Medieval Livonian county Metsepole and the historical and linguistic border between Livonians and Estonians.
Fifteen years ago, Enn Tarvel, a famous Estonian historian wrote that there was still an urgent need to study the Livonian settlement history even if "The borders of ethnic territories of the Livonian tribes in the 13th century are indeed very well defined on the basis of the Chronicle of Henry of Livonia, and the documents and records from the 13th to 15th centuries". The results of the [former] research are summarized in the fundamental work by A. Bielenstein (1892). According to Tarvel, nothing new has been written on the settlement history during the past hundred years (1995 : 58).
Most linguists and historians in Estonia and in Latvia agree with Tarvel on that the borders of ancient Livonians are well documented and the border between Estonians and (Metsepole) Livonians coincides approximately with the modern state border between Estonia and Latvia. What is unclear is the Livonian settlement history in Courland.
Encouraged by Tarvel's claim that there is still an urgent need to study the Livonian settlement history, the authors of this paper will try to show that the question of the medieval border between Estonians and Livonians is not yet resolved. Most problems arise while carefully reading the Chronicle of Henry of Livonia (CLH). Surprisingly, one can find modern maps in the literature where the territory of Livonians reached as far as the River Parnu in Estonia in the 13th century (e.g. Minahan 2000 : 424; Christiansen  1997 : XXII). Literature offers different interpretations as to the northern and north-eastern border of Metsepole. The territory of Livonians under discussion covers the southern region of modern Parnumaa--the Coastal Southern Parnumaa with Haademeeste, Tahkuranna, Surju, Paikuse and partially Saarde districts.
To the west, Metsepole bordered on the east coast of the Gulf of Riga, to the east on the northern Baltic Latgalian county Talava, to the southeast on the Livonian Idumea County, and to the south on the Livonian Turaida County with Gauja Livonians. The centre of Metsepole was the Turaida Castle.
The question of the Estonian-Livonian tribal border is at the same time the question of borders between medieval Estonian counties and districts. We have different questions to answer: first, was this part of Parnumaa really Livonian and did it belong to Metsepole and second, if not, did it belong instead to the Sackala County or even to the Soontagana district of the Maritima County?
In the following we will first shortly treat the question of the northern border of the Metsepole County in the CLH and in early treaties in the 13th century. Second, we will examine the historical traditions to see how the question of the medieval status of the coastal region of Southern Parnumaa is resolved. Third, we will offer some linguistic evidence to show that the Parnumaa dialects have many common and archaic features--in pronunciation, lexicon and grammar--with the Livonian language. Finally, different linguistic interpretations will be discussed.
2. The north border of Metsepole in the Chronicle of Henry of Livonia and in two treaties from the 13th century
The Chronicle of Henry of Livonia or the Livonian Chronicle is the primary source of the early medieval history of Estonia and Latvia (Heinrichs Livlandische Chronik 1955; Henricus Lettus 2003). (1) It is mainly from this chronicle that we have knowledge on the ancient Estonian and Livonian tribes, their social organisation, customs, religion, and territories. There are at least fourteen passages in the chronicle where the border between Metsepole Livonians and Estonians is mentioned in the beginning of the 13th century. Here, under the term border we mean not a line on a map, but a situation where somebody moves from one known point to another known point mentioned in the chronicle; the border lies between them.
The first references to the border are made in XIV, 10: "To take vengeance on the nations of Estonians the whole army followed Russin, Caupo, and others, who went ahead of the Rigans and pilgrims into Metsepole. They marched to the sea and, following day and night the direct route along the sea, they came to the first province, which is called Soontagana." Here it is quite interesting that Soontagana is called the first Estonian province, which means that from the direction of Riga, Estonia started beyond the River Parnu. Soontagana is called the first province of Estonia also in XIX, 8. Most passages with reference to the border do not give us enough information to decide on the location of the border, e.g. in the same passage (XIV, 10): "the Estonians who had first escaped came hurriedly from Soontagana and the other near-lying provinces with a great army to Metsepole".
In our context one of the most important places in the chronicle is (XV, 7): "They [Brothers, Livonians,Wenden, and Letts] assembled a great army inMetsepole, marched to the sea, and went for a three-day journey beside the sea. They turned, after this, toward the province of the Sackalians and journeyed for three days through forests and swamps by a very bad road. [--] At length, on the seventh day, they came to the villages. [--] They [--] returned by another road, dividing all the loot equally among themselves. With joy they returned to Livonia." Here the army started its journey somewhere in the heart of Metsepole, marched a three-day journey beside the sea. It means that they reached near the mouth of the River Parnu, but before Soontagana turned towards the direction of Sackala and continued their journey for three more days before they reached the Sackalian villages. The other road probably went from Sackala directly to Lake Burtnieki. From this passage it is possible to infer that the coastal region of Parnumaa did not belong to Sackala at that time.
The chronicle mentions several times how the Teutonic army marched from Sallaca (Metsepole) directly to the Soontagana or Maritime provinces, e.g. XVIII, 5; XIX, 3; XXI, 5; XXII, 9.
Two border treaties are also known from the second part of the 13th century. In 1259 Livonians from Metsepole and Estonians from Soontagana concluded a border treaty (see Berkholz 1886 : 44-46; Perlbach 1886 : 20-23). Both sides agreed that in 1259 Livonians inhabited the territories north of the River Salaca to the River Orwaguge [Orajoe], and the Cape of Laddekeriste [Laigaste nina] would remain the northern border of Metsepole.
