Experimental and theoretical examination of vowel harmony patterns (NKFI 119863)

1 Background and research questions

The study of vowel harmony, i.e. the agreement between vowels in the value of a designated feature (the harmonic feature) within a morphologically and/or phonologically circumscribed domain, has been central in phonological research (at least) since the advent of generative phonology (e.g. Halle 1962, Chomsky & Halle 1968). Although the intensity of attention paid to vowel harmony has been constantly high, the focus of research has shifted considerably over the years. Until fairly recently, generative phonological research of vowel harmony in general (in its many varieties, including lexical, autosegmental, government, optimality theoretic, etc. models, cf. van der Hulst & van de Weijer 1995, Archangeli & Pulleyblank 2007, Rose & Walker 2011, Gafos & Dye 2011) concentrated on (i) in terms of data: the categorical (invariable, non-gradual) aspect of vowel harmony patterns, (ii) in terms of analysis: (the formal properties of) phonological models designed to handle the long-distance effects found in vowel harmony patterns with reference to abstract underlying representations (e.g. autosegmental representations, underspecification, floating features, etc.). A further general trait of this research was that it usually did not include psycholinguistic or phonetic experiments or corpus studies to probe the abstract models or devices proposed. Hungarian vowel harmony had a central place in this research due to the coexistence of a number of features it displays (the transparency of neutral vowels, the intricate patterning of alternating and invariant suffixes, the existence of antiharmonic roots, and the combination and very different character of backness and roundness harmony) and it was used as a testing ground for various theories and theoretical devices of generative phonology (a veritable host of alternative abstract analyses exist e.g. Booij 1984, Clements 1976, Hulst 1985, Ringen 1978, 1980, 1988, Vago 1976, 1978, 1980, Rebrus 2000, Ringen & Vago 1995, 1998, Siptár & Törkenczy 2000) In the past two decades there has been a shift in phonological research towards the inclusion of variation and gradient data, experimental evidence, phonetic motivation and surface-oriented explanation (e.g. Hayes, Kirchner & Steriade 2004, Solé, Beddor & Ohala 2007, Bybee & Hopper 2001, Coetzee & Pater 2011). This change has influenced the study of Hungarian vowel harmony too where recent research focussed on (a) the gradient transparency of neutral vowels as manifested in the statistics of variation and/or (experimentally tested) native speaker intuition (Ringen & Miklós Kontra 1989, Hayes & Cziráky Londe 2006, Hayes, Zuraw, Siptár & Londe 2009, Rebrus & Törkenczy 2015), (b) the phonetics of neutral vowels and the connection between surface vowel quality and transparency (e.g. Beňuš, Gafos & Goldstein 2003, Beňuš & Gafos 2005, 2007, Blaho & Szeredi 2013, Szeredi 2010), and (c) the statistics of the lexicon and the antiharmonic behaviour of all-neutral roots. (Hayes, Zuraw, Siptár & Londe 2009, Rebrus & Törkenczy 2015). The proposed research is set within this new paradigm. Specifically, we want to address the following problems (grouped into four “strands” of research below):

(i) NEUTRALITY AND VARIATION IN HARMONY

What makes a vowel neutral in harmony? Exactly how does the gradient character of the harmonic vs. neutral distinction manifest itself? What are the properties that condition variation (lexical variation and vacillation) in Hungarian vowel harmony and in other harmony systems? How do the factors that are known to condition variation in Hungarian vowel harmony interact?

(ii) MORPHOLOGY AND HARMONY

How do morphophonological properties such as paradigmatic relatedness and the alternating vs. invariant character of a suffix influence harmony in Hungarian? What is the relationship between the domain of Hungarian vowel harmony and morphologically identifiable domains. Is domain delimitation categorical?

(iii) THE PHONETICS OF HARMONY

Phonetic effects in antiharmony and vacillation: What is the connection between the phonetic (subphonemic) properties of neutral vowels and variation in their transparency and/or (anti)harmonic behaviour? Neutrality and “distinguishing” dialects: How do the height differences between front unrounded non-high vowels influence vowel neutrality in dialects that distinguish two short front unrounded vowels: mid-high /e/ and mid-low /ɛ/?

(iv) THE TYPOLOGY OF HARMONY

How and to what extent are the key properties of Hungarian vowel harmony (variable transparency, antiharmony, harmonic uniformity, gradience of neutrality, etc.) occur in other (Finno-Ugric/Uralic, Turkic) harmony systems. What is a possible front/back harmony system? Are variable and invariable systems monotonic?

