Slavic *tъrgъ, Old Church Slavonic trъgъ. Their Origin and Distribution in Postclassical Times

This work was supported by the Romanian Ministry of Education and Research. Competing interests: no competing interests have been declared. Publisher: Institute of Slavic Studies PAS. This is an Open Access article distributed under the terms of the Creative Commons Attribution 3.0 PL License (creativecommons.org/licenses/by/3.0/pl/), which permits redistribution, commercial and non­ ­commercial, provided that the article is properly cited. © The Author(s) 2015. Sorin Paliga University of Bucharest


Their Origin and Distribution in Postclassical Times
The Basic Data When analyzing in deeper details the origin, evolution and distribution of Slavic tъrgъ, some basic problems are encountered: -the origin of the term; -its later evolution; -its distribution in vocabulary and placenames.I shall try to show that, while some related forms are quite clear, there are many others, which would require more attention.
One first point is that, according to most, if not all, investigations, the term *tъrgъ seems very old.How old?
Analyses of the topic, due to linguists like Machek, Skok and Rejzek indicate a possible relation with AssyroBabylonian tamgaru, a hypothesis suggested by Bedřich Hrozný.To us, this approximate similarity is due to chance, rather than a result of a loanword or an initial relationship.Nevertheless, the term is indeed old, as Skok suggests, labelling it as 'Mediterranean': "Postavlja se pitanje kako, kojim putem su Praslaveni došli do te mediteranske riječi" (Skok 1950; A question arises: how and by what means did the ProtoSlavs get hold of this Mediterranean word?).
Note that Petar Skok is the author of an outstanding book dedicated to the classification and analysis of the archaic PreSlavic placenames of the Adriatic Islands (Skok, 1950), in which he presented his seven principles of toponymical analysis and also the stratification of the PreSlavic linguistic heritage, in which the 'Mediterranean' (i.e.PreIndoEuropean) is reckoned to be the oldest1.
In Skok's terminology, also adopted by many other linguists, 'Mediter ranean' means, in most cases, 'PreIndoEuropean', and refers to a linguistic stratum spread over southern Europe, at least, prior to the arrival of the IndoEuropeans.Machek and Skok were indeed linguists who often referred to this stratum in their studies.In Skok's analysis of the Adriatic toponymy, and in his classification, the 'Mediterranean' stratum plays, therefore, an important role.I analysed Skok's contribution elsewhere and in repeated contexts, I will not insist here (Paliga, 2006b and the updated version Paliga, 2013, the chapter dedicated to placenames).
According to current data, most linguists have agreed upon the basic hypothesis that *tъrgъ is an old, 'technical' term, probably related to Illyrian placename Tergitio (Russu, 1969, p. 254), and its modern related form Trieste, Slovene Trst, explained from a Latin form Tergeste. Albanian treg is considered a loanword from SouthSlavic by Orel (see a longer discussion there), even if this may be debatable in a forensic analysis.2 The forms therefore show a certain distribution in antiquity, with attested forms in the Illyrian area, which may reflect chance, rather than indicate that Illyrian was the source of the term.
Nevertheless, as most linguists are inclined to acknowledge, the term is old and, perhaps from prehistoric times, refers to a 'market place'.This is in full accordance with archaeological data, which indicate southeast Europe as one of the first beneficiaries of the 'Neolithic Revolution', which started in the 1 Our colleague Corinna Leschber presented an interesting paper on the topic of the linguistic stratification of southeast Europe, so this paper may be also read in connection with her presentation, which in fact complements and enriches older data.
2 During the discussions in Toruń (2013), our colleague Xhelal Ylli was firm in denying the Slavic origin of Albanian terg, which is in accordance with the data presented here.See the case of the Baltic forms below.
Fertile Crescent, and then spread east and west, reaching the Balkans soon after the 8th mill.B.C.E.If the term is to be considered 'Mediterranean', as Skok believes, then this hypothesis is supported by archaeological data as well.