The next border contract from 1276, now between the archiepiscopacy and the Livonian Order agreed that the new border was along the rivers Coddeyogge [German Koddiak, Rozeni at the River Salaca] and Hainejecke [Ainazi] (see Berkholz 1886 : 45; Hildebrand 1880 : 376-277).
Half a century earlier, the data from the CLH and the treaties are only prima facie contradictory. It is possible that Soontagana skillfully took control over the territory in question using the chaos resulting from the Livonian Crusade. Passages from the chronicle clearly show that Soontagana was located beyond the River Parnu, since it took several days and nights to march from Salaca to Soontagana.
3. Whom did Coastal Southern Parnumaa belong to?
According to mainstream historians, Coastal Southern Parnumaa belonged to Estonians (see e.g. Strods 1963 : map on p. 7; Jaanits, Laul, Lougas, Tonisson 1982 : 408; Eesti aastal 1200 : maps on pp. 10 and 12; Latvijas vestures atlants 2005 : map on p. 12; Kriiska, Tvauri, Selart, Kibal, Andersen, Pajur 2006 : map on p. 44). The mainstream tradition is based on the influential book "Grenzen des lettischen Volksstammes" by Dr. August Bielenstein (1826-1907), a famous Latvian theologist, linguist, folklorist, and historian who placed the border between Soontagana and Metsepole on the River Orajogi (Bielenstein 1892 : map II). He comments that the north border of Metsepole was beyond the River Salaca, approximately at the border between Estonians and Latvians at the end of the 19th century (Bielenstein 1892 : 59). Bielenstein did not find any evidence from the CLH, because on the one hand he was enthusiastic about the 1259 and 1276 border contracts (see above), and on the other hand he relied on a comment by Hermann Hildebrand on the CLH "north of the River Salaca is Soontagana and in south Metsepole" (Hildebrand 1865 : 65). We can call the mainstream historians representatives of the contractual tradition.
But some authors hold that this territory belonged to Livonians. For example, in a recent historical dictionary of European national groups, James B. Minahan suggests that the territory of Livoninas included the Coastal Southern Parnumaa: "In Estonia, the Livonian homeland includes the coastal region south of the city of Parnu" (Minahan 2000 : 424 and map on the same page). Of the same opinion is Eric Christiansen, a leading scholar in the field of the Northern Crusades. He uses the map of Livonia and Estonia where the Livonians inhabit the territory covering the districts on the West Bank of the River Western Dvina up to Parnu in Estonia (see Figure 1) (Christiansen  1997 : map on p. XXII).
[FIGURE 1 OMITTED]
The tradition of Christiansen and Minahan continues an old tradition that is based on the careful reading of the CLH. We can call it the historiographer tradition. This tradition has a long and prestigious history. A German councilman of French origin, Johann Leonhard von Parrot (1755-1836), who was interested in the culture, history, language, and mythology of the Estonians, Livonians and Latvians, wrote that according to the CLH, Metsepole borders on many Estonian counties or districts (Parrot 1828 : 197). In the north at the River Parnu, Metsepole has its neighbour Soontagana, and in the east there is Sackala (Parrot 1828 : 204, 205, see also Parrot 1839 : map). Parrot gave an interesting but erroneous etymology to the name Metsepole, which according to him means the separation of counties, border of the fatherland, from Celtic met 'separated' and pole 'fatherland' (Parrot 1828 : 197).
In the same year, Estonian Lutheran pastor Heinrich Georg von Jannau (1789-1869) published his famous "Ueber die Grund- und Ursprache der Ehsten". He argued that the forest district Metsepole (Motsepole) extended from the River Salaca and from Lake Beverin [Lake Burtnieki?] up to the River Parnu in Estonia. It is interesting that Jannau, in general, follows the division of Old-Livonia published anonymously in Hupel's "Neue Nordische Miscellaneen" (Anonymous 1792). The anonymous author thought that the border between Livonians and Estonians was at the River Salaca, but on this Jannau disagreed with Anonymous.
A Baltic-German historian, Alexander von Richter (1803-1864) wrote in his history of the German Baltic Sea provinces of the Russian empire: "Livonians inhabited the territory close to Ventspils [Windau, in Courland] and up to Parnu (in Estonia). The Livonian settlement was interrupted by the Baltic people at the harbour of Zemgale [Semgaller Hafen] at the estuary of the River Aa 50 km west of Riga. North of the River Western Dvina lay the territories of Turaida, Idumea and Metsepole" (Richter 1857 : I, 48). If we take a look at the accompanying map we can see that Metsepole is located between the River Salaca and the River Parnu according to Richter (1857 : map).
A famous historian, writer and lawyer, Otto von Rutenberg (1802-1864), in his history of the Baltic Sea provinces Estonia, Livonia and Courland, argued that in the 13th century the Order of the Brothers of the Sword received the province of Sackala, and the bishop got the provinces of Idumea and Metsepole between the River Gauja and Parnu (Rutenberg 1859 : I, 67). On the accompanying map Metsepole is located between the rivers Salaca and Sauga (the Sauga is a small river 10 km north of the river Parnu); to the east Metsepole bordered on the rivers Navesti and Halliste (the Navesti is a tributary of the River Parnu and the Halliste is a tributary of the Navesti) (Rutenberg 1859 : map).
An Estonian journalist and amateur historian Hindrik Prants wrote in his textbook that Soontagana (Korbemaa) reached the River Parnu in the south and bordered on the Land of Metsepole. Metsepole was inhabited with mixed populations where Estonians had contacts with Livonians. Between Metsepole and Sackala there was Zara [Saarde] (Prants 1920 : 49). On the accompanying map issued by the topographical department of the Estonian army, the border of Metsepole is located between Parnu and the Estonian-Latvian border. Zara [Saarde] is erroneously located between the River Parnu and Metsepole (Prants 1920 : map).