All of these strands of questions have an empirical or “descriptive” aspect in addition to their theoretical or “analytical” significance. In spite of the fact that vowel harmony has been a major issue in phonological research and that it is the most studied aspect of Hungarian phonology there is a lot that is simply not known about it or has been often (over)simplified when addressed. The research questions are designed to investigate these “dark areas”.

The researchers of the project have done previous research to probe into some (aspects) of these questions, especially in the following publications (grouped into strands as above): (i) Fejes 2011, Rebrus, Szigetvári & Törkenczy 2012, Rebrus & Törkenczy 2015cef, Törkenczy 2011, Törkenczy 2013, Törkenczy, Rebrus & Szigetvári 2013; (ii) Rebrus & Szigetvári 2013, 2015, Rebrus & Törkenczy 2015d, Benkő 2014; (iii) Mády 2010, Mády & Bárkányi 2015; (iv) Fejes 2010, Rebrus & Törkenczy 2014ab Rebrus & Törkenczy 2015ab.

2 Hypotheses, key questions and the goals/aims/objectives of the project

The main goal of the project is (a) to gain a fuller understanding of the Hungarian vowel harmony pattern through empirical investigation and by using experimental methods, (b) explore the general properties of front/back (Uralic and some Turkic) harmony systems and (c) develop phonological models/analyses that can account for the key properties (gradience, frequency-dependence, variability) of these harmony phenomena by capturing the surface patterns directly, without reference to abstract underlying representations.

2.1 NEUTRALITY AND VARIATION IN HARMONY

Some vowels behave neutrally in Hungarian front/back (palatal) harmony. Neutrality is a gradient property because there are four neutral vowels (/iː, i, eː, ɛ/) and there are four patterns of neutrality (transparency, suffix invariability, vowel phonotactics in roots, antiharmony) and individual neutral vowels are involved differently in these four patterns. (i) Transparency: front/back harmony can apply long-distance, skipping neutral vowels, which are thus transparent. The gradience here manifests itself chiefly in the variability of transparency: /iː, i/ are fully transparent, there is some variation between transparency and opacity with /eː/ and there is massive variation with /ɛ/ (this is referred to as the “Height Effect” in the literature, cf. Hayes et al. 2009). More than one neutral vowel is also less transparent than a single one (this is referred to as the “Count Effect”, cf. Hayes et al. 2009). (ii) Suffix invariability: a further neutrality pattern is a vowel’s ability to occur in harmonically non-alternating suffixes. It is generally assumed that in a harmony system with neutral vowels those vowels of the inventory are the neutral ones, which cannot harmonically alternate since they lack their harmonic counterparts. Hungarian is interesting/problematic in this respect because, of the four neutral vowels, /eː, ɛ/ are also involved harmonic alternations in suffixes -- on the other hand, almost all of the suffixes that do not alternate harmonically have one of the neutral vowels (except /ɛ/). (iii) Vowel phonotactics in roots. The neutral vowels (which are phonetically front) can also regularly combine with back vowels within the root. (iv) Anti-harmony: Hungarian has front/back anti-harmony: some neutral-vowel roots consistently take an anti-harmonic back suffix alternant rather than the front one while others (the majority of similar roots) behave in the way required by front/back harmony (and take a front suffix) -- most of the relevant roots have/iː, i/, very few have /eː/ and none have /ɛ/. Independently and taken together, the participation of the four vowels in the four neutrality patterns motivate the following scale of neutrality: iː > i > eː > ɛ (where X > Y means X is more neutral than Y).

Our main aims and hypotheses in this strand of research:

Aim: to discover the conditioning factors of variation in Hungarian and develop a model that can accommodate these factors. We hypothesize (based on previous/preliminary research) that in addition to the vocalic shape of stem, other, possibly phonologically “unnatural” factors (Hayes et. al. 2009, Rebrus & Törkenczy 2013, 2015e), e.g. the phonological shape of the suffix, the final consonants, meaning, style, etc. determine variation.
Aim: to investigate how exactly the Count Effect and the Height Effect interact and develop a model to account for the interaction. We hypothesize (based on previous/preliminary research, cf. Rebrus & Törkenczy 2015e) that in addition to cumulative interaction (which is predicted by extant analyses and is an “inbuilt” property of standard derivational phonological models), non-cumulative interaction also occurs and has a frequency based explanation.

