Grant H. Lundberg 
Brigham Young University 
A Preliminary Report on Dialectological Fieldwork In 
Northwestern Croatia: Brezova Gora and 
the Croatian-Slovene Dialect Continuum 
This paper is a eontribution to the eorpus of dialeet data from regions along the Slo-
vene-Croatian national border. It provides a brief deseription of the phonemie inventory of the 
Croatian village dialeet of Mohenski in Brezova Gora. It al so reports on developments in the pro-
sodie system based on a speetrographie analysis of tonal oppositions in this dialeet and eompares 
the situation in Mohenski to the dialeets just aeross the border in Slovenia. 
Pričujoča razprava je prispevek k zbirki dialektološkega gradiva s področja vzdolž sloven-
sko-hrvaške državne meje. Podan je kratek opis fonološkega inventarja hrvaškega govora Mohen-
skega v Brezovi gori. Obravnava tudi razvoj prozodičnega sistema na osnovi spektografske raz-
člembe tonemskih nasprotij v tem govoru in ga primerja s stanjem govorov v neposredni bližini 
na drugi strani državne meje. 
1. Introduction 
One of the more interesting questions in Western South Slavic dialectology is 
the relationship, both historical and modern, between the dialects of the Slovene and 
the Kajkavian Croatian speech territories. The debate over the origin and genetic re-
lationship between these dialect regions goes back to Dobrovsky in the early 19th 
century and was carried on by such scholars as Belič, Ramovš, Ivšič and more re-
cently by Ivič, Vermeer and Greenberg (see Greenberg 2000: 42-50). 
Contemporary Slavic linguists agree that the dialects of the Slovene and Kaj-
kavian speech territories are part of a dialect continuum with almost all of the iso-
glosses which unite them being archaisms rather than shared innovations. Although 
it is c1ear that this is a dialect continuum, a great deal of weight is given to the po-
litical border between Slovenia and Croatia. Dialect maps of this region often cor-
respond exactly with the political border. This may be because little is known about 
the characteristics of the village dialects directly on and around the national fron-
tier. It is also because in some cases the political border does represent a linguistic 
border (Lončarič 156). Of course, linguistic isoglosses do not always correspond to 
political boundaries. For example, eastern Haloze, a Pannonian Slovene dialect, has 
merged the reflexes of the Common Slavic jat and jers, a Kajkavian development,! 
and does not exhibit circumflex advancement, a Slovene development (Lundberg 
1997, 1999). On the other hand, the Kajkavian dialect of Bednja does have some cir-
cumflex advancement (Vermeer 1979). Additional descriptions of dialects in this re-
gion could be very helpful to linguists analyzing development s across this national 
frontier, especially for the question of tone loss. 
! This important Kajkavian development was pointed out by Ivič (1968: 57). 
56 Slovenski jezik - Slovene Linguistic Studies 4 (2003) 
1.2. Brezova Gora 
This paper is a contribution to the pool of dialect data from this area. 1 will 
provide a brief description of the phonemic inventory of the Croatian village dialect 
of Mohenski in Brezova Gora. 1 will also report on development s in the prosodic 
system based on a spectrographic analysis of tonal oppositions in this dialect and 
compare the situation in Mohenski to the dialects just across the border in Slovenia. 
1.3. Location 
Brezova Gora is a small area in northwestern Croatia about five kilometers 
northwest of Trakošean. It is located directly across the border from the Slovene dia-
lect area of Haloze, specifically the village dialect of Velika Varnica. Brezova Gora 
runs southwest along the border for nearly five kilometers, starting from Jamno, and 
is made up of approximately fifteen villages. This is an interesting area for dialect 
studies because, in terms of the geography of the region, these villages are more 
closely connected to the Slovene villages across the border than they are to other 
Croatian villages. This raises the question of their linguistic connection to Pannoni-
an Slovene and Kajkavian Croatian dialects. 