If it is old, and there is no doubt it is, Skok's question may be invoked again: how did the term spread?and by which route?
Before attempting an answer, let us briefly review some other possibly related forms.
Trògir, Italian Trau.Ancient spelling Tragurium, Gr.Tragourion.The origin must again be Illyrian, like Trieste, Trst.See in Paliga, 2006a, p. 287.Polish Toruń, as far as I know remains unexplained in details, which may be mutually related to Finnish Turku, genitive form Turun (e.g.Turun Yliopisto 'University of Turku').The alternative is to explain it in connection with Bulg.Tărnovo, Czech Trnava etc. Baltic (Lith.and Latvian) form turgùs and tirgus, respectively, are held to be Slavic loanwords, which may again be debatable (see the case of Albanian form).
Romanian tîrg, târg is also held to be a Slavic loanword, which may or may not be debatable, depending on the ultimate solution agreed upon regarding the origin of the word.If it is a 'Mediterranean', archaic word, as Skok and Machek are inclined to accept, then the term is rather from the substratum language, i.e. a common Illyrian (where it is attested in placenames) and Thra cian form, even if there is no clear Thracian form to be assigned to this group.The list in Dečev (1957) does not quote any such form, but further research may be positive in this sense.
Romanian târg, tîrg is the source of PN Tîrgoviște,Târgoviște, which shows a similarity with the Bulgarian derivation, which is entirely normal.

The Origin
There are several ways of analyzing the origin of the Slavic form *tъrgъ.The possible scenarios are, in our view, the following: 1.A 'Mediterranean', i.e.PreIndoEuropean form.This would fully comply with the analysis of Skok, Rostaing and Paliga regarding the PreIE heritage of Europe.The word may possibly be related to the Sumerian form, but this would be-so far-the only word of this origin analyzable in a Euro pean language.Is this analysis acceptable?If the 'Mediterranean' substratum of southeast Europe may possibly be related to the Sumerian heritage is an interesting topic of discussion, but this is not directly relevant to our more modest purpose here.Such a relationship may be further analyzed in a larger context, aiming at identifying a more consistent set of such possibly related forms.This is a task for the future.
If this term related to a basic activity of trade and commerce is indeed of an archaic, presumably 'Mediterranean' (PreIndoEuropean) origin, then its distribution in southeast Europe, later all over the Slavic area, also in Baltic and Finnish, brings ample problems of interpretation, well noted by Machek and Skok who, by accepting an archaic origin, did not venture to further explain its gradual distribution.I assume that the term spread from south to north, prior to the Slavic expansion, which began some time before 550 C.E.Only an early circulation of the form could explain its PanSlavic situation, and its circulation further north to Baltic and even Finnish.
The term seems only attested in Illyrian, but not in Thracian-at least according to available data and their interpretation, which may be mere chance.If Illyrian is assumed to have been the focus of the later distribution of the term, then the hypothesis that it spread to the Thracian speakers first, hence to the ProtoSlavic (or BaltoSlavic) speakers, must be accepted.It is difficult to assume the scenario that the term could circulate directly from Illyrian to ProtoSlavic, as the two groups never met-also noting that Illyrian became extinct in the 2 nd century C.E.3 The alternative would be to consider the term spread to the whole Illyrian and Thracian world of Late Antiquity as a 'technical', specific term referring to trade.In this perspective, the distribution of this term to Slavic becomes easier to explain.As shown in our previous studies, there was a late Thracian, presumably Carpian and Costobocian influence on ProtoSlavic.In our view, and starting from Aleksandar Loma's analysis regarding ProtoSlavic, THREE satem strata concurring with ProtoSlavic (Loma speaks of TWO strata) may be envisaged: a. BaltoSlavic; b.Iranic; c. North Thracian, specifically Carpian, i.e. late Thracian still spoken until, perhaps, the 5th century C.E.