The question of the borders between the Livonians and the Estonians is at the same time the question of the borders of Estonian ancient counties Sackala and Rotalia and sometimes Maritima. Soontagana was the southernmost district of Maritima. If we have another look at the contractualists' views, we can see that the Coastal Southern Parnumaa formerly belonged to Soontagana (e.g. Hildebrant 1865 : 65; Bielenstein 1892 : 59). This position was also held by Karl Woldemar von Lowis of Menar (1855-1939), a historical geographer, archaeologist and cartographer. According to him, Soontagana coincided with the Coastal Southern Parnumaa (Lowis of Menar 1895; 1907). A Baltic-German historian, Leonid Arbusow the elder (1848-1912), following Lowis of Menar, places Soontagana between the River Parnu, the Sackala County and Metsepole in Kuppfer's "Baltische Landeskunde" (Arbusow 1911a : 390-391; 1911b : map "Historische Karte von Alt-Livland"; and also in 1908).
[FIGURE 2 OMITTED]
The authors of the academic history of ancient Estonia ("Eesti esiajalugu") take no clear position on the territory in question. It is said that traditionally the coastal forested and swampy territory from Ikla to Parnu is included in the Sackala County, but there is no clear evidence about the affiliation of the territory. The authors claim even that probably there was no need to regulate the administrative relations of this territory because the population density was very low (Jaanits, Laul, Lougas, Tonisson 1982 : 408).
The authors of the academic history of ancient Estonia follow the tradition of Heinrich Laakmann (1892-1955), a scholar in medieval Livonian history and historical cartographer. He wrote that north of the River Salaca there was a great wilderness; from Salaca to the north (up to Parnu) one can find forest areas on the coast of the Gulf of Riga (Laakmann 1939 : 206). On his map "Baltic countries: population in 1200", he left blank the territory between Parnu and Estonian-Latvian border (Metsepole) and between the Gulf of Riga in the West and the Sackala County with the districts Alistegunde and Ruyenia (see Figure 2). The border between the territory in question and Sackala coincides partly with the River Halliste, a tributary of the River Parnu (Laakmann 1954 : map; 1973 : map). Laakmann commented that north of the River Western Dvina there was no real wilderness but rather a no-man-land between the neighbouring tribes. The territories that are shown as inhabited on the map belong to the neighbouring districts, and the usage rights to such territories were mutually delimited (Laakmann 1954 : 7; 1973 : 313). One can find an echo of this "wilderniss" also in a miscellany Estonia at the year 1200 (Eesti aastal 1200 : map on p. 10).
4. Contacts of Salaca Livonian with its neighbouring dialects of Estonian
The data about the vocabulary and grammatical structure of Salaca Livonian were recorded by a number of scholars between 1655 and 1846 (see Winkler 1994; 1999a; 1999b; 2000; 2002; Pajusalu 2007).This corpus is not uniform but nevertheless rather comprehensive. The dictionary of Salaca Livonian (Winkler, Pajusalu 2009) systematically presents lexical and morphological data from all the available sources including around 1,450 stems and 8,500 word forms. More than two thousand sentences and phrases of Salaca Livonian are known altogether from different manuscripts. Examples of A. J. Sjogren's recordings were published in Wiedemann 1861a; 1861b. Considering all these materials it is possible to estimate the similarity between Salaca Livonian and Estonian dialects. The analysis of lexical relations is presented by Pajusalu, Kirkmann, Winkler 2009 (see the article in this volume). In this study we focus on various characteristics of Salaca Livonian regarding the closest Estonian varieties spoken in the north of the historical Salaca region, i.e. western Estonian sub-dialects of Haademeeste and Saarde, and the south-western Mulgi dialect of South Estonian. The common innovations of Livonian and Leivu South Estonian, which was once obviously an eastern neighbouring dialect of Vidzeme Livonian, are analyzed by Tiit-Rein Viitso (see Viitso 2009 in this volume).
The Estonian contact area of Salaca Livonian has been linguistically diverse, and Salaca Livonian also reveals extensive variation. It is possible to distinguish forms with different origins in the Salaca Livonian data (see Pajusalu 2007). Firstly, the occurrence of West Estonian-type forms points to the Estonian-Livonian transitional variety in the north of the River Salaca. It probably emerged in areas of Estonian-Livonian mixed settlement from Salaca to as far as Tahkuranna. Secondly, the exceptional features of Salaca Livonian which resemble Courland Livonian, especially its western dialect, can be remnants of the former Livonian variety of southern Vidzeme. The specific traits and some similarities with Mulgi, Leivu and other southernmost South Estonian dialects could be the remnants of the archaic Metsepole Livonian, which differed both from the southern Vidzeme dialects of Livonian and Courland Livonian. One can claim that the language usage of the last Salaca Livonians constituted a merged variety combining the features of various historical Livonian dialects.
4.1. Similarities between Salaca Livonian and Mulgi South Estonian
The Mulgi region has historically been a center of west-southern Estonia, and was known as the Sackala County already at the end of the prehistoric period. In south-western Sackala, the ancient Halliste-Karksi parish (Alistegunde in CLH) also contained the Ruijena area of present day northern Latvia, and had a close contact with the Metsepole Livonian area. Sackala was densely populated and administratively well-organised already at the beginning of the second millennium when the Livonian settlement was flourishing in the Salaca and Gauja area. Later the Halliste-Karksi parish became part of Parnumaa. The linguistic similarities between Salaca Livonian and Mulgi Estonian could be a result of mutual contacts over a very long period.