Figure 1: Map of Slovene-Croatian Border Region 
Slovenia 
Ptuj 
v. Varnica 
Varaždin 
Croatia 
1.4. Previous Scholarship 
To my knowledge almost nothing has been written about Brezova Gora in dia-
lect literature. Kolarič, in his 1964 article on the Haloze dialect, mentions it as a 
contrast to the Slovene dialect he is describing. Using about five forms to represent 
the diphthongal reflexes of the Common Slavic *e and *0 as ie and uo respectively, 
he indicates that Brezova Gora appears to be like Bednja, a Kajkavian dialect (397).2 
Jedvaj in his well-known work on the Bednja dialect mentions that a dialect like 
2 There are at least two reasons to be cautious about this conclusion. First, we are presented 
with very little information. We are given the reflexes of only two Common Slavic phonemes. 
Second, Kolarič's informant from Mahinski (possibly Mohenski), had lived for over thirty years in 
Slovenia when Kolarič spoke with her. She al so said that she rarely went home and that they spoke 
differently there at that time than when she left (397). 
G. H. Lundberg, Preliminary Report on Dialectological Fieldwork ... 57 
that of Bednja could be heard in Cvetlin (283). Cvet1in is not part of Brezova Gora, 
but it is only five kilometers to the northeast. The interesting point here is that Bre-
zova Gora is located right at the traditional dividing line between Pannonian Slo-
vene and Kajkavian dialects. This is also the location of the isoglosses for several 
important vocalic development s as well as the isogloss of tone loss. Northeastern Slo-
vene dialects have lost tone, while some Kajkavian Croatian dialects have retained it. 
This paper will attempt to determine if Brezova Gora is vocalically and tonemically 
more like Haloze or Bednja. 
2. Haloze 
ln order to make a comparison between dialect systems, I will start by briefly 
describing the system found in Haloze. The Slovene dialect directly across the border 
from Mohenski is Velika Varnica. There is no wholly reliable description of this dia-
lect,3 but my experience in this area confirms that it is a typical central Haloze vil-
lage dialect. Forms from the author's fieldwork in Belavšek, a neighboring central 
Haloze village dialect, will be used here. 
Figure 2: Belavšek Vowel System (Central Haloze) 
ilU iilii:i 
iilii:i Q:U 
gJg: (a:) 
9 
ulu: 
Belavšek, like Velika Varnica, Trdobojci and other central Haloze systems, has 
both monophthongs and diphthongs in its vocalic system. Judging from the historical 
development of the vocalic system, it is a Pannonian Slovene dialect. This can be 
seen in the merger of the reflexes of the Common Slavic *e, *e and *1JI*b > *<1 in ~, 
while the reflex of the Common Slavic *f! has remained distinct, 'p~:t 'five' < *e, 
'p~:č 'oven' <*e, 'd~:n 'day' < *<1, but z'vii:izda 'star' < *f! (Lundberg 1999: 100). This 
dia1ect has a rounded reflex of Common Slavic *a, t'rg:vg 'grass', and the reflexes of 
the Common Slavic *Q and *0 have merged in Q:U, gla'vQ:u 'he ad' < *Q , b'IQ:u 'was' 
< *0 (101). Belavšek has a fronted reflex of the Common Slavic *u, 'vii:ista 'mouth', 
'pii:ivali 'buili' , and the Common Slavic *1 has developed into areflex distinct from 
*Q and *0, 'du:go 'long', 'vu:k 'wolf' (101). 
2.1. Prosody 
Belavšek has a vocalic system in which all distinctions of word-level prosody 
are rea1ized in the accented syllable and in which the accent is free to falI on any 
syllable of the word. This system has distinctive quantity, b'rgt 'brother', b'ra:t (sup.) 
'to pick', a1though that quantity, stilI distinctive, carries low functiona1 load in Be-
lavšek because it is almost always accompanied by a quality distinction, 'd~lati 'to 
work', 'dii:ilaš 'you work'. Circumflex advancement occurs in the same voca1ic con-
3 A student of Professor Zinka Zorko at the University of Maribor, Anton Roškar, wrote a 
description of Varnica as part of his course work. In some ways it is a very helpful thesis, espe-
cially in the area of morphology, but his approach to the vocalic system and prosody is incon-
sistent. 
