Sō̃ma: Remeeting the Finngreek People

This post focuses on the reconstruction of people who would have been relevant to Helleno-Uralic contact.

There is a great deal of human action and interaction that can be inferred from verbs and economic terms I will share in separate posts. Here, I will only address the people themselves who are theorized to have existed within the Helleno-Uralic sphere.

The role of the noītá has already been discussed in an earlier post: They would have been a central figure in Helleno-Uralic contact, as an exalted spiritualist. In addition, it is assumed that nautical crew (rowers, dockmen, etc.), traders, and other workers were part of Helleno-Uralic exchange. However, terms for these roles have not yet been proposed.

Re: the proposals herein, the following Helleno-Uralic people are reconstructed: Warriors (apʰḗ?, iskʰús?, ry̌̃ma), crafters (daídalos), nobility (kǐ́rios), and Proto-Finns (sō̃ma).


strength (PS), wrestling grip (He.)
PS *āppē <? He. (h)apʰḗ / ἁφή < (h)áptō / ἅπτω ‘to attack, take hold of’ < ?

Regardless of whether He. (h)apʰḗ retained h- during PS < He. loaning, it appears that h- was not loaned into PS.

artistic (HF ~ HFS)
NSa. dáiddalaš << PS *tājδētēk <? PF *taitadak ‘to know how’ <?> He. daidállō / δαιδάλλω, daídalos / δαίδαλος <1? PIE *dens- ~ *dn̥s- ‘to learn’

Fi. taito ‘skill, art’ and SaN dáidu ‘skill, know-how’ imply HFS *daídos / *δαίδος ‘artistic skill, cunning, proficiency’. The He. verb daidállō is identified as PreΗ – but if Fi. taitaa is furthermore connected via F *taj- with taju ‘consciousness, sense’, then compare also with He. dáios / δάϊος ‘knowing, cunning’ ~ daē̃nai / δαῆναι ‘learn’, inf. aor. of dáō / δάω ‘to learn, know, teach’ <1 PIE *dens- ~ *dn̥s- (cf. He. διδάσκω). If the Sa. adjectival suffix -laš was loaned from He. or earlier IE (vs. being a “Calque of Finnish -llinen by replacing the inflectional part -ise- with the existing native Sami cognate -žža-.”), then it would be cognate with, or reflexed from, SaN -las < PS *-lës – although the phonological resemblance could rather be a coincidence, which would leave the reconstruction HF(S) *dai-, *daidV-.

attack, impact (Fi.); brute force, might (He.)
F isku ?> He. iskʰús / ἰσχύς < ?

Fi. iskeä would imply H *iskʰéō, whereas He. iskʰúō / ἰσχύω is instead the attested form. Because of this, it is difficult to argue for F < He. loaning. Along with the unknown He. etymology, a F > He. loan is considered.

lord, master, mister (HEr.)
Er. čirjaz / чиряз << He. kǐ́rios / κύριος <1 PIE *ḱeuh₂- ‘to swell, be strong’

NDS defines Er. čirjaz as ‘барин, господин (Ru.); herra (Fi.)’: These are all honorifics to politely address a man (historically a nobleman); and originally meant ‘lord, master’. This semantic development is paralleled in HEr. kǐ́rios.

friend (Man.); defense, protection (He.); ally? (HUg.)
Man. ruma / рума <<? He. ry̌̃ma / ῥῦμα < ἐρύω <1? PIE *u̯eru- ‘to draw’

In light of HUg. pʰy̌̃ma, the primary form proposed is ry̌̃ma vs. rũma.

smile (ΗMa.)
Ma. šərgəžmáš / шыргыжмаш << He. sarkazmós / σαρκασμός
to smile (HMa.)
Ma. šərgəž- << sarkaz(d)- / σαρκάζ- (cf. σαρκάζει ‘smiles’ / ‘μειδιᾷ‘ [Hsch.])
cheek, face (Ma.); flesh, body (He.)
Ma. šərɣə, šürɣö < PMa. *šǚrgə (Aikio 2014, p. 132) < He. sýrks ~ sýrka / σύρκ- <1 PIE *turḱ- ‘to cut’
cut, slit (PS); ? (He.)
PS *sārkē <? He. sárks ~ sárka / σάρκ- (cf. σαρκάζω ‘to tear flesh’)

It appears that the original vowel of the PMa. < He. loan was /y/ (PMa. *šǚrgə <? HMa. sýrka), which later, perhaps independently, evolved into Ma. -ə- / He. -a-. This must have been a popular word, because it can be compared through multiple derivations. It is unclear whether the U terms illustrate loaning from the acc. He. form, or if the absence of -s is due to a limitation on consonant clusters. It is also unclear whether PS *sārkē ‘cut, slit’ might be included in this proposal: Its phonology would be distinctly He., but its reconstructed meaning differs. However, the Sa. reflexes themselves have varied meanings (cf. SaI särgi ‘rib’, SaN sárgi ‘a single rib with the flesh on it’, SaS+SaSk. ‘mark/slit on a reindeer’s ear’), which might suggest a HS meaning ‘cut flesh’ that bridges the semantic gap. It is unclear whether the HMa. and HSa. forms represent separate He. loans, or if they might rather descend from a HFV loan: This ambiguity is reminiscent of HFS págos ‘frost’ vs. HEr. págos ‘hill, mountain’.

Finland, Finn (Fi.); body, human, person (He.)
Fi. Suomi, suoma- < PF *sooma < He. sō̃ma / σῶμα < PH tsṓmə < *twṓmə <? PIE *tu̯ō-mn̥
Con.: PrePG *ǵʰm̥-ōn ~ PrePBS *ǵʰom-yā- (Kallio 1998, p. 617); PFS *sama- < I zam (De Smit, unpublished)

Fi. Suomi is a puzzle for etymologists seeking its ultimate provenance: The conflicts listed are just two of many proposals. De Smit excellently summarizes the strengths and weaknesses of proposals from Kallio and Koivulehto. Although I have objections to each conflicting proposal, I will not argue here against them, as they all appear to me to be reasonable enough. I will rather offer my own solution, and let it be subject to its own share of criticism.

