om ("around that which flows", river bank)
Permodios (Premuda) < The supperlative form of Proto-Indo-European preposition *per (in front of), so that it means "the first island in the archipelago".
Pharos (Hvar) < related to Proto-Slavic *borŭ (pine wood), compare the Italian name Lesina (from a Slavic word meaning "wood").
Pisinium (Pazin) < *peyH (resin), compare Greek πισσα and Latin "pinus"
Pisunus (Psunj) < *peyH (resin), compare Greek πισσα and Latin "pinus"
Plitvice (the lakes) < *plew (to flow), compare also the German (perhaps borrowed directly from Illyrian) name for the Balaton lake, Plattensee.
Sava (the river) < *sewh1 (to water)
Scardona (Skradin) < *(s)kwor-dhos (big cliff)
Sčitarjevo < *skewH (shield), in the sense "embankment"
Serapia (unidentified river in northern Croatia, perhaps Bednja) < *ser-h2ep (flowing water)
Šibenik < from an Illyrian name such as *Sievennos (borrowed into Old Croatian through Vulgar Latin), perhaps meaning "flax field", derived from *syuh1 (to sew).
Siscia (Sisak) < *sek (to cut), in the sense "where Romans have cut the forest".
Sopje (perhaps from an ancient name *salapia) < *seh2l-h2ep (salt lake), so that it's an eponym of the near-by town Slatina (Croatian for "salt lake"), compare "Almissia">"Omiš" for the sound changes.
Sunja (the river) < *suh1-nt (that which waters the ground), the Illyrian name might have been *Sunneos (with a consonant cluster simplification seen in "Dinara"), the Illyrian nasalized 'u' might have been borrowed as nasalized 'o' in Old Croatian.
Sutla (the river) < *suh1-nt (that which waters the ground), the Illyrian nasalized 'u' might have been borrowed as nasalized 'o' in Old Croatian.
Tarda (Darda) < *ters (dry land)
Ulciria (Učka mountain) < *wlkwos (wolf)
Una (the river) < *unt (wave, the Illyrian nasalized 'u' could be borrowed as nasalized 'o' into Old Croatian, compare "Andautonia" for the change from Indo-European *nt to 'n' in Illyrian). There are some other cases around the world of a river being named after a word meaning "wave", for instance, the river Vilnia in Lithuania. The name might be even influenced by phonosemantics, because "Una" is the name to numerous rivers around the world, and also the Etruscan word for stream.
Valeriana (Baranja) < *wel-h1er (wet valley)
Vuka (the river) < *welk (moisture, the change from syllabic 'l' to 'u' is regular in Croatian). The names such as Vukovar, Vučica and Vučedol share the same root.
So, assuming those etymologies are true, it is, of course, possible to deduce quite a few things about the Illyrian grammar (show/hide details).
Here is what Dubravka Ivsic told me about some other toponyms I've attempted to interpret on this web-page: Poštovani g. Samaržija,
hvala Vam na mailu. Otvorila sam Vašu stranicu, no nažalost sada nisam u mogućnosti detaljno se baviti time. Ipak, nekoliko usputnih napomena. Toponimi poput Kalnik, Paklenica, Plitvice ipak jesu slavenskoga podrijetla (za korijene usp. kal, paklina, plit-ak), Essek je jednostavno mađarski izgovor hrvatskoga Osijek, a toponim Jozinci puno je vjerojatnije izvoditi od osobnog imena Jozo nego pretpostavljati "ilirski" *Iasona. Toponim Mariniana sasvim se lijepo da izvesti kao predijalni toponim, (rimskoga podrijetla), od osobnog imena Marinius. Problem s pojmom "ilirski" je taj što se u njega trpaju svi predslavenski toponimi na našem tlu koji nisu latinskog i grčkog podrijetla, a mogu se izvesti od nekih praindoeuropskih korijena. No, većina tih etimologija je "ad hoc", t.j. nema se s čime usporediti. Osim toga, "ilirski" toponimi su k nama dospjeli kroz latinski ili grčki filtar, pa mi zapravo ne znamo koji je bio njihov izvorni lik. Zbog toga ne možemo biti sigurni niti kako treba analizirati toponime, npr. Vaša analiza An-da-ut-on- ili ipak An-daut-on-ia. Zato u istraživanju toponima predslavenskoga podrijetla treba biti iznimno oprezan, a metodološki ispravan početak svakako je analiza povijesnih potvrda toponima. O mnogima od tih problema pisala sam u svojoj disertaciji, a upućujem Vas i na stariju literaturu o ilirskom (Krahe i Mayer).
Dubravka Ivšić Majić Here are the sound changes I propose happened from Proto-Indo-European to Illyrian:
[bh]>p (perhaps not in all contexts)
C[l]C>il (perhaps not between all consonants)
So, because of the epenthetic vowel 'e' appearing only in the first syllable (Ervenica), I suggest that the stress was always on the first syllable. I also have a temptation to think that the nominative singular actually ended in -i in Illyrian. The suffix -i- is seen in, for instance, Serapia, Krndija, Colapis, and possibly also in Andautonia. I believe that the primary ablaut changed from e/o in PIE to a/u in Illyrian, for instance, in the toponyms such as Mursa, Marsonia and Mariniana. The geminates in the toponyms such as Issa and Pannonia are probably explicable by some consonants being doubled after a short vowel, like in Middle English more or less. It's not quite clear what sound changes the Proto-Indo-European syllabic 'l' (and therefore probably other syllabic consonants) had undergone. It appears to have been retained after a 'w' (Vuka, Ulciria), but not after a 'd' (Dilj), where it appears to have been turned into 'il'.
So, here is what I think Pater Noster might have sounded like in Illyrian:
Patir nos, kis divi esi!
Kekluit ennomin tepi!
Puind ki tu vessi
Ki nos ne essenti dogi toimi, dodi!
Pagdike likona nos,
kom vike pagmos likettirmos nos.
Sentme ne dirtomos,
solaske inzme tirnomos.
The Roman Thermae
in Issa (Vis)
were getting the water
from a mineral spring
that doesn't exist
However, it's possible
that Issa was
named after it,
from the Indo-European
root *yos (spring).