58 Slovenski jezik - Slovene Linguistic Studies 4 (2003) 
texts as in other Pannonian dialects, me'sQ:u 'meal' , gla'vQ:u (acc. sg. fem.) 'head' , 
kU'kQ:Uš 'chicken', but 'vii:ižgali 'kindled' (95). 
There is no distinctive tone in Haloze, but there is a non -phonemic rising con-
tour on accented syllables (Lundberg 2001: 95). On short and long stressed syllables 
the pitch level starts low at the beginning of the syllable and peaks 75% to 80% 
through the duration of the syllable nuc1eus, then it falls off slight1y but finishes 
higher than it started. The basic measurements of the long and short syllables are 
almost exact1y the same. The average long syllable rises 41Hz to a pe ak at 274Hz 
and then falls 25Hz to the end of the syllable. The average short syllable rises 34Hz 
to a pe ak of 273Hz and then falls 16Hz to the end of the syllable. The striking dif-
ference is that the short syllable makes this contour fit within half the duration of 
the long syllable. The contour is therefore much sharper. 
3. Kajkavian 
As was mentioned before, Kolarič states that, according to his limited informa-
tion, Brezova Gora should be c1assified as a Kajkavian dialect. Lončarič and Ivšič 
also list it as Kajkavian on their dialect maps (Lončarič 65, 199). If it is Kajkavian, 
what is it likely to look like? Vermeer lists some characteristics shared by most Kaj-
kavian dialects, inc1uding a likely Common Kajkavian vowel system (1983: 456). 
Several of these points will be discussed below. 
Figure 3: Common Kajkavian Vowel System 
ili: iiiii: ulu: « Q, later D 
(!Ii:e iJ olu:o (olo:u) rlr: 
ele: ala: 
The vowel system above represents the most salient Kajkavian vocalic features. 
First, the reflexes of the Common Slavic jat and jers have merged,4 b'ri:eg 'hill', 
'di:en 'day' (Jedvaj 286, 288). Second, the reflexes of the Common Slavic *e and *e 
have merged, 'ma:se 'meat', 'ša:st 'six' (285, 288). Third, the reflex of the Common 
Slavic *u, though fronted in the early stages of Common Kajkavian, has in most 
modern Kajkavian dialects velarized. Fourth, the reflexes of the Common Slavic *Q 
and *l have merged in a phoneme distinct from that of the Common Slavic *0, 
mo:už 'husband', vo:uk 'wolf', but ny:es 'nose' (289, 285). To this list we might add 
several secondary Kajkavian characteristics. 1) There is a velarization of the reflex 
of *a. 2) The reflex of *e is low. 3) The reflex of long *e is distinct from i and e. 4) 
There is a fronting of the reflex of *0 to e or o. 5) The reflex of long *e and *0 are 
diphthongs of the type ie and uo respectively (Vermeer 1983: 440-1). 
3.1. Prosody 
In most Kajkavian dialects tonal oppositions are retained only on long syllables. 
The fundamental frequency (FO) contour is contained in one syllable, and it is the 
shape of that tone contour , rising or falling, that is distinctive and not the FO height 
of the following syllable, as is the case for Standard Serbo-Croatian and to some 
extent Slovene as well (Lehiste and Ivič 1986: 81). Kajkavian, like Slovene, has 
4 All Kajkavian examples are from Jedvaj 1956. 
G. H. Lundberg, Preliminary Report on Dialectological Fieldwork ... 59 
neo-cireumflex, which is a cireumflex that developed on a syllable with an original 
aeute (Lončaric 40). The neo-aeute in Kajkavian on etymologieally short syllables is 
long rising, sela (nom. pI. neut.) 'villages', ženska (nom. sg. fem.) 'woman' (Vermeer 
1983: 440). Finally, some Kajkavian dialeets have limited eireumflex advaneement. 
For example, Bednja has advaneement onto closed syllables but not onto open syllab-
les, so'de:il 'planted' (Jedvaj 296), ke'ky:eš 'hen' (283), but 'ma:se 'meat' (288). 
4. Mohenski 
Based on the diseussion of the Pannonian Slovene dialeet of Belavšek and of 
the common Kajkavian features listed above, it should now be possible to compare 
the Brezova Gora dialeet of Mohenski to these in order to determine which features 
it shares with the dialeets that surround it. 