PF *sooma can be regularly derived from HF(S?) sō̃ma (cf. PG *kookka < PG *χōka-; PF *rooka < PG *rōkā). The implication of a demonymic loan from He. is that HF(S) contact must be have been exceptionally intensive; or that it at least occurred in a geographical environment not subject to other IE adstrates (i.e. G, B, and II) to a degree greater than H. One tentative argument in favor of He. > F(S) meeting this criterion is that F(S) > He. is demonstrable in HU theory, which can not be so well-demonstrated elsewhere from U (perhaps notwithstanding Ma. > He.): Cf. HF(S) aktī́, iskʰús, laĩf͔os, lían, nákē, páskos, síka. This implies a direct line of contact between early F(S) and He. speakers – likely via the Dnieper river.

I do not attempt to connect the Sami ethnonym to this proposal, at least until a viable phonological environment can be replicated through further HU proposals.

Mukʰá: Reviewing the Finngreek Landscape

This post focuses on reconstructed geographical features of Helleno-Uralic contact, which would have primarily taken place in northeast Europe, through the lands known as Scythia. During Helleno-Uralic contact, the mukʰá (riverbend[s]) of the Scythian rivers would have been the center(s) of trade and travel. It is unclear whether this term was used by the ancient Greeks to describe the rivers themselves – which is possible within Greek – or whether a specific term just for ‘river’ can be reconstructed from terms not currently compared. The ancient Greeks certainly had names for the Scythian rivers, such as the Borysthenes (Dnieper), the Tanais (Don), and the Rha (Volga). There are historical records of Greeks traveling these rivers, and settling upstream among the locals (eg: Gelonus).

In addition to the mukʰá, there were dams, moors, hills, cliffs, and roads. There may have also been swamps. This is a smaller collection of the terms I originally wished to share; the terms not yet shared may offer a much broader description of the landscapes, but they need more in-depth research. Over time, I will expand on this post.

wave (HU)
PU *kumpa <? He. kũma / κῦμᾰ < PH *kūmə < PIE *ḱéwh₁-mn̥ < *ḱewh₁- ‘to swell’
Con.: PU *-p-; PU *kompa? (per Aikio)

If a U < H loan, it must have been an early loan, given its wide distribution, and /u/ vs. /y/. However, PU *-p- is unexplained. Also, I am awaiting more information re: *kompa (vs. the commonplace reconstruction *kumpa). This is already an irregular proposal due to *-p-; and if I am satisfied with the reasoning for *-o-, then I will remove this proposal.

dam, dike? (HEr.)
Er. kavaz / каваз < He. *kʰówos / *χόϜος (cf. χόος, LSJ:FB) < khéō / χέω <1 PIE *ǵʰew- ‘to pour’

A citation for the Erzya term kavaz has been difficult to find: It is plainly listed on Wiktionary as ‘dam, dike’, but is without a source; and it is allegedly defined as ‘dam, pond’ / ‘плотина, пруд’ in the Erzya-language book Сырнень човалят: Стувтовозь ды чуросто вастневиця валт (Russian: Золотые бисеринки: Забытые и редко встречающиеся слова [English: Gold beads: Forgotten and rare words]), by Evgeny Vladimirovich / Евгений Владимирович. As for He. kʰó(w)os / χόος, it is defined as ‘mound, bank of earth’ (cf. χοῦς, LSJ:EW), ‘dike’ (LSJ:E>GW); and described by Hsch. as “χῶμα” = ‘mound, dike, dam’.

pond, puddle, bog (PU); damp, mud, scum (He.); mud? (HU)
PU *lampe <?> He. lámpē / λάμπη < ?
Con.: Lit. klampa ‘marsh, swamp’ (cf. Junttila, p. 282); Geo. loṗo / ლოპო ‘slime?’

It appears that the unifying HU semantic value is ‘mud’: Compare Ma. lop ‘deep-lying moist soil’ / ‘tiefliegender feuchter Boden’ and Ne. limbad ‘sand with water on the riverbank, sand with an admixture of clay’ / ‘вязкий песок с водой на берегу реки, песок с примесью глины’. However, also compare Ma. lop and Geo. loṗo, as well as the He. variant lápē / λάπη. However, the National Parliamentary Library of Georgia defines loṗo as ‘slippery moisture that settles between tree bark and timber in the spring’ / ‘ლიპი სისველე, რომელიც ხის ქერქსა და მერქანს შორის ჩადგება გაზაფხულზე’, which does not describe moist ground as seen in the U, He., and Lit. terms. If the Sy. terms are correctly attributed to PU *lampe, then it could not likely be a B loan. Re: HU, the PU phonology is ideal for a U<H loan; but as the He. term is without etymology, the direction of loaning is unclear. If U>H, it would most likely be from disintegrating PFS – or a FS descendant language.

moss (HFP)
PFP *rekɜ (lichen) <? He. leikʰḗn / λειχήν <1 PIE *leiǵh- ‘to lick’

Cf. Ma. reγeńə. Ma. -ńə ~ -ńćə̑ ~ -ńće is listed as a denominative suffix; and it is unclear how this might (or might not) correspond to the He. -n(a[s]) suffix(es) (cf. λειχήνα ~ λειχήνας). FP *r- < He. *l- is unprecedented; and the primary FP < H form must have been rendered with FP *r-, perhaps due to II influence. If the FP term(s) were in fact loaned from He., it is relevant to note that reflexes are only found in Mari and Permic (i.e. Central Uralic): This might lexically nudge the range of HU contact towards the Volga-Kama interfluve, which would likely rely on travel on the Volga river.

mukʰá(s) ~ *mukʰā́ ~ múkʰē 
bend of a river (PU); bend of the shore, nook, a bay or creek running far inland (He.)
PU *mučka > PF *mutka ?> He. mukʰá / μυχά (cf. μυχός) ~ mukʰás / μυχάς ~ múkʰē / μύχη ~ mukʰḗ / μυχή (thus *múkʰā ~ *mukʰā́) < ?