Several authors, e.g. Salme Tanning (1958) and Karl Pajusalu (1996 : 56-64) have observed common features between Livonian and the western South Estonian. These features are mostly lexical (cf. Pajusalu, Krikmann 2009), but there are also some remarkable phonetic and morphophonological similarities. It is notable that the majority of Mulgi-like forms occur both in Courland and Salaca Livonian. However, a set of words, phonetic and grammatical forms are common only in Mulgi and Salaca Livonian, such as katik 'broken' (cf. Kur katki), Sal loud and Krk louda 'to find' (cf. Kur lieudo, (2) Sal secondarily lied, leud-), Sal siemil and Mul seemel 'seed' (Kur siemgoz), Sal velen ~ velan ~ valan and Mul vallan (Kur ullo) 'outside', among others.
4.2. Similarities between Salaca Livonian and Haademeeste-Saarde
The region of Haademeeste and Saarde is significantly different from Mulgi. Haademeeste and Saarde do not represent an old settlement and cultural center comparable to the Mulgi region. They form a transitional area in many senses, also linguistically. According to the traditional classification of Estonian dialects, Haademeeste and Saarde belong to the western dialect area of North Estonian; together they constitute the southern group of the western dialect (Pajusalu, Hennoste, Niit, Pall, Viikberg 2009). This classification is based on the grammatical and sound structure of the varieties. According to lexical characteristics, Haademeeste and Saarde are more closely connected to the Mulgi area (see Krikmann, Pajusalu 2000). Besides, the area is intrinsically uneven. It is obvious that the Haademeeste-Saarde region was populated from different directions and therefore it shares features with several Estonian dialects and also with Livonian. In the following, we give an overview of some common features of the Haademeeste-Saarde varieties and Salaca Livonian.
4.2.1. Phonetic similarities
Livonian reveals such central prosodic features of Southern Finnic as the strengthening of the primary stress and the related vowel loss in unstressed non-initial syllables, the emergence of the opposition of the long and overlong quantitities, quantity alternation, and foot isochrony (see Lehiste, Teras, Ernstreits, Lippus, Pajusalu, Tuisk, Viitso 2008). Weakening of the secondary stress and the related loss of unstressed vowels starting from the third syllable is typical of Salaca Livonian, too, e.g. kadutk 'with hands' (< *kadudeka), imistl 'to people (All.)' (< * imistele). An extensive loss of unstressed vowels is also characteristic of Haademeeste and Saarde, e.g. Haa lohkust 'they destroyed' (< * lohkusid), Saa pikukst 'tinies' (< * pikukesed). The late apocope, which appears also in southern Mulgi (see Pajusalu 1996 : 308-328), could result in the collapse of some cases, therefore the allative and the adessive are identical in many declensions in all dialects of this contact area, cf. Sal ommal ~ omal ~ omel, Haa omal 'to oneself, on oneself', Sal suzarel, Saa Hls Krk sosarel 'to sister, at sister'.
Besides vowel loss, Salaca Livonian and its neighbouring dialects of Estonian are characterized by the loss of certain consonants. The glottal spirant h does not appear in native words in either Courland or Salaca Livonian. In those cases where h was marked in Salaca Livonian recordings, it apparently denotes stod or vowel length, e.g. ra ~ raha ~ rah 'money' (Kur ro'). In many dialects of Estonian, the word-initial h has disappeared (see Pajusalu 2009 : 110) and h occurs only after the nucleus of the primary stressed syllable. In Haademeeste, Saarde, and also in southern Mulgi h is dropped sporadically also in the intervocalic position, cf. Sal ma tab and Haa taa 'I want' (< * taha), Sal mie, Saa mee 'man (Gen.)' (< *mehe). An interesting example is the verb laama 'to go (Sup.)' (< * lahema) that is presented in a dictionary compiled by Salomo Heinrich Vestring, a pastor of Parnu at the beginning of the 18th century. Vestring provides some valuable data about the Estonian language in the vicinity of Salaca 300 years ago; he uses a regional label Tackerort, Gudmansbach et Salis 'Tahkuranna, Haademeeste, and Salaca' that specifies the geographic area where the form laan 'I go' was used (Vestring 1998 : 105).
In Salaca Livonian, the labiodental spirant v was vocalized in the wordfinal position after the late apocope, e.g. alu 'bad' (cf. SE halv), jaru 'lake' (Kur jora), saru 'horn' (Kur sora). The vocalization of v is known also in the Estonian contact dialects, e.g. Haa kaju : kao 'well (Nom., Gen.)' (< * kaivo : * kaivon). The change v > u takes place also in the case of syncope, e.g. Sal aruldi ~ arulti 'rarely' (< * harv-), Haa arune 'rare'.
In many words with the * kl-, * kr-cluster, the plosive was vocalized and a secondary long monophthong developed, for example, Sal kal 'neck' (cf. Kur ka'ggol) and Saa, SE kaal; Sal kara 'oats' (cf. Kur ka'ggor) and Saa, SE kaar.