2.2 MORPHOLOGY AND HARMONY

Morphology and harmony interact in various ways. Our research aims and hypotheses concerning this interaction:

Aim: to investigate mismatches between morphologically identifiable domains and domains of harmony and develop and analysis/model to account for the the effects that result from these mismatches (e.g. húsvét-ra/re ‘Easter (lit. meat-taking) -SUBL’). We hypothesize (based on preliminary research) that domain delimitation, morpheme separability is gradient (cf. Hay & Baayen 2005) and this results in variation in harmony.
Aim: to investigate how harmonic uniformity (the requirement that an alternating suffix must agree harmonically with the harmonic class of the root) interacts with purely phonological properties driving harmony, sometimes overriding their effect. Like antiharmony, harmonic uniformity shows that front/back harmony is absolutely divorced from its phonetic origins. (e.g. a monomorphemic word with a back vowel followed by two neutral vowels exhibits harmonic variation in the suffixes attached to it, a monomorphemic root with a back vowel, followed by suffixes containing neutral vowels will invariably takes back suffixes (e.g. ɑlibi-rɑ/*rɛ ‘alibi-SUBL’ but mɑdrid-i-rɑ/*rɛ ‘Madrid-ADJ-SUBL’). We hypothesize (based on previous/preliminary research, cf. Rebrus & Szigetvári 2013, 2015; Rebrus, Szigetvári & Törkenczy 2012; Törkenczy, Rebrus & Szigetvári 2013) that harmonic uniformity effects are evidence that front/back harmony and vowel neutrality are (at least partially) divorced from their phonetic origins/motivation.
Aim: to investigate the relationship between suffix invariance and transparency and develop analyses/models that can account for the relationship. We hypothesize (based on previous/preliminary research, cf. Törkenczy 2011; Törkenczy, Rebrus & Szigetvári 2013) that transparent behaviour is associated with non-alternation and that this also extends to the “traditional” Hungarian neutral vowels /iː, i, eː, ɛ/, which are opaque when alternating, unusual behaviour for a neutral vowel (e.g. kolibri-é-na/ek ‘hummingbird-ANAPOSS-DAT’ vs. kolibri-jé-n*a/ek ‘hummingbird-POSS3SG-DAT’). As in 2.2(b) above this suggests that neutrality is not simply derivable from phonetic properties.
Aim: to investigate how variation in Hungarian vowel harmony is influenced by a larger “phrasal” context. It has been reported that the harmonic properties of the suffix of an “antecedent” word can influence the behaviour of vacillating stems: e.g. ebben a hotelben ‘in this hotel’ vs. abban a hotelban ‘in that hotel’ (Kontra, Ringen & Stemberger 1991). We hypothesize (based on previous/preliminary research, cf. Benkő 2014) that this effect is significant and must be accounted for in an analysis.

2.3 THE PHONETICS OF HARMONY

As described in section 2.2 above, Hungarian front/back harmony is gradient across vowel categories across height differences, which manifests itself both in antiharmony and transparency. Our main aims and hypotheses in this strand of research:

Aim: to investigate the subphonemic (phonetic) properties of the neutral vowels involved in antiharmony [iː, i, eː]. The production of Hungarian high /i/ has been studied by Beňuš & Gafos (2003, 2005, 2007), who claim (based on electromagnetic articulography (EMA) studies) that speakers articulated the [i(ː)] vowel with a retracted tongue position in antiharmonic roots (e.g. hiːd 'bridge') compared to non-antiharmonic ones (e.g. hiːr 'news') even when they occur in isolation, i.e. when the retraction is not a coarticulation effect. If this effect is significant it requires an exemplar-based interpretation (e.g. Johnson 1997) as a remaining neighbourhood effect of suffixes frequently appearing next to antiharmonic roots rather than an abstract generative account. We hypothesize that this effect extends to mixed Bɛ type roots (e.g. hɑvɛr ‘friend’, konʦɛrt ‘concert’, hotɛl ‘id.’), which behave in more than one way harmonically. Subphonemic differences in the production and/or the perception of /ɛ/ in mixed stems would be an argument for an exemplar-based rather than a generative account.
Aim: to investigate the Height Effect in dialects that distinguish “close” /e/ and “open” /ɛ/. We hypothesize that the Height Effect extends to this distinction too, resulting in the neutrality relationship e>ɛ in distinguishing dialects. Such a result would suggest that neutrality is (partially) phonetically motivated.