If our interpretation, based on Loma's scheme, and extended to THREE, not TWO satem4 components of ProtoSlavic is true, then the distribution of *tъrgъ is interpretable as a 'technical', specific term referring to trade, which began to circulate BEFORE the Slavic expansion, i.e. before mid6th century C.E.I may continue by asserting that the case of *tъrgъ should be discussed together with other forms, e.g.*sъto, to which a full analysis was dedicated a long time ago (see also the criticism of Marko Snoj, followed by an answer in a polemical note: Paliga, 2012b); 2. An old PreSlavic origin, but following another route.If an Oriental (AssyroBabylonian?Sumerian?)ultimate origin is to be accepted, I have no expla nation, my competence stops here.However I would add that this would be the only word of this origin identifiable in Slavic or in a European language in general.
3. An IndoEuropean origin.The problem is that the different forms are all isolated, not allowing a deeper analysis in the field of IE comparative linguistics.Slavic *tъrgъ is isolated in relation to other IE languages for which no direct borrow ing via the suggested route can be supposed.As a consequence, a PIE origin may be dismissed as no other form outside the quoted areas could possibly be invoked.

Interim Conclusions
Our interpretation is that *tъrgъ is indeed an archaic, 'Mediterranean' (i.e.PreIndoEuropean) term specific to southeast Europe, and corresponding to the spread of the 'Neolithic Revolution'.If a relationship with Ass.Bab.tamgaru is ever to be accepted, then it may be interpreted in the light of such archaic cul tural and linguistic developments.As the Neolithic revolution emerged in, and then spread from, the Fertile Crescent both west and east, then the relationship of such terms is not enigmatic.Nevertheless this is another direction of research, which cannot be developed here.If a Sumerian origin of Slavic *tъrgъ might be accepted, then this should be analyzed in a larger context of other similar forms spread in the European languages.This indeed seems a difficult task.To date, it is indeed difficult to accept a unique, isolated borrowing from Sumerian into a European language.A larger context of analysis is required.
I am rather inclined to see here a common Illyrian and Thracian heritage, with a possible AdriaticIllyrian origin as a term specific to trade, initially perhaps related to markets on the shore, then spreading to the Thracian world as well.The term must have spread before the fall of the Roman Empire, as only such an old distribution may explain how it reached the Slavic homeland before the 6th century C.E. just like other forms analyzed elsewhere (for other cases referring e.g. to *sъto and Perunъ (on *sъto see Paliga, 1988 and the recent polemical paper Paliga, 2012b; on Perunъ see Paliga, 2009).
The case of *tъrgъ is NOT, therefore, an isolated case, but an example, relevant indeed, among other examples of the north Thracian component (stratum C) of ProtoSlavic.Further research will either support or not support our interpretation.For a recent analysis of the early Slavic and East Romance relations, see Paliga and Teodor (2009), so far the only interdisciplinary approach published in Romania on this topic; very recently, Paliga, 2012a.Sl. *tъrgъ may be ultimately assigned to the group of words derived from PreIE *T-R-, analyzed in a larger context in an older paper, Paliga, 1989, with an earlier approach to the specific terms related to 'township' and 'fortress' in Thracian (Paliga, 1987), without analyzing the possible forms derived from *T-R-, which may be added in the wake of more recent investigation.

Addendum
The data analyzed above may be better understood in the light of other archaic ('Mediterranean') terms referring to urban settlements.The problem is complex, therefore I will briefly refer to some previous data published by the author, updating some information and making new connections.
The case of Sl. *tъrgъ might be easier understood if placed in a larger context of terms originating in the 'Mediterranean' (i.e.PreIndoEuropean) stratum.I note that a series of archaic terms are already attested for: -Greek asty, of 'Mediterranean' origin and polis of presumed IE origin.
-Latin urbs, urbis related to orbis, of Etruscan origin, which means another possibly Mediterranean origin, or in any case nonIE archaic origin.