Like Estonian neighbouring dialects and Courland Livonian, Salaca Livonian distinguishes between short and long geminates. This can be proved by the spelling of stop consonants, cf. utab 'takes' (NE votab [vottab]) and utten 'taken'; the single dental stop is marked by d: kadiz 'disappeared' (NE kadus). Accordingly, Salaca Livonian reveals a distinction between the long and overlong degree of quantity and three quantity degrees. In addition to quantity alternation, Salaca Livonian includes grade alternation of consonants where the stop is dropped in the weak grade, e.g. jalg 'foot' : jalad 'feet', tunda 'to feel' : tunab 'feels'. Such alternation is unknown in Courland Livonian but common in Estonian dialects. The commonality between Salaca Livonian and its contact varieties of Estonian appears also in the similar use of the grades. Unlike in Standard Estonian, both Haademeeste-Saarde and Salaca Livonian apply the weak grade in plural forms, cf. Sal jalatk 'with feet (Com.)' and Haa jaladeg. The weak grade was also used with the plural in the case of quantity alternation: Sal and Haa patudest 'from sins (El.)'.
4.2.2. Similarities in grammar
Common innovations become evident also in the grammatical structure of Salaca Livonian and the Haademeeste-Saarde varieties. In the nominal inflection, a proper example is the comitative-instrumental marked with the kending, e.g. Sal jamandek 'with landlady' (cf. Kur jemandoks), Haa emandag(a), pavak 'in a day', Haa paavag(a). The declension of Salaca Livonian does not involve the special n-marked form of the dative which is typical of Courland Livonian. To express the dative meaning the allative-adessive forms with the ending -l were used similarly to Haademeeste-Saarde, such as jumalel 'to god' (Kur jumalon). The use of the inessive in the function of the illative, e.g. kus 'where to', maas 'down' could well be a morphosyntactic influence of Latvian both in Salaca Livonian and its northern contact area, but it can also be related to the apocope of the illative sse-marker and the areal restructuring of the grammatical system.
In Haademeeste-Saarde, some forms of personal pronouns are strikingly similar to Salaca Livonian. For example, in the formation of the plural the de-marker is attached to a singular base also in the case of the forms mede 'our' and tede 'yours (Pl.)', which correspond exactly to Livonian forms mad ~ med and tad. The forms mede and tede have spread in southwestern dialects of Estonian (see Viitso 2008 : 94, figure 3). Among others, the partitive form of the third singular pronoun is the same: Haa, Saa tend ~ tand 'him/her/it', and Sal tand (Kur tanda). The tend-forms have also spread in the southern parts of the western Estonian dialect area, see Figure 3.
In verb inflection, the most noteworthy similarity between Livonian and Haademeeste is connected to an innovative category--that is, the indirect speech which expresses reported eviedentiality. In Courland Livonian and in Haademeeste the category is marked by forms which are similar to the agent noun (see the Livonian data in Viitso 2008 : 325). Such forms have not been earlier described in Salaca Livonian; however, because even Sjogren's texts show it as ji-marked: pateji 'is said to do' (Winkler 1994 : 353). The form shows that it corresponds to the agent noun in -(j)i-suffixed forms in Courland Livonian, cf. Kur teji, and the in -je- ~ -ja-forms in Haademeeste. The form had originally also been the present participle in these languages. However, in southern Parnumaa the Livonian-like indirect forms such as oleje 'is said to be', votja 'is said to take', is known only from the area of the former Tahkuranna manor. In the other part of the Haademeeste parish and in Saarde, the va-ended forms of the present participle are used in this function, e.g. oleva, votva. Once the southern border of the Tahkuranna manor reached the Salaca parish, direct contacts between Livonian and Estonian peasants and also some cases of language shift were likely to have taken place.
[FIGURE 3 OMITTED]
In this contact area, the formation of verb forms reveals, similarly to other southern Finnic languages, various simplifications of the paradigms and related analogical innovations. In Salaca Livonian, the conditional mood reveals, similarly to the neighbouring Estonian dialects, identical ks-marked forms in all persons: oks '(I, you, he/she, we, you, they) would be', panuks 'would put', etc. Similarly analogical is-suffixed forms of the preterite may be used in all persons, e.g. taplis '(I, you, he/she, we, you, they) fought', nuolis 'licked'.
In addition to the imperative, the jussive has been used. The jussive is formed by using the particle las 'let' and the imperative form of the third person: las mina olg 'let me be', las me olg 'let's be'. This pattern is similar to Haademeeste: las ma olga, las me olga, etc.
The formation of negative forms shows a distinction between the present and preterite negative particles, which is also characteristic of Haademeeste-Saarde and South Estonian. In Salaca Livonian, the present negative particle is ab (ap), and the preterite negative particle is iz (is). In Haademeeste, ep and es are used respectively, cf. Sal ap pan 'do not put' and is pan 'did not put'; Haa ep pane, es pane.
As for the tenses, Sjogren presents in addition to the present, preterite, perfect, and past perfect also two future tenses--the future (ma lib kuts ~ minnel lib kuts ~ kutsub ~ ma kutsumis 'I will invite') and the exact future, which is formed by means of the auxiliary verbs sab or lib and the past participle: ma sab tien 'I'll be done', ab li kutsen 'will not be invited'. Nowadays, the auxiliary verb of the future leeb is known only in the insular dialect of Estonian. However, Vestring presented it as a word of the Parnu district at the beginning of the 18th century (see Vestring 1998 : 111).
We can also find some syntactic similarities between Salaca, on the one hand with Livonian, and on the other hand with and the southern Parnumaa dialects, e.g. the genitive object in the imperative: Sal pan suomis lod pal 'put the food on the table', cf. Est. pane sook laua peale and Haa pane soogi laua paal.