Figure 4: Vocalic System of Mohenski 
u 
ie 
e o 
a (,l 
4.1. Examples 
i < *i: - ve'li (3rd pers. sg.) 'to say', 'zimu (aee. sg. fem.) 'winter', 'pil (I-pep. mase. 
sg.) 'to drink', 'bili (I-pep. mase. pI.) 'to be', 'vinsko (adj. nom. sg. neut.) 'wine', 
s(,l'dil (I-pep. mase. sg.) 'to plant' 
< *i - žg(,l'nica (nom. sg. fem.) 'brandy' , 'hiža (nom. sg. fem.) 'house', 'hiži (Ioe. sg. 
fem.), ja'zik (nom. sg. mase.) 'tongue' , ke'siti (inf.) 'to eut' 
ie < *e: - s'vietu (Ioe. sg. mase.) 'world', 'bieži (imp. sg.) 'run', 'liet (gen. pI. neut.) 
'year', d'rievo (nom. sg. neut.) 'tree', k'liet (nom. sg. fem.) 'cellar' 
< *;}: - g'nies (adv.) 'today', 'dien (nom. sg. mase.) 'day' 
? < *e - 'l?to (nom. sg. neut.) 'year', 'b?lega (adj. gen. sg. masc.) 'white', 'd?lati 
(inf.) 'to do' 
< *;} - 'd?š (nom. sg. masc.) 'rain', 'P?S (nom. sg. masc.) 'dog' 
e < *r - 'serce (nom. sg. neut.) 'heart', der'vana (adj. nom. sg. fem.) 'wooden' 
< *0 - ge'veriš (2nd pers. sg.) 'to talk', 'hečeš (2nd pers. sg.) 'to want', pede'm(,l-
čen (adv.) 'like at home', kes'm(,lti (inf.) 'to pluek', he'diii (l-pep. mase. pI.) 
'walked' 
a < *e: - 'pat (num.) 'five', 'maso (nom. sg. neut.) 'meat', masa (gen. sg. neut.) 
< *e - tride'satega (adj. gen. sg. neut.) 'thirtieth', 'davat (num.) 'nine' 
< *e - š'tari (nom. sg. mase.) 'whieh', 'male (gen. sg. fem.) 'flour' , 'žan ska (nom. 
sg. fem.) 'woman', 'dasat (num.) 'ten', 'davat (num.) 'nine', 'nabren (1st pers. sg.) 
'cannot', 'rakel (l-pep. mase. sg.) 'to say' 
(,l < *a: - s't(,lri (adj. nom. sg. mase.) 'old', pep'r(,lvili (l-pep. mase. sg.) 'to repair' , 
z'n(,lju (3rd pers. pI.) 'to know', z'n(,lš (2nd pers. sg.) 
60 Slovenski jezik - Slovene Linguistic Studies 4 (2003) 
< *a - g'rgba (nom. sg. fem.) 'valley', g'rgx (nom. sg. masc.) 'bean', k'rgve (acc. 
pI. fem.) 'cow', 'jgbuka (nom. sg. fem.) 'apple' 
o < *aN - gi'bojnca (nom. sg. fem.) 'gibanica' 
< *0 - p'ropalo (l-pcp. neut. sg.) 'ruined', 'pojel (l-pcp. masco sg.) 'to eat', 'počel 
(l-pcp. masco sg.) 'to begin', 'kosti (dat. g. fem.) 'bone', 'dobil (l-pcp. masco sg.) 
'to receive', ke'koš (nom. sg. masc.) 'chicken', ke 'koši (nom. pI. masc.), š'kolu 
(acc. sg. fem.) 'school' 
u < *l - 'puno (adj. nom. sg. neut.) 'full', 'suncu (loc. sg. neut.) 'sun', 'jgbušnica 
(nom. sg. fem.) 'apple wine' 
< *No - pe 'muči (dat. sg. fem.) 'power' 
< *Q: - pe'sudili (l-pcp. masco pI.) 'to judge' , 'sused (nom. sg. masc.) 'neighbor', 
'bum (fut. 1st pers. sg.) 'to be', 'buš (2nd pers. sg.), 'ruke (nom. pI. fem.) 'hands' 
< *Q - 'su (3rd pers. pI.) 'to be', s'mu (1st pers. pI.) 