The language or proto-language from which U > H would have occurred is unclear. The term could not have been inherited from Sy. due to lack of -k-; but the PU and PF forms are too similar for disambiguation. As -tk- does not occur in He., -t- loss might be expected. However, there are examples of intra-U -t- loss in both F (cf. Fi. mukka; McCoy 2017, p. 207) and S (PS *mokkē). However, Fi. mukka is not likely synchronous with HU contact; and the S reflexes exhibit -o-, which has not demonstrably resulted in He. -u- (but compare Cypriot loc. mokʰoĩ / μοχοῖ). Although not listed in the UEW, I also tentatively suggest Ma. mugər / мугыр ‘bend, wind, meander, curve (of river, road)’ as a reflex of PU *mučka – this would open loaning into He. from the FV continuum – although the Ma. -r suffix is unclear.

Fi. ‘peninsula / Halbinsel, spit of land / Landzunge’; He. ‘thread’
Fi. niemi < PF *neemi <? He. nē̃ma / νῆμα < néō / νέω ‘to spin’ << PIE *sneh₁-

Also compare He. nē̃sos / νῆσος ‘island’, of a debated etymology: Either ultimately from PIE *sneh₂- ‘to swim’ via He. néō / νέω ‘to swim’, or from an unknown or Pre-Greek origin. Perhaps the semantics of He. nē̃sos influenced nē̃ma: Spits of land are very common along the northern Black Sea and Sea of Azov; as well as along the southeastern Baltic Sea. In this context, HF nē̃ma would have meant ‘thread of land’.

heath, moor (F); pasture (He.)
PF *nummi < He. nomī́ / νομή << PIE *nom-éh₂ < *nem-
Con.: PIE *nom-éh₂?

If the word can not be attested within U outside of F, it is not likely to have been inherited into F from PIE via U.

hill, mountain (HMor.)
Er. pando / пандо << He. págos / πάγος <1 *peh₂ǵ- ‘to attach’

It is unclear whether HFS págos ‘frost’ and HMor. págos ‘hill, mountain’ were unified, or if they represent separate borrowings: Both meanings are found in He. págos. If unified, they could represent HFV págo(s). The Mor. result -nd- would apparently be regular (cf. PFU *towkɜ > Er. tundo).

psīlḗ ~ *psī́lē
tall (HU)
PU *pide <? He. psīlḗ / ψηλή <1 PIE *h₃ewps-

Aphetic, fem. form of He. hupsīlós / ὑψηλός. The gender might connect with HU fem. nouns, such as HU pʰúē ‘tree’. Assymetric HU -īlḗ is reminiscent of -ḗlǣ: A possibility that medial -η- evolved before final -η is considered.

grass (PUg.); that which grows (He.)
PUg. *pimɜ < He. pʰy̌̃ma / φῦμα << PIE *bʰúH-mn̥

In the Uralisches Etymologisches Wörterbuch, Károly Rédei describes Ug. -u- as secondary (cf. Khanty and Mansi pum < PUg. *pimɜ). However, this is not the outcome of U *(-)imɜ in the Ug. languages (cf. PU *imɜ > Kha. imə ~ imi; PFP? *simɜ > Kha. sɑ̆mĭ ~ sĭm- etc., Man. sämi ~ siməl etc.). Unfortunately, I haven’t found any example of PU *-üm- reflexed in PUg. for a potential regular phonological comparison: In this event, the best explanation I offer is that its origin as a loanword explains the unusual vowel correspondences. In light of this, I suggest the correction of PUg. *pimɜ to *pümɜ ~ HUg. pʰy̌̃ma.

raktós ~ ráktos
ravine (HF)
Fi. rotko <<? He. raktós / ῥακτός ~ ráktos / ῥάκτος <1? PIE *wreh₁ǵ- ‘to break’
Con.: Irr.

Although a perfect semantic match, the phonology is unusual: Fi. *rahto might be expected (cf. HF ah͔tī́); but if Sa. -tk- < U *-kt- is a demonstrable irregularity (cf. PU *tuktɜ >> Sa. totko), perhaps the same could be applied to a F < He. loan. Fi. ro- << He. ra- is unexplained.

rótʰos ~ *róθē
winter road (SaN); mountain path, path, trail (He.)
SaN raŧŧi < PS *rëθē <? He. rótʰos / ῥόθος ~ *róθē / ῥόθη < PreH

Due to HU phonological standards, HS /θ/ is presented as a possibility. However, secondary development is not ruled out. The PS form suggests *rëθē < He. *róθē. However, a feminine variant of He. rótʰos is not attested.

edge (Fi.); chasm (He.)
Fi. särmä <<> He. sárma / σάρμα < saírō / σαίρω < ?

The ultimate origins of both Fi. särmä and He. sárma are unknown, so the direction of loaning can not be ascertained.