4.2.3. Lexical similarities
In the dialectometric study of the relations between Salaca Livonian and Estonian dialects (see Pajusalu, Krikmann, Winkler 2009), which takes into account the stereotypical character of words, that is, the
extent of their spread, Salaca Livonian has had more specific contacts with the southernmost South-Estonian dialects, and it shares the largest number of common features with the Leivu linguistic enclave. There are also a number of lexical commonalities between Salaca Livonian, Mulgi South Estonian, and the southern group of western Estonian, which show close communication over a long period. A number of Salaca Livonian words have equivalents in the southern Parnumaa dialects (altogether 880 common words with Haademeeste, 545 words with Saarde) but the overall lexical similarity with Haademeeste-Saarde is smaller than the similarity between Salaca Livonian and the south-western dialects of South Estonian.
The development of Courland and Salaca Livonian was influenced by various contacts with the neighbouring Estonian dialects. Paul Ariste (Arista 1954) claimed that the insular dialect of Estonian is closest to Livonian, which is apparently true in the case of the Courland dialects of Livonian. The dialects of Courland Livonian and southern Saaremaa are similar mostly with regard to common phonetic developments and lexical borrowings in both directions. However, joint grammatical innovations of Livonian and the insular dialect are almost missing.
The data of Salaca Livonian do not reveal a close relationship with the insular dialect. There are a number of similarities with the western sub-dialects of South Estonian and the usage of southern sub-dialects of western Estonian. In addition to the archaic common features shared by Livonian and these Estonian varieties, one can find a number of joint innovations of Salaca Livonian and its Estonian neighbouring dialects in pronunciation, grammar, and vocabulary.
The common features of Salaca Livonian and Parnumaa Estonian dialects could be divided into three main groups: (1) the features which show the common origin of the language in prehistoric times, e.g. the plural genitive forms of the personal pronouns mede and tede and several words for natural objects, (2) the similarities which are the result of language contacts and adstratum phenomena--most common traits belong to this group, and (3) the phonological and morphosyntactic features that raise the question about the Livonian substratum in southern Parnumaa dialects, e.g. indirect speech expressed by the forms of nomen agentis in Tahkuranna.
Johann Heinrich Rosenplanter (1782-1846), a Baltic-German pastor and scholar wrote in his famous "Beitrage zur genaueren Kenntniss der estnischen Sprache" that "After La Peste [numerous epidemics in the 18th century] many Estonians from other regions inhabited the Parnu County; and one can find very few people whose forefathers lived in this county already before the plague" (Rosenplanter 1816 : 42). This note supports the substratum theory; very few original people remained and newcomers formed the new majority.
Careful reading of the Chronicle of Henry of Livonia supports the possibility that the Livonian Metsepole County reached the River Parnu and at that place shared a border with Soontagana at the beginning of the 13th century. The Sackala County did not reach the sea; to the west, Sackala had a swampy and forested border with Metsepole. The border contract of 1259 determined the border between Metsepole and Sontagana along the line of the Cape of Laigaste and the River Orajogi. The next treaty from 1276 established a new border along the line from Rozeni at the River Salaca to Ainazi.
There are two main traditions in treating the northern border between Livonians and Estonians. Mainstream historians represent the contractual tradition. According to them, the historical border between Livonians and Estonians was along the line of the Cape of Laigaste and the River Orajogi. This tradition has three sub-branches: (1) the Coastal Southern Parnumaa belonged to Soontagana, according to the strict contractualists who interpret word for word the border contract of 1259 between Soontagana and Metsepole; (2) the Coastal Southern Parnumaa belonged to Sackala; (3) the Coastal Southern Parnumaa formed a "wilderniss" with no clear affiliation to the Estonian counties Sackala or Soontagana.
The historians who belong to the historiographer tradition interpret literally the text of the Chronicle of Henry of Livonia. According to them, the Coastal Southern Parnumaa belonged to Metsepole and was inhabited by Livonians. The Livonian substratum phenomena in the coastal dialects of Southern Parnumaa support the historiographer tradition.
Ad.--Adessive-Allative; Com.--Comitative; El.--Elative; Est.--Estonian; Gen.--Genitive; Hls--Estonian sub-dialect of Halliste; Haa--Estonian sub-dialect of Haademeeste; Krk--Estonian sub-dialect of Karksi; Kur--Courland Livonian; Mul--Mulgi South Estonian; NE--North Estonian; Part.--Partitive; Saa--Estonian sub-dialect of Saarde; Sal--Salaca Livonian; SE--South Estonian.
Anonymous 1792, Versuch einer alten Geographie von Livland, zu besser'm Verstand der alten lieflandischen Geschichtsschreiber. Nebst etlichen Anmerkungen zu dieser Geographie.--Neue Nordische Miscellaneen von August Wilhelm Hupel. Erstes und Zweytes Stuck, Riga, 11-134.
Arbusow, L. 1908, Grundriss der Geschichte Liv-, Est- und Kurlands. Mit 1 Karte und 2 Lichtdrucktafeln. Dritte umgearbeitete Auflage, Riga.
--1911a, Geschichtliche Landeskunde (Hierzu die historische Karte des Atlases).--Baltische Landeskunde. Text, Riga, 387-406.
--1911b, Historische Karte von Alt-Livland.--Baltische Landeskunde. Atlas, Riga.
Berkholz, G. 1886, Vermischte Bemerkungen zu der vorstehenden Mittheilung Dr. Perlbach's.--Mittheilungen aus dem Gebiete der Geschichte Liv-Est-und Kurlands, herausgegeben von der Gesellschaft fur Geschichte und Alterthumskunde der Ostsee-Provinzen Russlands (Riga) 13, [Riga], 24-48.