< *u: - 'pujcice (acc. pI. fem.) 'girls', d'rugi (adj. nom. pI. masc.) 'other' 
< *u - 'vuzen (nom. sg. masc.) 'Easter' , 'vuzma (gen. sg. masc.), 'gujdika (nom. 
sg. fem.) 'chicks', 'buberek (nom. sg. masc.) 'cucumber', 'buberke (acc. pI. masc.) 
4.2. Vocalic Developments 
Based on these forms, Mohenski has much more in common with Kajkavian 
than with Pannonian Slovene dialects. 1) It has the merger of the jat and the jers, 
'liet, 'dien. 2) The reflex of Common Slavic *0 is distinct from that of the reflexes 
of Common Slavic * l and * Q , which have merged in u, ke'koši, but 'puno and pe-
'sudili. Mohenski also exhibits several of Vermeer's secondary Kajkavian features, all 
of which it also shares with Bednja. 1) It has a velarized a. 2) The reflex of the jat 
is distinct from i and e. 3) The reflex of Common Slavic *0 is fronted to e. 4) The 
reflex of the jat is a diphthong with rising sonority, ie. 
4.3. Prosody 
All prosodic opposlt1ons including quantity opposlt1ons have been lost in Mo-
henski. This is based on the author's perception as well as an instrumental analysis, 
which will be discussed below. There is almost no circumflex advancement in this 
dialect. As would be expected for a Kajkavian dialect, there is no advancement onto 
open syllables. This is true no matter what the relative syllable weight of the word 
is, 'maso 'meat', 'kosti 'bone', 'bili 'were' , 'bilo 'was', d'rievo 'tree'. 
The picture for circumflex advancement onto closed syllables is not so clear. 
This is likely due to lack of sufficient examples. In Bednja and eastern Haloze cir-
cumflex advancement onto closed syllables tends to function according to a hierar-
chy of syllable weight (Vermeer 1987, Greenberg 1992, Lundberg 1997). According to 
this hierarchy, advancement is most likely to take place from a long syllable onto a 
long closed syllable. This study records only one example of this kind of advance-
ment in Mohenski, sg'dil 'planted'. The next step in the hierarchy is advancement 
from a short syllable onto a long closed syllable. This stage of advancement is not at-
tested in Mohenski, 'daviit 'nine', 'dasiit 'ten', 'pojel 'ate', 'počel 'began' . The final 
stage of the hierarchy on closed syllables is from a short syllable onto a short closed 
syllable. One word with this type of advancement appears in several forms in this 
G. H. Lundberg, Preliminary Report on Dialectological Fieldwork ... 61 
study, ke'koš 'chicken'. It is difficult to explain why this final stage of advancement 
is attested while other more likely types of advancement are not. It could be that 
this inconsistency is due to the borrowing of a form like ke'koš. This is the regular 
form just across the border in Haloze. This explanation is not satisfying because it 
is ad hoc and because, except for advancement, the form has little in common 
phonetically with the form from Haloze, kU'kQ:Uš. At this point, there simply are 
not enough examples to make a good argument. 
4.3.1. Spectrographic Analysis 
As was mentioned above, all tonemic oppositions have been lost in Mohenski. 
This is surprising based on the fact that it is clearly a Kajkavian dialect that is simi-
lar in many ways to Bednja. Bednja is less than ten kilometers away and has re-
tained tonemic oppositions. 