Áīma: Refeeling the Finngreek Climate

This post focuses on the description of the climate of the Helleno-Uralic contact period, through the semantics of topical term proposals.

air, weather (S); wind (He.)
PS *ājmō <? He. áīma / ἄημα < ἄημι <1 PIE *h₂weh₁-

The shift PS *-ō < HS -a is unprecedented and unclear: There are no other HS terms involving PS *-ō; and in loans, PS *-ō appears to primarily reflex from either PG *-ō (PS *ājrō < PG *airō) or PF *-o (PS *vājmō < PF *vaimo; PS *arvō < PF *arvo [if not directly from PFU *arwa]). However, there are other phonemes which result in PS *-ō: PS *pārō < Old Norse bára is one example; although it’s unclear how this change could apply to HS áīma over the result of PS *vājmō. Perhaps PS *-ō occurred within PS to avoid homophony with *ājmē ‘needle’ – the expected outcome of He. áīma – which would have already been present in the language. Perhaps PS *ājmō could have also been influenced by *vājmō.

brokʰā́ ~ brokʰḗ ~ brokʰī́; purkaiā́
snowstorm, blizzard, smoke, spray (U); rainstorm (He.); pyre, fire, flame (He.)
PFP ~ PU? *purkɜ ~ PU ~ PFU *purka ~ *purki (pp. 25-27) < He. brokʰḗ / βροχή <1 PIE *mergʰ-?; <? He. purkaiā́ / πυρκαϊά < pũr / πῦρ + kaíō / καίω << PIE *péh₂wr̥ + *keh₂w-?
Con.: Nostratic *burV ‘storm’, but Lyle Campbell considers the absence of -ki/-kV unjustifiable (p. 132).

The U variation *-a ~ *-i might be compared to the dialectal variation and/or chronological progression of He. -ā ~ (-ǣ) ~ -ē ~ -ī. U CVLC- < H CLVC- is a common feature of HU proposals, due to Uralic restrictions on initial consonant clusters. A phonological and semantic puzzle arises when considering He. purkaiā́ re: U ‘smoke’: It is unclear whether He. brokʰā́ and purkaiā́ could have both influenced the outcome of U *purkV, which would ultimately depend on whether the original meaning was ‘(snow)storm’ or ‘smoke’. Aikio describes the semantic bridge as ‘pouring, whirling in the air’. He. purkaiā́ is included as material for consideration; but brokʰā́ ~ brokʰī́ > U *purka ~ *purki is considered regular, while there is no precedent from which to analyze -iā́ loss from He. purkaiā́ into Uralic – nor the Uralic variation *-a ~ *-i.

hot, red hot, fever (PFV); heat, burning heat, fever heat (He.)
PFV *kūma < He. kaũma / καῦμα

U < H loans appear to lose their initial vowel in HU -aũ-/-eũ- diphthongs – and possibly in other environments – although the nature of this process is unclear due to few comparable proposals. In this proposal, the original He. term would have been adjectivized after loaning into Finno-Volgaic.

cold, hot (PFS); frost (PF *pakkainen); frost (He.)
PFS *pakka- ~ PS *pākkës < He. págos / πάγος <1 PIE *peh₂ǵ-

In HS, PS *-ës < He. -os / -ος. As with HFV kaũma, HFS págos is a noun that has been adjectivized in either PFS or PS: I suppose the latter, as PF seems to preserve the original meaning. The suffixation of PF *pakka- with *-inen is like *hepoinen ‘horse’, derived from *hepo(i?) ‘horse’ (cf. Fi. hepo ‘horse’). It is curious that, in their combining forms, Fi. pakkas- (eg: pakkaskausi ‘period of frost’) and hevos- (eg: hevosvoima ‘horsepower’) resemble He. -os (i.e. págos / πάγος, híppos / ἵππος) – although this is more likely a coincidence, given the overwhelming regularity of He -os >> Fi. -o/u∅.

sky, god (PU); air, spirit, spiritual being (He.)
PU *nu-mɜ <? He. pneũma / πνεῦμα << PIE *pnéwmn̥ ‘breath’
Con.: PIE *pnéwmn̥

Although phonologically and semantically suitable, chronology is an issue: Loaning into Proto-Uralic is considered permissible in the context of “Pseudo-Proto-Uralic” (i.e. loaning across a a wide geolinguistic span), but PU *nu-mɜ only reflexes in the Ugro-Samoyedic languages, while some presence in the West Uralic languages would be expected from a Helleno-Uralic loan. Although only reconstructed with the meaning ‘breath’, PIE *pnéwmn̥ includes the Albanian reflex frymë ‘breath, wind, spirit’, which might suggest that the semantics ‘sky/air/wind’ and ‘spirit’ could be reconstructed back to PIE (if the Alb. semantics were not the result of secondary contact with He., given their long history of lexical exchange).

storm, wind (U); stormwind (He.)
PFP ~ PFU? *tule ~ *tuuli ~ *tuulə̑ < He. tʰúellā / θύελλα < He. tʰúō / θύω + áella / ἄελλα < PH *aweľľā / *ἀϜελ-ι̯α < PIE *h₂eu-el- ~ *h₂ewh₁eleh₂ ~ *h₂uh₁-l-yeh₂

In his 1991 work “Uralische Evidenz für die Laryngaltheorie”, Jorma Koivulehto compares PFP *tuxli with PIE *dʰuh₂-li-s < *dʰewh₂- ‘smoke, mist, haze’, on the basis of the laryngeal parallel PFP *-x- = PIE *h₂. However, the UEW presents *tule; Ante Aikio presents *tuuli (p. 229); and Petri Kallio presents *tuulə̑ (p. 166). The “long-vowel; *x ~ *w” problem is reminiscent of the varying reconstructions of PU *puwe ~ *puxi > PF *puu; PU *luwe ~ *luxi > PF *luu; and PU *śuwe ~ *śuxi > PF *suu. Re: the phonetic value of the final vowel, a H > U loan can be interpreted regardless: He. tʰú(w)ellā > U *tuulə̑; or HFP *tʰú(w)ellē (cf. θυέλλη) > PFP *tule ~ *tuuli.