Latin developed during the expansion of the Empire a series of 'technical' terms like oppidum, civitas, urbs, orbis and others.
At least an interesting case, which supports our interpretation, may refer to the forms derived from PreIE *OR-, *UR-'big, huge', specifically in some place and mythological names of antiquity, some preserved to modern and contemporary times.I once made a new interpretation of these forms in the context of analyzing Rom.oraș, dial.uraș 'town, city', showing that the currently accepted hypothesis, a Hungarian origin, is erroneous, it does not comply with either phonetic evolution or extralinguistic context: the Magyars could not have a term referring to an urban settlement prior to the 10th century C.E. (see, among others, the situation in Finnish, where kaupunki and kauppala, the two terms referring to the urban settlements, are of Germanic origin-the Finnish groups settles approximately 8 centuries earlier than the Magyars.
In reality, the Romanian forms must be explained in close connection with the Thracian placenames ending in -ora, -oros, -oron, well attested and correctly analyzed in Dečev's Thrakische Sprachreste.
Other forms compete with being accepted in this group, I would mention at least two: -Varna, in our analysis derived from a substratum, Thracian root *war-, *or-, and not derived from a root 'black' as in vrana, which is-in our hypoth esis-an erroneous explanation.Varna witnesses in fact the same phonetic evolution as in Vrbas, which does not reflect Slavic vrba 'willow', but relates to the ancient city of Urpanus.
-Warszawa, unexplained etymologically.It is, in our analysis, from the same PreSlavic root *war-'big, huge' with direct reference to the urban settlements of the time.I am inclined to analyze the placename Warszawa in connection with rivername Warta and with the Bulgarian placename Varna.
In Slavic, the initial v < w is a normal phonetic phenomenon, with later evolutions specific to the Slavic languages.Liquid r in Vrbas is specific to the SerbianCroatian area, but not to Bulgarian and Polish.If Warszawa may be indeed assigned to the PreIndoEuropean linguistic stratum, then future research may well be able to probe some interesting details.The placename Toruń may be also understood in this context.
Further research in the field of archaic PreSlavic placenames may clarify the origin of Varna and Warszawa as well.

Their Origin and Distribution in Postclassical Times
Slavic *tъrgъ, Old Church Slavonic trъgъ, preserved in the modern Slavic languages as well, has had an impressive distribution in both vocabulary and placenames, to note just Bulg.Tărgovište (also an important archaeological site), Rom.Târgoviște, also spelled Tîrgoviște (the political centre of Wallachia for some time, approx.80 kms northwest from Bucharest) and as far as Finnish Turku (gen.Turun).See below the discussion regarding the Polish placename Toruń.The origin has been debated, but it cannot be analysed independently from ancient Illyrian town of Tergitio, later Tergeste, the precursors of modern Slovene Trst, Italian Trieste.The ultimate origin has been looked for even in remote areas like Sumer, e.g.Václav Machek, who quotes AssyroBabylonian tamgaru 'trader', in fact following a suggestion of the orientalist Bedřich Hrozný, the decipherer of Hittite (he published the study in August 1915).
The author assumes that the origin of the word must be accepted as 'Balkanic' or, in a perhaps better phrasing, as a common Illyrian and Thracian 'technical term' referring to trade and commerce.Its spread from south to north is entirely normal, following the spread of economic relations from the Roman, then the Byzantine world northwards at a date difficult to determine, but definitely prior to the Slavic expansion, i.e. before the 6th century C.E.It is unlikely that we have to do here an Oriental term.If indeed that were so, the term should have spread first to Classical Greek, then should have migrated northwards at an earlier date.It is rather likely that we have here a 'Mediterranean', perhaps even a PreIndoEuropean term, in Machek's terminology, 'praevropský původ' (of Old European origin).Keywords: Illyrian language; Thracian language; tъrgъ; IndoEuropean language; PreIndoEuropean language; 'Mediterranean language'; Sumerian language