2.4 THE TYPOLOGY OF HARMONY

There is very little information in the “mainstream” “theoretically oriented” literature on (front/back) harmony (e.g. Beňuš 2005, Kiparsky & Pajusalu 2003, Krämer 2003, Nevins 2010) about whether and how (many of) the properties of Hungarian vowel harmony occur in other languages.

Our main aims and hypotheses in this strand of research:

Aim: to investigate how the harmony properties found in Hungarian (discussed in Sections 2.1-2.3 above) occur in Finno-Ugric (especially Erzya, Moksha, Hill Mari and Meadow Mari) and some Turkic vowel harmony systems (especially Uyghur). The properties to be studied are: antiharmony, variation in transparency/opacity, lexical variation, gradience in neutrality, the Height Effect, the Count Effect, harmonic uniformity, gradience in domain delimitation, frequency effects, phonologically unnatural factors conditioning harmony, neutral vowels in invariant/alternating affixes. We want to collect data about these properties in order to test the typological hypotheses proposed in the literature, e.g. the Uniformity principle (according to which all neutral vowels with a given value of the harmonic feature are either all opaque or all transparent, Kiparsky & Pajusalu 2003) and the monotonicity (which requires (among others) that a system with antiharmony must also have transparency, Rebrus & Törkenczy 2015ab).

3 Research methods

We will employ the following methods to answer the research questions and accomplish the objectives identified in Sections 2 and 3.

3.1 CORPUS STUDIES

In each of the four strands of research proposed we plan to carry out detailed corpus studies as well as statistical analysis of the gathered data (type and token frequencies). For Hungarian we intend to use the Szószablya webcorpus (cf. Halácsy et al. 2004), the Hungarian National Corpus, for other languages electronic databases (if available) and Google searches (cf. Zuraw 2006).

3.2 PHONETIC EXPERIMENTS

We will examine bisyllabic mixed stems containing a back vowel and a subsequent /ɛ/. These stems show three different patterns in suffix harmony: back preference (hɑvɛr-ok/*ɛk ‘friend-PL’), front preference (konʦɛrt-ɛk/*ok ‘concert-PL’) or free variation (hotɛl-ok/ɛk ‘id.-PL’). We plan to conduct the following phonetic experiments: (i) /ɛ/ in words with a front/back or vacillating suffix preference will undergo formant analysis. (ii) A listening experiment in which /ɛ/ is manipulated from its maximally open to its maximally closed realisation will be conducted with vacillating stems and nonwords. The question is whether more closed realisations trigger more back suffixes. (iii) According to current observations, /ɛ/ is produced more open by many young speakers, especially in certain environments. After testing this observation experimentally, it will be investigated whether a more open production of /ɛ/ (if present a speaker’s speech) leads to a higher preference of front suffixes. (iv) In a follow-up the same experiment will be conducted with native speakers of Hungarian dialects in Romania (Székely and Csángó) that distinguishes “close” /e/ and “open” /ɛ/ phonemically.

3.3 PSYCHOLINGUISTIC EXPERIMENTS

The effects modifying the basic patterns of vowel harmony have been amply described in theoretical approaches, but so far they gained little support from psycholinguistic experimental work. One of our aims in the project is to empirically test the psychological reality of some of these effects. We plan to investigate empirically how linguistic behavior is driven by the Height Effect, the Count Effect, their interactions and how analogical forces of similarity and frequency act upon them by (i) systematically gathering production data on real words for parts of the paradigms with scarce available corpus data and (ii) by testing predictions in elicited production tasks and grammaticality judgments using nonwords allowing for systematically manipulating the vowel structure of the items with respect to vowel harmony.

3.4 LITERATURE SURVEY

Very little is known about variation in front/back harmony and the harmony properties discussed in section 2 in languages other than Hungarian. We plan to carry out a thorough survey of the descriptive literature on (some key) Uralic and Turkic languages with front/back harmony in order to find data that conform to or falsify typological generalisations.

3.5 PHONOLOGICAL ANALYSIS AND MODELLING

We will conduct a surface-oriented phonological analysis of our findings which does not employ underlying representations, is sensitive to the frequency of usage and uses analogy with existing forms as the central explanatory device (cf. Blevins & Blevins 2009).