The He. semantics are preferred in comparison with PFP *tuuli vs. PIE *dʰuh₂-li-s, as the PFP form’s reflexes only show ‘wind, weather, storm, breeze’, while the PIE form’s reflexes show none of these meanings. Although not listed as a primary meaning in the UEW, I reconstruct ‘storm’ back to PFU due to its presence in Finnic, Mari, and Ugric (exclusively so in the latter two, according to UEW).

mist, fog (F); rain (He.)
Fi. usma << He. húsma / ὕσμα

A variant of Fi. usva, I consider usma to represent an original value -m-. Although written as -s- in He. (h)úsma, the pronunciation would have originally been /z/: This is reflected in HMa. proposals (eg: sarkazmós). As there was no /z/ in Proto-Finnic, /s/ would have been the closest approximation available in its phonology. It is unclear whether He. h- would have been inherited into the F(S?) term.

Noītā́: Reviving the Finngreek Shaman

As my research on Helleno-Uralic theory continues, I have pinpointed two essential areas of focus for the identification of the time, place, and genre of Helleno-Uralic contact: A material culture, and an immaterial culture. This post addresses the latter, in its current state of research.

I draw on the research of Ante Aikio in his paper Proto-Uralic (pp. 47-48) for the comparison of topical lexemes regarding Uralic immaterial culture; and also offer terms of my own, to contribute to the illustration of Helleno-Uralic spiritualism.

First, on the term “shaman”: In the Helleno-Uralic sense, I define a shaman as ‘a spiritual figure who, in an altered state of consciousness, communicates with the metaphysical realm, especially the deceased, in order to tell the future’. This definition is based on the semantics of proposed terms which are relevant to Hellenic and/or Uralic mystic rites.

Second, on the scope of Helleno-Uralic immaterial culture: Helleno-Uralic, as I see it, is a specific convergence and syncretism between Hellenic and Uralic peoples, who otherwise originated from mutually exclusive immaterial cultures. One will not find significant parallels in the foundational Hellenic and Uralic mythologems, which rather formed from the spacetimes of their respective histories prior to Helleno-Uralic contact. Helleno-Uralic religion is rather based on the aforementioned “shaman”, who in a frenzied state, communicated between the physical and imaginary realms. This imaginary ability would have been exalted above all other forms of spiritualism, much as “divine madness” was revered by the ancient Greeks.

The primary Greek god associated with Helleno-Uralic immaterial culture would have been Apollo (Phoibos); and Dionysus, Gaia, and Artemis may have been significant as well. In this post, I will discuss Apollo centrally, and Dionysus peripherally: In this sense, the Helleno-Uralic religion was probably related to (and synchronous with) Orphism, the Oracle of Delphi, and perhaps the Eleusinian Mysteries – although it’s unclear in what manner these practices corresponded to each other, and to Helleno-Uralic shamanism. The underlying theme, however, is a focus on the realm of the dead: And an altered state of consciousness in order to connect with it.

Likewise, Helleno-Uralic shamanism was not necessarily representative of all forms of Uralic shamanism. The Samoyedic peoples, for example, have a complex shamanistic culture consisting of roles and terms I do not (at least currently) reconstruct in the Helleno-Uralic contact period. Even in Sami shamanism, with which I draw the most parallels to Helleno-Uralic shamanism, there are various aspects I can not yet, and possibly will never be able to, reconstruct with confidence.

Regardless, I find it evident that, somewhere between the Uralic and Hellenic spiritualities, a syncretic immaterial culture was formed, which centered on worship of the sun, divine madness, and shamanic necromancy. That having been said, I will now discuss some of the terms on which I have based this claim. The first two are relevant to Aikio’s description of the Proto-Uralic immaterial culture (PU *nojta and *jada-), and the other seven are based on my own research. These are not all the proposed terms I correlate with the Helleno-Uralic immaterial culture, but represent a core selection.

(Note: These proposals are based on a phonologically synchronous orthography, which I devised to represent the values I reason should be attributed to the terms. I will discuss this in much greater detail in an upcoming post titled Phonology: But for now, it is to be understood that the Latin script I use is not identical to the transliterations of ancient Greek, although they represent the same phonemes. Also: The labels Proto-Uralic, Proto-Finno-Ugric, etc. [abbreviated at PU, PFU, etc.] are not indicative of the chronology of these stages of the proto-languages when in the context of Helleno-Uralic proposals: These terms are rather proposed to have been loaned between Hellenic and Uralic at a later date, c. 700 BC onward.)


shaman (PFU); seer (He.)
PFU *nojta < He. noītā́(s)* / νοητά(ς) ~ noītḗs / νοητής << ?

I define the meaning of He. νοητής ‘seer’ through the analysis of its source verb noéō / νοέω ‘to observe, perceive, think, conceive’. GM defines νοητής as “αυτός που σκέπτεται, που συλλαμβάνει τα πάντα με τον νου, δηλ. ο Θεός.” = ‘he who thinks, who grasps everything with the mind, i.e. God’. Νοητής is ultimately from nóos / νόος ‘mind, sense, intellect’, of an unknown etymology. In his book Hermetica, Walter Scott defined νοητός θεός as a “soul without body… direct from God.” (Scott 1985, p. 116). The Helleno-Uralic term noītā́ is not connected to an attested historical role in ancient Greece (notwithstanding the νοητάρχης); and its Uralic reflexes are only found in the Finno-Samic and Ugric languages (cf. HFU séhō). However, I roughly equate noītā́ with He. pʰoiv͔ás / φοιβάς ‘priestess of Phoebus, inspired woman, prophetess’ (cf. HF pʰoiv͔ǣtḗr, discussed below), although noītā́ may have just as well been a role originally exclusive to Uralic, and subsequently retitled with a He. term from a people whose closest approximation to the Uralic shamanic tradition would have been roles relevant to Apollo.

