The Phonology of Slovenian

Dissemination summary

The Unique Sounds of Slovenian

Dr P (Peter) Jurgec (m) 12-10-1979, University of Tromsø (Norway), Leiden

University Centre for Linguistics

Slovenian is a small South Slavic language with some remarkable properties. This
project provides a complete description of Slovenian sound and stress patterns. The
data will be recorded, analyzed digitally, and confirmed by production and perception
experiments.

Time frame: September 2012—August 2014

Summary

This project investigates the sound patterns of Slovenian. This South Slavic language
is vastly underdescribed. The existing studies are hardly accessible to the outside
world, as they are published mostly in Slovenian and furthermore they tend to be
based on less than up-to-date empirical methodologies. The main focus of this project
is a complete and comprehensive description of the behavior of vowels and
consonants and stress distribution, to be published as a monograph.

Slovenian exhibits several recently identified properties that are found in very few
other languages. Better knowledge of these noteworthy patterns would contribute to
the general phonological and linguistic theory. A considerable part of the project is
concentrated on three such phenomena: (i) the interaction of tones with vowel quality,
(ii) vowel alternations in unstressed positions, and (iii) the phonological properties of
loanwords borrowed into Slovenian. These patterns will be examined experimentally,
including acoustic analysis and several production and perception experiments.

KEYWORDS: Slovenian, phonology, Slavic, theoretical phonology,

experimental/laboratory phonology

Research proposal (1.687 words)

1 The subject

The subject of this project is the phonology of Slovenian. The principal aim is to
provide a comprehensive description of Slovenian sound system, segmental
distributions, alternations, stress, and intonation patterns. The description is
complemented by a contemporary theoretical account. Moreover, the patterns
particularly interesting from the viewpoint of the general phonological theory are to
be examined by means of digital speech analysis and experimental/laboratory
phonology. The combination of these diverse methods is rarely employed on a single
language, and this project will serve as a model for subsequent analyses of other
languages.
2 An underdescribed language

Phonological facts of Slovenian are practically unknown in theoretical linguistics.

There are several reasons for such a situation. First, only a handful of publications are
available in widely spoken languages (e.g. Lehiste 1961, Srebot Rejec 1988, Šuštaršič
et al. 1995, Jurgec 2010a). Second, the existing literature deals only with rather
fundamental issues (e.g. the inventory of sounds) and is typically unaware of more
recent methodologies. To make things worse, there is also no comprehensive
phonological description of any variety of Slovenian. Third, even the existing
literature is sometimes lacking in quality and insight. The best way to illustrate this is
to look at one recently discussed issue. When linguists try do describe a language, one
of the very first priorities is to determine the inventory of segments. Standard
Slovenian has been described as having eight vowels (Toporišič 1976/2000, Srebot
Rejec 1998, Šuštaršič et al. 1999). However, Jurgec 2011 has shown that Slovenian
actually has an additional vowel. The fact that such a conclusion has not been reached
much earlier illustrates that Slovenian lacks even the very basic description.

This project provides a comprehensive and informed description of Slovenian. The

data will be collected primarily using contemporary methods, which includes spoken
corpora, speech recordings, and digital speech analysis. The description is
complemented by a theoretical analysis, based on the current theoretic approaches,
which includes Optimality Theory (Prince and Smolensky 1993/2004) and feature
theory (Myers 1997, Moren 2001, 2006, Hyde 2002, Blaho 2008, Jurgec 2010b). The
main result is a comprehensive, monograph-sized phonological description, which
would contribute to our understanding of the sound patterns of Slovenian.

3 A unique language

As we have just seen, Slovenian sound patterns lack a comprehensive and detailed
description. Some recent literature, however, has hinted that Slovenian might have
several cross-linguistically rare phenomena. I now look at three such patterns.

3.1 Loanwords

Loanwords constitute an important part of any language. Often, some sounds are
found in loanwords, but not in native words (for instance, in Dutch, the 'zj' sound in
the middle of garage is only found in French loanwords). Slovenian, however, also
exhibits a different pattern. When a loanword is borrowed in Slovenian, certain
sounds are avoided, even though they are frequent in native Slovenian words (Jurgec
2010a). Vidmar 2011 points out that the Slovenian data show more variation in this
respect.

This project will look at Slovenian loanwords in greater detail. In particular, the
differences between two groups of vowels in loanwords vs. native words will be
fleshed out. Several experiments are planned, and I will now describe one. First, a list
of nonexistent, but possible words with different kinds of vowels is compiled. Some
of these words contain other patterns that we find only in loanwords, whereas others
contain patterns found only in native words. The randomized list of words is fed into
a script that presents one word at a time on a computer screen. The subjects (N=50,
native speakers) are asked to give the plural form of the stimuli, and their answers are
recorded. Since the relevant distinction in vowel quality is not marked in spelling,
there are two possible pronunciations per word. The recordings are examined by three
native speakers, who determine which of the two vowels is pronounced for each
token. The consistently judged words are analyzed statistically (using R and SPSS).
The hypothesis is that speakers prefer some vowels in words containing other
loanword structures, while there is no such preference in words with only native
structures. If the hypothesis is confirmed, the Slovenian data will be important for the
general theory of loanwords (see Calabrese and Wetzels 2009, Kang 2011).