4 Expected results

The expected results are partly empirical, partly theoretical. Empirically, the results consist in the systematic collection of reliable data (i) about the phonetics and phonology of front/back harmony and harmony-related phenomena in Hungarian and in other (Uralic and Turkic) languages and (ii) about the typology of front/back harmony. Theoretically, the result is a surface-oriented analysis of (Hungarian) front/back harmony that can handle graduality in data and stochastic information and conforms to general (typological, cross-linguistic) principles of structural organization (monotonicity).

We will create a web-based (i) typological database covering the (front/back) harmony phenomena and neutrality patterns collected from the Uralic and Turkic vowel harmony systems examined and the literature surveyed and (ii) Hungarian dialect database covering the distinguishing dialects examined.

We will publish the results in international publications and disseminate them at international and Hungarian conferences e.g. Manchester Phonology Meeting (MPM), Old World Conference in Phonology (OCP), Annual Meetings on Phonology (AMP), Biennial Conference on Laboratory Phonology (LabPhon), International Conference on the Structure of Hungarian (ICSH).

We plan to organise a vowel harmony workshop on Finno-Ugric harmony systems as part of the XIII International Congress for Finno-Ugric Studies (to be held in Vienna in 2020).

5 Researchers, infrastructure

The researchers participating in the project are the following:

PÉTER REBRUS (principal investigator), senior researcher at the Research Institute for Linguistics of the Hungarian Academy of Sciences; research tasks in the project: collecting empirical data (corpus study, literature), reviewing existing theoretical approaches, developing theoretical models; expertise: formal linguistics, phonology/morphology.

ÁGNES BENKŐ, PhD student at the English Linguistics Doctoral Programme of Eötvös Loránd University Budapest; research tasks in the project: assistance in psycholinguistic and phonetic experiments, conducting and organizing experiments, developing theoretical models; expertise: phonological theory, experiments.

LÁSZLÓ FEJES, research tasks in the project: literature review (Uralic/Turkic data), developing typological databases and models; expertise: Uralic languages.

ÁGNES LUKÁCS, associate professor at the Department of Cognitive Science of the Budapest University of Technology and Economics; research tasks in the project: developing and conducting psycholinguistic experiments, developing theoretical models, expertise: psycholinguistics.

KATALIN MÁDY, senior researcher at the Research Institute for Linguistics of the Hungarian Academy of Sciences; research tasks in the project: developing and conducting phonetic experiments, developing theoretical models; expertise: phonetics, laboratory phonology.

PÉTER SZIGETVÁRI, associate professor at the Department of English Linguistics of Eötvös Loránd University Budapest; research tasks in the project: collecting empirical data (corpus study, literature), reviewing existing theoretical approaches, developing theoretical models; expertise: phonological theory.

MIKLÓS TÖRKENCZY, full professor at the Department of English Linguistics of Eötvös Loránd University Budapest, research professor at the Research Institute for Linguistics of the Hungarian Academy of Sciences; research tasks in the project: collecting empirical data (corpus study, literature), reviewing existing theoretical approaches, developing theoretical models; expertise: phonological theory.

The host institution of the research project (the Research Institute for Linguistics of the Hungarian Academy of Sciences) has the infrastructure necessary for the successful completion of the project such as a non-echoic recording room, the necessary equipment for high-quality recordings (microphones, external sound cards) and sound-proof earphones.

References

Archangeli, Diana & Douglas Pulleyblank. 2007. Harmony. In Paul de Lacy (ed.), The Cambridge Handbook of Phonology, 353–378. Cambridge: Cambridge University Press.

Benkő, Ágnes. 2014. Vacillating stems in Hungarian. In Mark Newson & Péter Szigetvári (eds.) The Even Yearbook 11: ELTE SEAS Working Papers in Linguistics.

Beňuš, Stefan. 2005. Dynamics and transparency in vowel harmony. PhD dissertation, New York University.

Beňuš, Stefan, Adamantios I. Gafos, and Louis Goldstein. 2003. Phonetics and phonology of transparent vowels in Hungarian. In Proceedings of the 29th Annual meeting of the Berkeley Linguistic Society . P. M. Nowak, C. Yoquelet, and D. Mortensen, 485–497. Berkeley Linguistic Society.

Beňuš, Stefan, and Adamantios I. Gafos. 2005. Qualitative and quantitative aspects of vowel harmony: A dynamics model. In Proceedings of the 27th Annual Conference of the Cognitive Science Society. Edited by Bruno G. Bara, Lawrence Barsalou, and Monica Bucciarelli, 226–231. Mahwah, NJ: Lawrence Erlbaum.