Bielenstein, A. 1892, Die Grenzen des lettischen Volkstammes und der lettischen Sprache in der Gegenwart und im 13. Jahrhundert. Ein Beitrag zur ethnologischen Geographie und Geschichte Russlands (Mit einem Atlas von 7 Blattern = Beilage, Atlas der ethnologischen Geographie des heutigen und des praehistorischen Lettenlandes), St. Petersburg-Riga-Leipzig.
Christiansen, E. 1997, The Northern Crusades, London [first published 1980].
Eesti aastal 1200, Tallinn 2003.
Ernstreits, V. 1999, Livokiel-letkiel, letkiel-livokiel sonarontoz. Libiesu-latviesu, latviesu-libiesu vardnica, Riga.
Heinrichs Livlandische Chronik. Heinrici Chronicon Livoniae. 2. Aufl. Bearbeitet von Leonid Arbusow und Albert Bauer (Scriptores Rerum Germanicum in usum Scholarum ex Monumentis Germaniae Historicis separatim editi), Hannover 1955.
Henricus Lettus 2003, The Chronicle of Henry of Livonia. Translated with a new introduction and notes by James A. Brundage, New York (Records of Western Civilization).
Hildebrand, H. 1865, Die Chronik Heinrichs von Lettland. Ein Beitrag zu Livlands Historiographie und Geschichte, Berlin.
--1880, Zehn Urkunden zur alteren livlandischen Geschichte aus Petersburg und Stockholm.--Mittheilungen aus dem Gebiete der Geschichte Liv- Estund Kurlands, herausgegeben von der Gesellschaft fur Geschichte und Alterthumskunde der Ostsee-Provinzen Russlands (Riga) 12, [Riga], 367- 380.
Jaanits, L., Laul, S. Lougas, V. Tonisson, E. 1982, Eesti esiajalugu, Tallinn.
Jannau, H. G. v. 1828, Ueber die Grund- und Ursprache der Ehsten und die Mittel zu einer allgemeinen Schriftsprache zu gelangen. Eine historisch, philologisch-critische Untersuchung, nebst kurzer Zusammenstellung, Vergleichung und Wurdigung der Finnischen, Livischen, Reval-Ehstnischen und Dorpat-Ehstnischen Sprache, Pernau-Reval (Beitrage zur genauern Kenntniss der ehstnischen Sprache 19).
Kriiska, A., Tvauri, A. Selart, A. Kibal, B. Andersen, A. Pajur, A. 2006, Eesti ajaloo atlas, Tallinn.
Krikmann, A., Pajusalu, K. 2000, Kus on keskmurde keskpunkt.--Inter dialectos nominaque. Puhendusteos Mari Mustale 11. novembril 2000, Tallinn (Eesti Keele Instituudi toimetised 7), 131-172.
Laakmann, H. 1939, Estland und Livland in fruhgeschichtlicher Zeit.--Baltische Lande. Bd.1, Ostbaltische Fruhzeit, Leipzig, 204-262.
--1954, Die Baltischen Lande, Munchen (Staats- und Verwaltungsgrenzen in Ostmitteleuropa: Historisches Kartenwerk 1. Der Gottingen Arbeitskreis: Veroffentlichung 114).
--1973, Erlauterungen zu den Karten.--R. Wittram, Baltische Geschichte. Die ostseelande Livland, Estland, Kurland 1180-1918. Grundzuge und Durchblicke. Mit 7 Karten nebst Erlauterungen von Heinrich Laakmann. Unveranderte Abdruck der ersten Auflage von 1954, Darmstadt, 311-323. Latvijas vestures atlants, Riga 2005.
Lehiste, I., Teras, P., Ernstreits, V., Lippus, P., Pajusalu, K., Tuisk, T., Viitso, T.-R. 2008, Livonian Prosody, Helsinki (MSFOu 255).
Lowisof Menar, K. v. 1895, Livland im Mittelalter. Eine kartographische Darstellung. Mit einem Heft Erlauterungen, Reval.
--1907, Erlauterungen zur Karte von Livland im Mittelalter (Separatabzug aus "Baltische Geschichte im Grundriss"), Reval.
Minahan, J. B. 2000, One Europe, Many Nations. A Historical Dictionary of European National Groups. Westport-London.
Pajusalu, K. 1996, Multiple Linguistic Contacts in South Estonian. Variation of Verb Inflection in Karksi, Turku (Turun yliopiston suomalaisen ja yleisen kielitieteen laitoksen julkaisuja 54).
--2007, Kuidas aratada keelt: salatsiliivi valjavaateid.--Merkityksen ongelmasta vahemmistokielten oikeuksiin. Juhlakirja professori Helena Sulkalan 60-vuotispaivana, Oulu (Acta Universitas Ouluensis, Humaniora B 79), 211-227.
--2009, Dynamics of Estonian Phonology.--STUF. Language Typology and Universals 62, 109-121.
Pajusalu, K., Hennoste, T., Niit, E., Pall, P., Viikberg, J. 2009, Eesti murded ja kohanimed, Tallinn.
Pajusalu, K., Krikmann, A., Winkler, E. 2009, Lexical Relations between Salaca Livonian and Estonian Dialects.--LU XLV, 283-298.
Parrot, J. L. v. 1828, Versuch einer Entwicklung der Sprache, Abstammung, Geschichte, Mythologie und burgerlichen Verhaltnisse der Liwen, Latten, Eesten, mit Hinblick auf einige benachbarte Ostseevolker, von den altesten Zeiten bis zur Einfuhrung des Christenthums. Nebst einer Topographie und einer topographischen Karte des Landes zu Anfang des dreizehnten Jahrhunderts. Erster Band, Stuttgart.