3.2 Tone and vowels

Slovenian exhibits interactions between vowel quality and tone (Becker and Jurgec
2009). Tone is a phonological property that refers to the distribution of pitch at the
lexical level. That is, some vowels have higher pitch than others. Slovenian has two
lexical tones (Jurgec 2007). In most languages, tones appear unrestricted on all
vowels. Slovenian, however, falls in a small group of languages in which the tone on
some vowels is predictable (other cases include Tupuri, Cantonese, Shua, Japanese).

The project will examine the Slovenian interactions between tone and vowel quality
in greater detail. The first part is a production experiment. The stimuli are preselected
and randomized words with different tone–vowel combinations. The subjects (N=20)
are asked to pronounce the words in a carrier sentence, and their response is recorded.
The relevant parts of the recordings are then segmented and labeled. In acoustic
phonetics, vowel quality is determined by formants, which are parts of the spectrum
with the greatest intensity. In this experiment, the first two formants are measured by
running a script in Praat (http://www.praat.org). The measurements are analyzed
statistically using SPSS and R.

The second part is a perception study. The stimuli are prepared by recording various
words by two reference speakers. These words are then acoustically manipulated,
with typical tonal patterns superimposed on them (two patterns per word, both
manipulated). The stimuli are randomized and presented to subjects (N=50), who are
asked to rate each word in terms of acceptability (on a 7 grade scale). The results are
then statistically analyzed. The hypothesis is that subjects will rate licit vowel–tone
combinations significantly better than the illicit ones.

3.3 Vowel inventory

Slovenian has five stressed mid vowels, but only three unstressed ones. Jurgec 2006
reports that the quality of unstressed vowels is affected by position. What these data
suggest is the quality of unstressed vowels is determined by the stressed vowel. Such
patterns are frequent in other languages and fall under the term vowel harmony (see
Baković 2000, Krämer 2003, Finley 2008, Mahanta 2008, Archangeli and
Pulleyblank 2007 for overviews). Yet Slovenian seems to be unique in the sense that
vowel harmony applies only to the positions before the stressed vowel, but not after.

To determine the distribution of unstressed vowels, I will perform a production

experiment. The first step is to compile a list of actual words with different
combinations of stressed and unstressed vowels. These words are randomized and
presented to the subjects (N=20). The task is to pronounce the word on the screen.
The recordings are segmented and labeled using Praat, and a script is run to measure
the first formant of the unstressed vowels. The subsequent statistical analysis would
determine to what extent the first formant of different unstressed vowels correlates
with the quality of the stressed vowel.

Why is it important to better understand these unique patterns in Slovenian? One

insight of linguistic theory is that most languages work similarly in display universal
properties. In this context, Slovenian can provide important evidence for how all
languages work. For example, Slovenian loanwords avoid some segments even
though they are possible in native words. The results may hint why that is, and in
what contexts. In short, the Slovenian data can be instrumental in advancing the
general phonological and linguistic theory.

4 An innovative project

4.1 Methods

To describe the phonological patterns as accurately as possible, a combination of

several methods will be employed. In the initial stage, I will set up an online archive
of articles that deal with Slovenian phonology and phonetics. When possible, the
corpus resources will be consulted (Fida, http://www.fida.net, and GOS,
http://www.korpus-gos.net/). Additional data will be collected from a large pool of
native speakers of Slovenian (in the Netherlands and Slovenia).

Specific patterns will be analyzed by means of acoustic analysis using Praat. In

addition, the project will include two sets of phonological experiments. These are
detailed in the previous section.

4.2 Innovative aspects

This project has at least two main contributions. The empirical contribution is to
provide an in-depth description of all known phonological patterns of an understudied
language. To achieve optimal results, I will employ standard tools of digital speech
analysis. The theoretical contribution is to analyze the pattern using contemporary
phonological theory. In particular, this work will specifically look for patterns that are
cross-linguistically rare.

Extensive holistic descriptions of various languages do exist, but their scope is

typically limited to the traditional elicitation or speaker introspection. This project
takes on an integrated approach to data description, combining phonetics,
experimental methods in phonology, and theoretical phonology. I will develop a
brand new methodology that can be replicated by other linguists.

4.3 Dissemination

The main goal of this project is a unified description and analysis of the phonology of
Slovenian, published as a monograph (in Oxford’s The Phonology of World's
Languages). In addition, individual phenomena will be described and analyzed for
publication in top-tier linguistic journals (e.g. Phonology, Linguistic Inquiry,
Language). At least two top-tier publications are planned in this project. The
experimental results are also suitable for publication in phonetic and laboratory
phonology journals (Journal of Phonetics, Phonetica). Some of the results will be
publishable in the journals focusing on Slavic languages (International Journal of
Slavic Linguistics). In addition, this work will be presented at a variety of linguistic
conferences (Old World Conference in Phonology, Manchester Phonology Meeting,
NELS, WCCFL).