ln order to confirm the author's perceptual observation, an instrumental analy-
sis was performed using Speech Analyzer, software created by the Summer Institute 
of Linguistics. Twenty-six words were examined on which the traditional accent 
types might be expected.5 There was, of course, variation, but I found no consistent 
connection to the historical accent types. As a general rule, the FO peak of the 
word is also the peak of the accented syllable. The FO rises gradually through the 
pretonic syllables to the peak in the accented syllable and then falls gradually 
through the posttonic syllables. On all measured words, the FO height of the post-
tonic syllable turns out to be lower than that of the accented syllable.6 The average 
peak of the FO on the accented syllable is at 33% of the duration of the syllable nu-
cleus. The defau1t contour of accented syllables seems to be falling, but there are 
several forms in this study with a pe ak from 45% to 68% of the duration. In fact, 
the shape of the syllable, especially the type of consonant that precedes the ac-
cented syllable, has the most influence on the location of the FO peak. This is con-
sistent with Lehiste and Peterson's work on English accented syllables, in which they 
argue that a preceding voiceless consonant, especially avoiceless fricative causes the 
peak of the accented syllable to be early, while a preceding voiced consonant causes 
the FO to rise slowly with a peak near the center of the syllable (Lehiste and Peter-
son 420). For example, š'kola has a pe ak located 22% of the way through the syl-
lable nucleus, and 'salu peaks at 8% of the duration. On the other hand, žgg'nica 
peaks at 68% of the duration, 'buš peaks at 45%, and 'maso has a peak located at 
60% of the duration of the syllable nucleus. 
5. Conclusions 
If we compare Mohenski to the Pannonian dialects to the west and to the Kaj-
kavian dialect to the south and east, it clearly has more in common with Kajkavian. 
The best examples of this are the merger of the reflexes of the Common Slavic jat 
and jers as well as the merger of the reflexes of the Common Slavic *Q and *l in a 
phoneme separate form the reflex of the Common Slavic *0. The dialect also has a 
velarized reflex of the Common Slavic *a and shows some evidence of the fronting 
of the reflex of *0, especially in pretonic position. 
5 These twenty-six forms were made up of ten circumflex, eight acute and eight neo-acute. 
6 The one exception to this was on the word 'počel. The posttonic syllable in this word was 
higher than the accented syllable, but this can be explained as a result of the palatal affricate 
(Lehiste and Peterson). 
62 Slovenski jezik - Slovene Linguistic Studies 4 (2003) 
It is interesting to note that the dialect has changed significantly in the seventy 
years since Kolarič's informant left Brezova Gora? The innovations in Mohenski over 
nearly three quarters of a century are partly due to influence from Standard Ser-
bo-Croatian. Kolarič's informant was bom in 1901 and moved to Slovenske Gorice in 
Slovenia in 1930. She could neither read nor write. In the early 1930's a school was 
bui1t in Brezova Gora. Both of my major informants, now in their eighties, went to 
grammar school there. This may explain the loss of the diphthongal reflex of the 
Common Slavic *0. The neighboring Slovene dialects in the area are probably also 
part of this development. People from Mohenski have family relationships that ex-
tend across the national frontier. They also cross the border into Slovenia regularly 
to shop and even to work if possible. Many people in Brezova Gora believe that 
prices are lower and social benefits are better in Slovenia. Anecdotal evidence, based 
on interviews with the owner of a small grocery store and an elementary school 
teacher in Haloze, suggests that the Zagorci try to use forms that they think are 
Slovene when doing business in Slovenia. 
Finally, it must be noted that Brezova Gora is likely not uniform in its dialect. 
The author's limited experience there suggests that some village dialects, like Ilijevci, 
which is c10ser to the border, have a dialect very similar to that found in central 
Haloze and some, like Mohenski, are of Kajkavian origin. More fieldwork is needed 
in Brezova Gora to c1arify the situation and to more specifically locate the isogloss 
of tone loss and circumflex advancement in this border region. 
Works Cited 
Greenberg, Marc L. 1992. "Circumflex Advancement in Prekmurje and Beyond." 
Slovene Studies. 14.1: 69-91. 
-. 2000. AHistorical Phonology of the Slovene Language. Heidelberg: Carl Winter 
UniversiUitsverlag. 
Ivic, Pavle. 1968. "Procesi rasterecenja vokalskog sistema u kajkavskim govorima." 
Zbornik za filologiju i lingvistiku. 11: 57-68. 
Jedvaj, Josip. 1956. "Bednjanski govor." Hrvatski dijalektološki zbornik. 1: 279-330. 