In Helleno-Uralic is also the term noerā́ ~ *nógera* (Fi. nokkela ‘clever’ << He. νοερά ‘intellectual, spiritual, wise’; also an epithet of Apollo): This would make Fi. noita and nokkela derived from the same no- root (i.e. νόος). Although the He. etymology is unknown, I propose that U < H loaning occurred, due to the absence of a more basal root in Uralic. This also agrees with adjectival F -lV < H -rV.

*The phonology of HU noītā́s requires a brief discussion on the dialects of ancient Greek. It is difficult to derive PU *nojta from the Attic form noītḗs, because -ā́- is taken as the original He. value; and corresponds to PU and F *-a > PS *-ē, as seen throughout Helleno-Uralic proposals. However, derivation from Doric would also be problematic, in that the form *noātā́s might instead be expected (cf. Att. γοητής = Dor. γοατάς < γοάω; Att. βοητής = Dor. βοᾶτις [with fem. suffix] < βοάω). Re: noītā́s, the phonology does not seem as plainly bound: < νοέω (vs. -άω); and the adjectival νοατός listed as “hyper-Doric” for Att. νοητός, which describes a hypercorrection rather than a natural formation (although I do not assume this as universal: There may have very well been He. dialects where νοατάς was natural). I hope to eventually identify the He. dialect(s) specific to HU contact if enough attestations can be relied upon – but until then, I can only assert that the He. form, as loaned into “PU”, must have been noītā́(s): So noētā́s would be suitable, but not noātā́s. He. -οη- always corresponds to Uralic <oj>/<oi>. Ancient Greek consisted of myriad sub-dialects, which complicates identification. In addition, it may very well be that PU *nojta was adopted from the voc. He. νοητά, since Sa. -s (which is demonstrable in HU) is absent; and in light of kóptēs. Vocative is rare in HU, but might be suitable in the context of an honorary title (cf. Latin Kyrie < He. Κύριε, voc. of Κύριος) – although this may rather contradict Erzya čirjaz < He. κύριος. Assuming loaning from voc. case would remove the need for a dialectal quest (and explain -s absence in Sami), but I’m not prepared to make that assumption at this time. On top of that, it is worth consideration that noītā́ could be a nominalization of the fem. adj. νοητά (masc. νοητός).

**This is an example of phonological variation involving a recessive accent, epenthetic HF <g> (cf. νογώ) , and the ambiguity of number and gender re: U -a.

to perform necromancy (HU)
PU *jada- <?> He. iálō / ἰάλω ~ iállō / ἰάλλω

Regarding HU semantic reconstruction:
*jada- ‘to conjure, tell fortunes, shamanize; curse/quarrel?’ (Aikio, p. 48)
ἰάλλω ‘to send forth, assail, send oneself on, flee, fly’
LSJ:GM also offers ἰάλω = προπέμπτω ‘to send forth, conduct, escort, “esp. follow a corpse to the grave”, pursue’. GM does not list examples specific to ἰάλω: But the paralleled semantics ‘to send forth’ and ‘assail/pursue’ stand out between ἰάλλω and προπέμπτω. This might represent an intense metaphysical dialogue between the noītā́ and the spirit.

Etymologizing this proposal is difficult, due to the isolated affinity of He. ἰάλλω with Sanskrit iyarti / इयर्ति < PIE *h₃i-h₃ér-ti. On that note, an important phonological discussion must take place: The value of the unclear Proto-Uralic phoneme *-d- (Aikio, pp. 7-10). I will not completely discuss this topic until Phonology; but in Helleno-Uralic theory, PU *-d- = H *-l- (HU psīlē = PU *pide ; HFU tele- = PFU *täwde ~ *tälke, etc.). However, that is not to say that PU *-d- was /l/. Relative chronology is a problem here; and will have to be saved for later. For now, this is a tentative proposal.

pʰoív͔ǣ ~ pʰoĩv͔os; *paiwán
sun? (PU); Apollo/Phoibos (He.)
PU *päjwä <> He. pʰoív͔ǣ / φοίβη ~ pʰoĩv͔o / φοῖβος <? ; *paiwa(n) / *παῖϜα < Pre-Greek

This is perhaps the most unusual proposal in Helleno-Uralic theory, due to ambiguities throughout the Uralic and Hellenic proto-forms and reflexes. First, I present HU pʰoív͔ǣ as the primary form for two reasons:
1. The expected Finno-Samic suffixal variation of PF *-a/-ä, PS *-ē (>NSa. -i), which indicates a feminine He. noun ending in -ā>-ǣ>-ē>-ī. In these diachronica, F -ä and H -ǣ* are also synchronous with archaic Greece (to be discussed in the upcoming post “Phonology”; this correspondence dates the exchange of pʰoív͔ǣ to c. 700 BCE), which is suitable to the northeastward expansion of the Hellenic colonists to the northern shores of the Black Sea, and the historical rise of Apollonian worship. As a feminine noun, it suits the primarily female association with the Sami sun-deity Beaivi. Furthermore, in this context, the deer is relevant as sacrificed to Beaivi, and sacred to Artemis (φοίβη). The shift U Pai- < H Poi would have taken place, with Uralic front-vowel harmonization due to -ǣ.
2. The HF term pʰoiv͔ǣtḗr / φοιβητήρ ‘prophet’, which I propose as the source of Fi. Päivätär ‘Sun goddess’ (contradicting PF *tüt’är < PBS *duktḗ [I rather etymologize PF < He. tʰýgatēr / θύγατηρ], which I will, once again, discuss in Phonology, in order to remain topical).