Beňuš, Stefan, and Adamantios I. Gafos. 2007. Articulatory characteristics of Hungarian ‘transparent’ vowels. Journal of Phonetics 35, 271–300

Blevins, James P. and Juliette Blevins (eds.). 2009. Analogy in Grammar: Form and Acquisition. Oxford: Oxford University Press.

Booij, Geert. 1984. Neutral vowels and the autosegmental analysis of Hungarian vowel harmony. Linguistics 22. 629–641.

Blaho, Sylvia, and Dániel Szeredi. 2013. Hungarian neutral vowels: a microcomparison. Nordlyd 40.1, 20–40, special issue ‘A Festschrift on the Occasion of X Years of CASTL Phonology and Curt Rice’s Lth Birthday’. Edited by Sylvia Blaho, Martin Krämer, and Bruce Morén-Duolljá. [http://www.ub.uit.no/baser/nordlyd]

Bybee, Joan and Paul Hopper. 2001. Frequency and the Emergence of Linguistic Structure. Amsterdam/Philadephia: John Benjamins.

Chomsky, Noam; Halle, Morris (1968) The Sound Pattern of English. New York: Harper and Row.

Clements, George N. 1976. Clements, G. N. 1976. Neutral vowels in Hungarian vowel harmony: An autosegmental interpretation. Papers from the Annual Meeting of the North East Linguistic Society 7. 49–64.

Coetzee, Andries W. & Joe Pater. 2011. The place of variation in phonological theory. In J Goldsmith, J Riggle & A Yu (eds.), The Handbook of Phonological Theory. 2nd Edition, 401–434. Oxford: Blackwell.

Fejes László. 2011. Miért nem harmonizál a -kor? In Kádár Edit, Szilágyi N. Sándor szerk. Szinkronikus nyelvleírás és diakrónia. Kolozsvár: Erdélyi Múzeum-Egyesület.

Fejes László. 2010. A mari morfológiai elemző néhány tanulsága., Ünnepi írások Bereczki Gábor tiszteletére. Urálisztikai Tanulmányok 19. CD-ROM. Budapest: ELTE BTK Finnugor Tanszék.

Gafos, Adamantios I. & Amanda Dye. 2011. Vowel Harmony: Opaque and Transparent Vowels. In M van Oostendorp, C J Ewen, E Hume & K Rice (eds.), The Blackwell Companion to Phonology 2164–2189, Cambridge, MA & Oxford: Blackwell.

Halácsy, Péter, András Kornai, László Németh, András Rung, István Szakadát, and Viktor Trón (2004). Creating open language resources for Hungarian In: Proceedings of Language Resources and Evaluation Conference (LREC04). LREC, 203–210.

Halle, Morris (1962). Phonology in generative grammar. in J. Katz and J. Fodor (eds.) The Structure of Language, Englewood Cliffs, N. J. Prentice Hall, pp. 334–352.

Hay, J. and Harald Baayen. (2005) Shifting paradigms: gradient structure in morphology. Trends in Cognitive Sciences 9(7): 342–348.

Hayes, Bruce, and Zsuzsa Cziráky Londe. 2006. Stochastic phonological knowledge: the case of Hungarian vowel harmony. Phonology 23. 59–104.

Hayes, Bruce, Robert Kirchner and Donca Steriade (eds). 2004. Phonetically Based Phonology. Cambridge: Cambridge University Press.

Hayes, Bruce, Kie Zuraw, Péter Siptár, and Zsuzsa Londe. 2009. Natural and unnatural constraints in Hungarian vowel harmony. Language 85. 822–863.

Hulst, Harry van der. 1985. Vowel harmony in Hungarian: a comparison of segmental and autosegmental analyses. In Advances in nonlinear phonology. Edited by Harry van der Hulst, and Norval Smith, 267303. Dordrecht: Foris.

Hulst, Harry van der & Jeroen van de Weijer. 1995. Vowel harmony. In J. A. Goldsmith (ed.), The handbook of phonological theory, 495–534. Cambridge, MA & Oxford: Blackwell.

Johnson, Keith. 1997. Speech perception without speaker normalization: an exemplar model. In: K. Johnson & J. W. Mullennix: Talker variability in speech perception. San Diego: Academic Press, 145–166.