4.4 Timing

The author is familiar with all existing literature on Slovenian phonology, thus no
time for literature review is required. Instead, the initial stage will be devoted to an
online literature archive, which will make the literature more available to other
researchers. The remaining stages are outlined in the sections above.

Year 1 1 2 3 4 5 6 7 8 9 10 11 12
Online literature archive
Elicitation, recording
Experiment I
Article I
Experiment II

Year 2 1 2 3 4 5 6 7 8 9 10 11 12
Experiment II
Fieldwork
Experiment III
Monograph writing
Article II

References

Archangeli, Diana & Douglas Pulleyblank (2007). Harmony. In Paul de Lacy (ed.)
The Cambridge Handbook of Phonology, Cambridge: Cambridge University
Press. 353–378.
Baković, Eric (2000). Harmony, dominance, and control. Ph.D. dissertation, Rutgers
University, New Brunswick, NJ.
Becker, Michael & Peter Jurgec (2009). Interactions of tone and ATR in Slovenian.
Ms., University of Massachusetts Amherst and University of Tromsø.
Available on Rutgers Optimality Archive, ROA 995, http://roa.rutgers.edu.
Blaho, Sylvia (2008). The syntax of phonology: a radically substance-free approach.
Ph.D. dissertation, University of Tromsø.
Calabrese, Andrea & Leo Wetzels (eds.) (2009). Loan phonology. Amsterdam: John
Benjamins.
Finley, Sara (2008). Formal and cognitive restrictions on vowel harmony. Ph.D.
dissertation, Johns Hopkins University, Baltimore, MD.
Hyde, Brett (2002). A restrictive theory of metrical stress. Phonology 19. 313–339.
Jurgec, Peter (2006). O nenaglašenih /e/ in /o/ v standardni slovenščini. Slavistična
Revija/Slavic Review Ljubljana 54. 173–185.
Jurgec, Peter (2007). Acoustic analysis of lexical tones in contemporary standard
Slovenian. In Proceedings of the 16th International Congress of Phonetic
Sciences, Saarbrücken: Universität des Saarlandes, vol. 2. 1089–1092.
Jurgec, Peter (2010a). Disjunctive lexical stratification. Linguistic Inquiry 41. 149–
161.
Jurgec, Peter (2010b). Feature Spreading 2.0: A Unified Theory of Assimilation.
Ph.D. dissertation, University of Tromsø.
Jurgec, Peter (2011). Slovenian has nine vowels. Slavistična Revija/Slavic Review
Ljubljana 59. 243–268.
Kang, Yoonjung (2011). Loanword phonology. In Marc van Oostendorp, Colin J.
Ewen, Elizabeth Hume & Keren Rice (eds.) The Blackwell Companion to
Phonology, Malden, MA: Blackwell. 2258–2282.
Krämer, Martin (2003). Vowel harmony and Correspondence Theory. Berlin: Mouton
de Gruyter.
Lehiste, Ilse (1961). The phonemes of Slovene. International Journal of Slavic
Linguistics and Poetics 4. 48–66.
Mahanta, Shakuntala (2008). Directionality and Locality in Vowel Harmony: With
Special Reference to Vowel Harmony in Assamese. Utrecht: LOT.
Moŕen, Bruce (1́999/2001). Distinctiveness, coercion and sonority: a unified theory of
weight. New York: Routledge.
Moŕen, Bruce (2006). Consonant-vowel interactions in Serbian: features,
representations and constraint interactions. Lingua 116. 1198–1244.
Myers, Scott (1997). OCP effects in Optimality Theory. Natural Language and
Linguistic Theory 15. 847–892.
Prince, Alan & Paul Smolensky (1993/2004). Optimality Theory: Constraint
Interaction in Generative Grammar. Malden, MA: Blackwell.
Srebot Rejec, Tatjana (1988). Word Accent and Vowel Duration in Standard Slovene:
An Acoustic and Linguistic Investigation. Munich: Otto Sagner.
Srebot Rejec, Tatjana (1998). O slovenskih samoglasniških sestavih zadnjih 45 let.
Slavistična Revija/Slavic Review Ljubljana 46. 339–346.
Šuštaršič, Rastislav, Smiljana Komar & Bojan Petek (1995). Slovene. Journal of the
International Phonetic Association 25. 86–90.
Šuštaršič, Rastislav, Smiljana Komar & Bojan Petek (1999). Slovene. In Handbook of
the International Phonetic Association, Cambridge: Cambridge University
Press. 135–139.
Toporišič, Jože (1976/2000). Slovenska Slovnica. Maribor: Obzorja.
Vidmar, Carly Ann (2011). Loanword Phonology in Slovene. Master’s thesis, Leiden
University Centre for Linguistics, Leiden.