However, there are other important considerations: That Apollo was male (although the sun’s gender is evidently somewhat flexible in Sami mythology); and that there was another term relevant to Apollo: Pai(w)án ‘song addressed to Apollo or Artemis, epithet of Apollo, physician-healer’. If Beekes’ ideation is correct, the meaning “who heals illnesses through magic (Apollon)” is also to be attributed, which is relevant to the role of the Sami noaidi as a healer; and would derive παιάν from paíō / παίω ‘to strike, smite’: Compare with *päjwä and *päjä reflexes in Khanty and Sami, semantically ‘thunder, lightning’, which I consider a preferable attribution to PFU *päjä vs. ‘fire’, given its wider geographic distribution. Perhaps the meanings ‘smite’ and ‘thunder, lightning’ could relate παίω with *päjä, although morphological affinity is unestablished. With the He. terms lacking certain etymologies, the direction of loaning is unclear for both HFU pʰoív͔ǣ and *paiwán.

*on He. /æ/: Samuels, B: Vocalic Shifts in Attic-Ionic Greek (pp. 91-98)

(h)iaró ~ íaro
noble (Fi.); holy, sacred, (under) divine (protection), glorious, super-human, supernatural, imperial (He.)
Fi. jalo < He. (h)iarós / ἱαρός ~ íaros / ἴαρος<< PIE *ish₁ros

Two points can be inferred from this proposal: That it was loaned from Doric or Northern He. (although this is not certain: F -al- < H -o/eL- is a topic to discuss in Phonology); and that it was used to describe a masculine or neuter noun (eg: It could refer to Phoibos or Apollo, but not Phoibe or Artemis). The Aeolic form íaros / ἴαρος is noted for its recessive accent and smooth breathing, although neither feature is necessarily required in HU proposals.

kópt– ~ kóptēs
shaman’s drum (Sa.); *’striker, beater, knocker, mourner’ (He.)
Sa. goavddis, gobdes, koptês <<? He. kóptēs / κόπτης < kóptō / κόπτω <1? PIE *kop- ‘strike, smite, hew’

I tentatively suggest a direct loaning into Sami from He. κόπτης, with a reconstructed meaning *’striker, beater, knocker, mourner’ (cf. σκώπτω ‘to scoff’ > σκώπτης ‘scoffer’, λάπτω ‘to slurp’ > λάπτης ‘slurper’) – compare English terms such as toaster and computer – but this runs into a problem: The active vs. passive connotation of HFU kóptēs. In light of HF ennustázdō, one speculative alternate proposal is a Sa. back-formation from He. kóptestʰai / κόπτεσθαι ‘to beat the breast, mourn for the dead’, which might bring into relevance an act directed to the deceased (cf. HU iálō; the death of Orpheus as mentioned in maínō) – however, this example would be otherwise morphologically unprecedented in the HS etymological paradigm. With confidence, I can only reconstruct the HS root kópt-, from which would ultimately be derived the name for the Sami shaman’s drum.

*kǐ̃ke- ~ kǐkeí-
frenzy / frenzied (HF)
Fi. kiihkeä << He. kǐkeíā / κυκεία

Fi. kiihkeä is likely from kiihko, which makes it easy to reconstruct H(F) *kǐ̃kos / *κῦκος, as the suffix in κυκεία is identical to the feminine singular form of the adjectival suffix -eĩos / -εῖος, whence nominal -os can be traced (compare oĩkos / οἶκος and oikeíā / οἰκείᾱ). Semantically, I reconstruct the meaning ‘frenzy’ from He. mĩksis / μῖξις ‘mixing (*of the κυκεών), intercourse’ and tarakʰḗ / ταραχή ‘disturbance, commotion’. This seems to me to be relevant to Aikio’s description of shamanistic “sexual and spiritual excitement” (Aikio, p. 48 re: PU *kixi-; although I do not infer an etymological connection between Fi. kiihk- and PU *kixi-). In this context, HF *kǐ̃ke- (with recessive accent) might be contextually related to the entheogenic drink kykeṓn / κυκεών , with which the root kǐk- is shared.

s͔éhō; séhos ~ séos
mixture (Fi.), inside (of a vessel [Man.]), disorder (Sa.); earthquake, commotion, shaking/stirring (He.)
PFU *seka < He. sé(h)ōs / σέως < seí(h)ō / σείω < PH *tséhō << PIE *tweys- ~ *tu̯ei(s)-
Fi. se(k)os <? He. sé(h)os ~ séos / σέος
Con.: In HF séhos, He. and Fi. -os may be a false equivalence. Fi. seos can not be directly < sé(h)ōs, because Fi. -a << He. -ō. Fi. seos (assuming earlier *sekos as seen with teos < *tekos) must either be directly < He. sé(h)os, or be a native Fi. -os suffix unconnected with He. -os.

The HU value (/ts/) can not be reconstructed in PFU, but is expected because of PH *ts-; and medial PFU *-k- < H -h-. UEW describes PFU *seka as semantically and existentially uncertain, due to only being found in FS and Mansi with divergent meanings. However, the attested reflexing of a FU term only in FS and Man. is not unprecedented: This is paralleled in PFU *nojta = HFU noītā́. In light of this – although it could be a mere coincidence – I consider at least the remote possibility that two or more HFU noītā́ shamans were part of Helleno-Uralic contact, perhaps one in a western and one in an eastern zone (i.e. Finno-Samic and Ugric). This hypothesis will require more research.

In this context, my speculation for the semantics of HFU (t)sé(h)o(s) is ‘a mixture or shaking/stirring inside a vessel’. My inclination is to broach the nature of a drink such as the aforementioned κυκεών (cf. HF kǐ̃ke-). I do not reconstruct the meaning ‘earthquake’, due to the Uralic geographical range* of HU contact not being seismically active; and the absence of such a meaning in U reflexes (Sa. ‘disorder’ is also a contextually limited meaning).