Kiparsky, Paul & K. Pajusalu. (2003) Towards a typology of disharmony. The Linguistic Review 20:217–241.

Kontra, Miklós, Catherine O. Ringen & Joseph Paul Stemberger. 1991. The Effect of Context on Suffix Vowel Choice in Hungarian Vowel Harmony. In Proceedings of the Fourteenth International Congress of Linguists. Edited by W. Bahner, J. Schildt, and D. Viehweger, 450–453. Berlin: Akademie-Verlag 1.

Krämer, Martin. (2003). Vowel Harmony and Correspondence Theory. Berlin and New York: Mouton de Gruyter.

Mády, Katalin (2010). Hungarian vowel quantity neutralisation as a potential social marker. Acta Linguistica Hungarica 57/3–4, 167–188.

Mády, Katalin and Bárkányi, Zsuzsanna (2015). Voicing assimilation at accentual phrase boundaries in Hungarian. Proc. International Congress of Phonetic Sciences, Glasgow, ICPHS0796, 5 p. Nevins, Andrew (2010). Locality in Vowel Harmony. Cambridge, Mass: MIT Press.

Rebrus, Péter. 2000. Morfofonológiai jelenségek. [Morphophonology] In Strukturális magyar nyelvtan 3. Morfológia. [A Structural Grammar of Hungarian 3. Morphology.] Edited by Ferenc Kiefer, 763–947. Budapest: Akadémiai Kiadó.

Rebrus, Péter and Péter Szigetvári. 2013. Antiharmony, transparency, truncation. Poster presented at the 21st Manchester Phonology Meeting, 25 May 2013. [http://seas3.elte.hu/szigetva/papers/att.pdf]

Rebrus, Péter and Péter Szigetvári. 2015. Diminutives: Exceptions to the exceptions. Talk delivered to the Exceptionality in Phonology Workshop at the 12th Old-World Conference in Phonology, Barcelona, 28 January 2015. [http://seas3.elte.hu/szigetva/papers/exception-ocp12.pdf]

Rebrus, Péter, Péter SzigetvárI and Miklós Törkenczy. 2012. Dark secrets of Hungarian vowel harmony. In: Sound, Structure and Sense: Studies in memory of Edmund Gussmann, Eugeniusz Cyran—Henryk Kardela—Bogdan Szymanek (ed.) Lublin: Wydawnictwo KUL. 491–508

Rebrus Péter and Törkenczy Miklós. 2013. Magánhangzó-diszharmónia. Talk delivered to MTA Nyelvtudományi Intézet, Budapest on 22 October. [http://budling.nytud.hu/~tork/owncikk/mgh_diszharmonia_ea.pdf]

Rebrus, Péter and Miklós Törkenczy. 2014a Monotonicity and the typology of front/back harmony systems. Poster presented at Eleventh Old World Conference in Phonology (OCP11), Leiden University Centre for Linguistics (LUCL) and the Meertens Instituut Amsterdam, Leiden/Amsterdam, The Netherlands 22–25 January. [http://budling.nytud.hu/~tork/owncikk/Rebrus&Torkenczy_monotonicity_poster_OCP11.pdf]

Rebrus, Péter and Miklós Törkenczy. 2014b Monotonicity and the limits of disharmony. Poster presented at Phonology 2014, Massachusetts Institute of Technology, Cambridge MA 19–21 September [http://budling.nytud.hu/~tork/owncikk/MIT_2014_poster_5hasab.pdf].

Rebrus, Péter and Miklós Törkenczy. 2015a. Monotonicity and the typology of front/back harmony. Theoretical Linguistics 41: 1–2, 1–61.

Rebrus, Péter and Miklós Törkenczy. 2015b. The monotonic behaviour of language patterns. Theoretical Linguistics 41/3–4:241–268.

Rebrus, Péter and Miklós Törkenczy. 2015c. Variation and subpatterns of disharmony in Hungarian. Poster presented at 12th Old World Conference in Phonology (OCP12), Barcelona 27–30 January [http://budling.nytud.hu/~tork/owncikk/OCP_2015_poster_3hasab.pdf].