Kolarič, Rudolf. 1964. "Haloški govor." Prace filologiczne 18.2: 395-401. 
Lehiste, Ilse, and Gordon E. Peterson. 1961. "Some Basic Considerations in the 
Analysis of Intonation." Journal of the Acoustic Society of America. 33.4: 
419-25. 
Lehiste, Ilse, and Pavle Ivic. 1986. Word and Sentence Prosody in Serbocroatian. 
Cambridge, Massachusetts: The MIT Press. 
Lončaric, Mijo. 1996. Kajkavsko narječje. Zagreb: Školska knjiga. 
7If Kolarič's forms can be trusted, the diphthongal reflex of *0 has not been retained, and 
he list one form with circumflex advancement onto an open syllable, žre'bie (398). This has also 
not been retained. It should al so be noted that, although the present study is more detailed than 
Kolarič's description of Brezova Gora, it is also based on a small data set. This description is 
based on one hour of recorded dialog and notes from several additional conversations with resi-
dents of the village. Because of the small amount of material, the conclusions here must be 
viewed with some caution. 
G. H. Lundberg, Preliminary Report on Dialectological Fieldwork ... 63 
Lundberg, Grant H. 1997. "Circumflex Advancement in Haloze." Slovene Studies. 
21.1: 61-81. 
-. 1999. "Peliminary Report on Dialectological Fieldwork in Haloze, Slovenia." Slo-
venski jezik-Slovene Linguistic Studies. 2: 91-108. 
-. 2001. "Loss of Tonemic Oppositions in Eastern Haloze, Slovenia: An Instrumental 
StUdy." Balkanistica. 14: 83-100. 
Vermeer, Willem R. 1979. "Innovations in the Kajkavian Dialect of Bednja." Dutch 
Contributions to the Eighth International Congress of Slavists. Zagreb, Ljub-
ljana, 1978 (ed. Jan M. Meier), 347-81. Lisse: Peter de Ridder. 
-. 1983. "The Rise and FalI of the Kajkavian Vowel System." Studies in Slavic and 
General Linguistics. 3: 439-77. 
1987. "The Treatment of the Proto-Slavic Falling Tone in the Resian Dialects of 
Slovene." Dutch Studies in South Slavic and Balkan Linguistics. (ed. A. A. 
Barentsen), 275-98. Amsterdam: Rodopi. 
Prispelo januarja 2003, sprejeto marca 2003 
Received January 2003, accepted March 2003 
A Preliminary Report on Dialectological Fieldwork 
in Northwestern Croatia: Brezova Gora and the 
Croatian-Slovene Dialect Continuum 
One of the more interesting questions in Western South Slavic dialectology is 
the relationship, both historical and modern, between the dialects of the Slovene and 
the Kajkavian Croatian speech territories. The debate over the origin and genetic re-
lationship between these dialect regions goes back to Dobrovsky in the early 19th 
century and was carried on by such scholars as of Belic, Ramovš, Ivšic, and more 
recent1y by Ivic, Vermeer and Greenberg. 
Contemporary Slavic linguists agree that the dialects of the Slovene and Kaj-
kavian speech territories are part of a dialect continuum with almost all of the iso-
glosses which unite them being archaisms rather than shared innovations. Although 
it is c1ear that this is a dialect continuum, the political border between Slovenia and 
Croatia has had an important influence on dialect development in this area. This 
paper is a contribution to the pool of dialect data from the border region. It pro-
vides a brief description of the phonemic inventory of the Croatian village dialect of 
Mohenski in Brezova Gora. It also reports on development s in the prosodic system 
based on a spectrographic analysis of tonal oppositions in this dialect and compares 
the situation in Mohenski to the dialects just across the border in Slovenia. 