--1839, Versuch einer Entwicklung der Sprache, Abstammung, Geschichte, Mythologie und burgerlichen Verhaltnisse der Liwen, Latten, Eesten, mit Hinblick auf einige benachbarte Ostseevolker, von den altesten Zeiten bis zur Einfuhrung des Christenthums. Neue Ausgabe mit einer Topographie und einer topographischen Karte des Landes zu Anfang des dreizehnten Jahrhunderts und einem Polyglotten-Atlas, Berlin.
Perlbach, M. 1886, Urkunden des rigaschen Capitel-Archives in Furstlich Czartoryskischen Bibliothek zu Krakau.--Mittheilungen aus dem Gebiete der Geschichte Liv- Est- und Kurlands, herausgegeben von der Gesellschaft fur Geschichte und Alterthumskunde der Ostsee-Provinzen Russlands (Riga) 13, [Riga], 1-23.
Prants, H. 1920, Eesti wanem ajalugu, Tallinn. Appendix: Laanemere maade kaart XIII aastasajal. Sojawae Topograafia osakonna waljaanne, s.a., s.l.
Richter, A. v. 1857, Geschichte der dem russischen Kaiserthum einverleibten deutschen Ostseeprovinzen bis zur Zeit ihrer Vereinigung mit demselben. Theil 1. Die Zeiten der reingermanischen Entwickelung. 1158-1562. Band 1. Die Eroberung 1158-1237. Die Behauptung des Landes und die innern Einrichtungen 1237-1347. Riga (Nachdruck Hannover-Dohren 1972).
Rosenplanter, J. H. 1816, Ueber die Wanderungen der Ehsten gleich nach der Pest.--Beitrage zur genaueren Kenntniss der estnischen Sprache 6, Pernau-Reval, 42-47.
Rutenberg, O. v. 1859, Geschichte der Ostseeprovinzen Liv-, Esth- und Kurland von der altesten Zeit bis zum Untergange ihrer Selbstandigkeit. Bd. 1, Leipzig.
Strods, H. 1963, Vidzemes etnografias avotu un petijumu aspektas.--Arheologija un etnografija (Rakstu krajums V), Riga, 5-24.
Tanning, S. 1958, Mulgi murde ja liivi keele suhetest.--KKIU II, Tallinn, 105-117.
Tarvel, E. 1995, Uber die Notwendigkeit, die livische Siedlungsgeschichte zu Erforschen.--Latvijas Zinatnu Akademijas Vestis. A Dala: Socialas un Humanitaras Zinatnes. Proceedings of the Latvian Academy of Sciences. Section A: Human and Social Sciences 9-10, [Riga], 57-61.
Vestring, S. H. 1998, Lexicon Esthonico Germanicum, Tartu.
Viitso, T.-R. 2008, Liivi keel ja laanemeresoome keelemaastikud, Tartu-Tallinn.
--2009, Liivi and Leivu: Shared Innovations and Problems.--LU XLV, 269-282.
Wiedemann, F. J. 1861a, Joh. Andreas Sjogren's Livische Grammatik nebst Sprachproben, St. Peterburg.
--1861b, Joh. Andreas Sjogren's Livisch-deutsches und deutsch-livisches Worterbuch, St. Peterburg.
Winkler, E. 1994, Salis-livische Sprachmaterialien, Munchen (Veroffentlichungen des Finnisch-Ugrischen Seminars an der Universitat Munchen, Serie C: Band 21).
--1999a, Altere Livische Sprachmaterialien, Munchen (Veroffentlichungen des Finnisch-Ugrischen Seminars an der Universitat Munchen, Serie C: Band 22).
--1999b, Katkeintonatsiooni tekkimisest laanemeresoome keeltes.--Odagumeresoomo veeremaaq. Laanemeresoome perifeeriad, Voro (Voro Instituudi Toimotiseq 6), 201-206.
--2000, Salatsi liivi keele asendist. Vahekokkuvote.--Odagumeresoomo piirisuameq. Keskused laanemeresoome piiridel, Voro (Voro Instituudi Toimondusoq 10), 127-133.
--2002, Laensonakihid Salatsi liivi keeles ja mis sellest jareldub.--Vaikeisi kiili kokkoputmisoq. Vaikeste keelte kontaktid, Voro (Voro Instituudi Toimondusoq 14), 61-69.
Winkler, E., Pajusalu, K. 2009, Salis-livisches Worterbuch, Tallinn (Linguistica Uralica. Supplementary Series. Volume 3).
[TEXT NOT REPRODUCIBLE IN ASCII.]
(1) The format of the citations from the chronicle is to the chapter (Roman number) and paragraph (Arabic number) according to the critical edition by Leonid Arbusow and Albert Bauer (Heinrichs Livlandische Chronik 1955) and the translation of the passages follows the English translation by James A. Brundage's (Henricus Lettus 2003). The chronicle is not divided into chapters and paragraphs in the same way in these issues of the chronicle.
(2) Spelling of Courland Livonian words is presented according to Ernstreits 1999.
URMAS SUTROP (Tallinn-Tartu), KARL PAJUSALU (Tartu)
Institute of the Estonian Language and University of Tartu
University of Tartu
|Printer friendly Cite/link Email Feedback|
|Author:||Sutrop, Urmas; Pajusalu, Karl|
|Date:||Dec 1, 2009|
|Previous Article:||The role of duration ratios and fundamental frequency in spontaneous Livonian.|
|Next Article:||Livonian and Leivu: shared innovations and problems.|