*I will discuss the theoretical locations of Helleno-Uralic contact in another post.

to praise (PS); to rave (He.)
Fi. mainoa << PS *mājnōtēk < He. maínō / μαίνω ~ maíno- / μαίνομαι <1 PIE *men- ‘to be excited’

I assume Fi. << PS as it’s been stated in the Wiktionary article for Fi. mainos, although no source is provided. I reject the relation of Fi. mainoa with mainita < PG *mainijaną < PIE *mey-: The phonological similarity of Fi. main-oa and main-ita is rather a coincidence, as the He. semantic and contextual similarities (cf. Maenad < μαινάς) denote the relevance of Dionysianesque worship in the HU immaterial culture. The act of attaining a state of divine “ecstatic frenzy” through intoxication and music is very similar to the shamanic template: Compare the artwork of this ancient Greek mixing bowl, depicting a Maenad beating a hand drum in the company of Dionysus and a satyr (I also tentatively compare Nganasan satərə / сатәрә ‘arctic’ fox <?> He. sáturos / σάτυρος ‘a satyr‘ < Pre-Greek). This mixing bowl also depicts the death of Orpheus: A central theme in Orphism; and relevant to death in HU shamanism.

to predict the future (Fi.); “para someterse a la oniromancia” = to undergo oneiromancy (He.)
Fi. ennustaa < He. ennustázdō / ἐννυστάζω

This proposal, if tenable, would invalidate the Fi. etymology “ennus + -taa” (thus rejecting “ennus < [probably] ensi- + -us”; although ensi is in fact in a HF proposal, which I’ll discuss in a more relevant post), instead etymologizing ennus as a back-formation from ennustaa < He. ἐννυστάζω. The semantic value of ennustázdō is thus ‘to divine the future through the interpretation of dreams’. This is central to the role of the Pythia (cf. egkoimisi) of the Oracle of Delphi, and perhaps relevant to HF nukʰeúō (Fi. nukkua ‘to sleep’ < He. νυχεύω ‘to pass the night’).


I would like to conclude this post in reiteration that I do not equate Uralic shamanism with Helleno-Uralic shamanism. In Aikio’s description of the Proto-Uralic immaterial culture – for which he describes evidence as “very limited” – I do not have proposals to compare with Proto-Uralic *wajŋi, *lewli(w), *eśi / iśi / ićći (I may have a comparison for this term, but not one I’m yet confident to reconstruct), or *kixi-. Furthermore, PU *pi̬ŋka evidences the well-established contact between Uralic and Indo-Iranian peoples: So when Hellenic would have entered the horizon of the Uralic immaterial culture(s), shamanism was already present for millennia. The same can not be said of ancient Greek religion, where any practices comparable to (although not necessarily equivalent with) shamanism are generally limited to specific locations (eg: Delphi) and figures (eg: Orpheus, the Maenads, and the Pythia).

Helleno-Uralic contact and religion were esoteric. The terms that I propose were exchanged are topical to isolated material and immaterial cultures, which were intentionally syncretized; and lexically influential on the Hellenic and Uralic languages due to their status vs. population size. I could go on about this subject: The role of the Pythia in the establishment of Greek colonies on the northern Black Sea, the adoption of Apollo as the patron deity of those colonies; the parallels between archaic Greek religion and the northern Eurasian locale (such as the Ceryneian Hind) – but in the spirit of the noītā́, I’d rather end with some music. Kǐ̃dos pollʲón.

Kiitos Paljon = Kǐ̃dos Pollʲón = Κῦδος Πολλόν

So much done, yet so much to do. I am now proposing new terms on Reddit on a daily basis – we’ll see how long that can last.

I have been working on Helleno-Uralic theory for almost two years now. When I started, I had no idea what I was doing, with only maniacal enthusiasm to guide me in lieu of an education in historical linguistics. The current form of Finngreek is actually the third version I created (plus a recent, important orthographical revision): The other two versions now a distant memory. I look back at words and ideas I proposed two years ago with self-pity – and yet somehow, through every flaw, I am at a point now where I am more confident than ever to share my research with the world.

The main project I am working on now is my first academic paper. I hope for it to be complete before the year’s end. In this paper, I will divulge the anthropological and linguistic reasonings behind my theory. It will be an expansive – but hopefully succinct – foray into my current state of mind regarding Helleno-Uralic contact.

I have removed virtually all content from this website, because it is terribly outdated. The “About” page also requires extensive revision: I will probably rewrite the entire page from scratch, once I have finished my current studies regarding the background information behind my proposed contact period(s).

When I had started Finngreek – back when it was not Helleno-Uralic theory but “Helleno-Finnic theory” – my inspiration was to create an auxiliary language through which Finnish and Greek speakers could easily communicate, helping to re-cement a bond I had felt existed since 3,000 years ago.

More than ever, I continue to believe that around 2,700 years ago, a Hellenic adstrate in the Uralic language continuum resulted in a considerable amount of lexical affinity between (ancient) Greek and Finnish, Estonian, Sami, Hungarian, Khanty, Mansi, Mari, Moksha, Erzya, Udmurt, Komi, Nenets, Nganasan, Selkup, and so on. However, the more new proposals that have come into play, the more painfully evident it has been made that a mutually intelligible auxlang for any single Uralic language with Greek is simply out of the question. Even just within Greek, it is impossible to achieve mutual intelligibility with Finngreek, due to the considerable amount of obscure and obsolete Hellenic terms involved.

On the bright side, this inevitable truth has allowed for Finngreek to become a unique and abundant – if not grammatically stunted – (re)constructed language, offering great insight into the cross-cultural basis for Helleno-Uralic theory. While there are still various problems requiring attention (eg: Differentiating between Hellenic and Proto-Indo-European in ambiguous contexts), I have finally established a reliable phonological paradigm for the evaluation of my proposals: And through this, I am steadily nearing the moment when I can publish my research to share with the world. Once my paper is finally released, I will be able to contribute more time to blogging here, and creating educational and artistic content on Youtube and other platforms for the discussion and celebration of Finngreek and Helleno-Uralic theory.

I’d like to wish you all onḗn kǐkeíān (frenzied fortune) and kǐ̃dos pollʲón (much praise)!