Based on the dialect forms in this study, Mohenski has much more in common 
with Kajkavian than with Pannonian Slovene dialects. (1) It has the merger of the 
jat and the jers, 'liet, 'dien. (2) The reflex of Common Slavic *0 is distinct from 
that of the reflexes of Common Slavic *l and *Q, which have merged in u, ke'koši, 
but 'puno and pe'sudili. Mohenski also exhibits several of Vermeer's secondary Kaj-
kavian features, all of which it also shares with Bednja. (1) It has a velarized a. (2) 
64 Slovenski jezik - Slovene Linguistic Studies 4 (2003) 
The reflex of the jat is distinct from i and e. (3) The reflex of Common Slavic *0 is 
fronted to e. (4) The reflex of the jat is a diphthong with rising sonority, ie. 
All prosodic oppositions including quantity oppositions have been lost in Mohenski. 
This is based on the author's perception as well as an instrumental analysis. There is al-
most no circumflex advancement in this dialect. As would be expected for a Kajkavian 
dialect, there is no advancement onto open syllables. This is true no matter what the re-
lative syllable weight of the word is, 'maso, 'kosti, 'bili, 'bilo, d'rievo. 
Additional dialect descriptions from this area are needed in order to more spe-
cifically locate the isoglosses of circumflex advancement and tone loss in this part 
of the Slovene-Croatian dialect continuum. 
Uvodna opažanja iz dialektološke terenske raziskave na 
severozahodnem Hrvaškem: Brezova Gora in 
hrvaško-slovenski narečni kontinuum 
Eno od zanimivejših vprašanj v zahodni južnoslovanski dialektologiji je tako 
zgodovinsko kot sodobno razmerje med govori slovenskega in hrvaškega kajkavskega 
jezikovnega ozemlja. Razpravljanje o izvoru in sorodnosti teh govorov sega v začetek 
19. stoletja k Dobrovskemu, nadaljevali so ga Belic, Ramovš, Ivšic, v novejšem času 
pa Ivic, Vermeer in Greenberg. 
Današnji slovanski jezikoslovci se strinjajo v ugotovitvi, da so govori slovenske-
ga in kajkavskega jezikovnega ozemlja del narečnega kontinuuma, v katerem so sko-
raj vse izoglose, ki povezujejo ta dva dela, arhaizmi in ne skupne inovacije. Čeprav 
je jasno, da gre za na rečni kontinuum, pa je imela politična meja med Slovenijo in 
Hrvaško pomemben vpliv na razvoj narečij tega področja. 
Pričujoča razprava je prispevek v zbirko narečnega gradiva z mejnega področja. 
Podaja kratek opis glasovnega inventarja hrvaške vasi Mohenski v Brezovi Gori. Na 
osnovi spektografske analize tonemskih nasprotij v tem govori avtor poroča tudi o 
razvoju prozodičnega sistema ter stanje v govoru Mohenskega primerja z govori na 
slovenski strani meje. Narečne oblike iz te raziskave kažejo, da ima govor Mohen-
skega veliko več skupnega s kajkavskimi kot s slovenskimi panonskimi govori: (1) 
sovpad jata s polglasnikoma, npr. 'liet, 'dien; (2) refleks psI. *0 se razlikuje od re-
fleksov psI. *1 in *Q, ki sta sovpadla v u, npr. ke'koši, vendar 'puno in pe'sudili. Go-
vor Mohenskega izkazuje tudi nekatere Vermeerove drugotne kajkavske poteze, ki 
jih ima vse tudi govor Bednje: (1) zaokroženi a; (2) refleks jata se razlikuje od i in 
e; (3) refleks psI. *0 se je pomaknil naprej v e; (4) refleks jata je dvoglasnik z ras-
točo zvočnost jo ie. 
Vsa prozodična nasprotja, vključno s kolikostnimi, so se v govoru Mohenskega 
izgubila. To opažanje temelji na avtorjevem slušnem vtisu in na strojni analizi. Go-
vor skoraj ne pozna pomika cirkumfleksa. Kot bi bilo pričakovati za kajkavski go-
vor, pomika na odprte zloge ni. To velja ne glede na relativno zložno težo besede, 
npr. 'maso, 'kosti, 'bili, 'bilo, d'rievo. 
Za natančnejšo določitev izoglos pomika cirkumfleksa in izgube tonemskosti v 
tem delu slovensko-hrvaškega narečnega kontinuuma bodo potrebni dodatni opisi 
govorov s tega področja.