Rebrus, Péter and Miklós Törkenczy. 2015d. Co-patterns and uniformity in Hungarian vowel harmony. The 12th International Conference on the Structure of Hungarian (ICSH12), Leiden University (Leiden University Centre for Linguistics), Leiden, 22–23 May 2015 [http://budling.nytud.hu/~tork/owncikk/icsh2015_prez_3_short_2x.pdf]

Rebrus, Péter and Miklós Törkenczy. 2015e An ‘unnatural’ pattern of variation in vowel harmony: a frequency-based account. Poster presented at 2015 Annual Meeting on Phonology (AMP 2015), University of British Columbia and Simon Fraser University, Vancouver, Canada 9–11 October. [http://budling.nytud.hu/~tork/owncikk/Vancouver_2015_poster.pdf].

Rebrus, Péter and Törkenczy Miklós. 2015f Abigéllal matinére? A gyakoriság szerepe a hangrendi ingadozásban. [Frequency effects in variation in vowel harmony]. Talk delivered to MTA Nyelvtudományi Intézet, Budapest on 19 November. [http://budling.nytud.hu/~tork/owncikk/nyti-prez.pdf]

Ringen, Catherine. 1978. Another view of the theoretical implications of Hungarian vowel harmony. Linguistic Inquiry 9. 105–115.

Ringen, Catherine. 1980. A concrete analysis of Hungarian vowel harmony. In Issues in vowel harmony. Edited by Robert Vago, 135–154. Amsterdam: Benjamins.

Ringen, Catherine. 1988. Transparency in Hungarian vowel harmony. Phonology 5. 327–342.

Ringen, Catherine, and Miklós Kontra. 1989. Hungarian neutral vowels. Lingua 78. 181–191.

Ringen, Catherine O., and Robert M. Vago. 1995. A constraint based analysis of Hungarian vowel harmony. In Approaches to Hungarian: Levels and structures, vol. 5. Edited by István Kenesei, 309319. Szeged: Attila József University.

Ringen, Catherine O, and Robert M. Vago. 1998. Hungarian vowel harmony in Optimality Theory. Phonology 15. 393–416.

Rose, Sharon & Rachel Walker. 2011. Harmony Systems. In J Goldsmith, J Riggle & A Yu (eds.), Handbook of Phonological Theory. 2nd ed., 240–290. Cambridge, MA: Blackwell.

Siptár, Péter and Miklós Törkenczy. 2000. The phonology of Hungarian. Oxford: Oxford.

Solé, Maria-Josep, Patrice Speeter Beddor and Manjari Ohala (eds). 2007. Experimental Approaches to Phonology. Oxford: Oxford University Press.

Szeredi, Dániel. 2010. Vowel centralization and vowel harmony in Hungarian. In Odd Yearbook 8. Edited by Márton Sóskuthy, 111–137. Budapest: Department of Linguistics Eötvös Loránd University. [http://www.budling.hu/~cash/odd/odd8/OYB8-6-Szeredi.pdf]

Törkenczy, Miklós. 2011. Hungarian Vowel Harmony. In M van Oostendorp, C J Ewen, E Hume & K Rice (eds.), The Blackwell companion to phonology, 2963–2990. Malden, MA & Oxford: WileyBlackwell.

Törkenczy, Miklós. 2013. The close-ups can get rough: exceptional behaviour, transparency and variation in Hungarian vowel harmony. Paper presented at The Twenty-First Manchester Phonology Meeting, Manchester, UK, 23–25 May. [http://budling.nytud.hu/~tork/owncikk/torkenczy_manchester_VH_talk.pdf]

Törkenczy, Miklós, Péter Rebrus and Péter Szigetvári. 2013. Harmony that cannot be represented. In J Brandtler, V Molnár & C Platzack (eds.) Approaches to Hungarian. Volume 13: Papers from the 2011 Lund Conference Volume 13, 229–252. Amsterdam/Philadelphia: John Benjamins.

Vago, Robert. 1976. Theoretical implications of Hungarian vowel harmony. Linguistic Inquiry 7. 243–263.

Vago, Robert. 1978. Some controversial questions concerning the description of Hungarian vowel harmony. Linguistic Inquiry 9. 116–126.

Vago, Robert. 1980. A critique of suprasegmental theories of vowel harmony. In Issues in vowel harmony. Edited by Robert Vago, 155–183. Amsterdam: Benjamins.

Zuraw, Kie. 2006. Using the web as a phonological corpus: a case study from Tagalog. EACL–2006: Proceedings of the 11th Conference of the European Chapter of the Association for Computational Linguistics/Proceedings of the 2nd International Workshop on Web As Corpus. pp. 59-66.