The 16th volume of the Approaches to Hungarian series (currently published by John Benjamins) contains 10 papers selected from presentations at the 13th International Conference on the Structure of Hungarian (ICSH13), held in Budapest in June 2017, organized by the (then) Research Institute for Linguistics of the Hungarian Academy of Sciences. The papers presented in this volume discuss various topics from the realm of syntax (synchronic and diachronic), semantics, pragmatics, phonology and phonetics. All topics are related to language phenomena observable in Hungarian, but the description of the data and their theoretical analysis do not stop here. In most papers, the patterns are set against and compared with relevant data from other languages – German, English, Turkish, Hindi, Nepali, Spanish, Arabic and Serbo-Croatian among others – leading to discoveries about language-particular differences as well as cross-linguistic commonalities and to theoretical advances on long-standing research questions. As the following overview of the contents of the volume shows, the book contributes to many currently investigated issues in theoretical linguistics, with some well-known themes related to Hungarian, such as vowel harmony, stress placement and perception, the focus–background subdivision of clauses and syntactic agreement.
Julia Bacskai-Atkari’s contribution, ‘Non-degree equatives and reanalysis: A case study of doubling patterns in German and Hungarian’, looks at the status and the development of non-degree equative markers in German and the historically unrelated Hungarian. Both languages exhibit doubling patterns in which a matrix equative element (als in als wie and oly in oly(an) mint) was reanalyzed, from the matrix clause as the C head of an embedded clause, undergoing a case of relabeling. The author argues that in the case of Hungarian, we can be certain that the combination of the two elements did not correspond to the combination of two existing complementizers, and this finding allows us to interpret the observed changes in the same way in German as well. The paper also shows that degree equatives are more restricted than non-degree equatives when it comes to the use of operators and that syntactic innovations affecting equatives always start from non-degree equatives.
In the paper ‘Anatomy of Hungarian aspectual particles’, Anikó Csirmaz & Benjamin Slade show that one can provide a uniform, templatic meaning to the various readings that the aspectual adverbial még can have: the temporal, comparative, marginality and additive meanings. The differences of meaning are due to the fact that certain aspects of the templatic meaning can vary: the various uses differ in terms of the identity of the scalar argument of the adverbial particle, the set of focus alternatives it operates on (times, degrees or subconstituents), and the ordering relations. The morphological form of this adverbial is also compared to similar expressions in German, Nepali and Hindi, which shows that adverbial elements like még have distinct but related functions in some contexts in other languages as well.
Andrea Deme, Márton Bartók, Tekla Etelka Gráczi, Tamás Gábor Csapó & Alexandra Markó investigate the amount of voicing and the sound quality (expressed in harmonics-to-noise ratio) of the laryngeal fricative /h/, in the paper titled ‘Intervocalic voicing of Hungarian /h/’. On the basis of the analysis of the realizations of /h/ in syllable onsets in words occurring in meaningful sentences, the authors show that the intervocalic voicing of /h/ is a purely phonetic process. The voicing of /h/ is dependent on its position: /h/ shows very little voicing when it appears in post-pausal positions, but it is largely voiced when it stands between two vowels (this is irrespective of word boundaries or pitch-accent). In the latter case, the exact amount of voicing depends on the horizontal tongue position of the flanking vowels.
Tamás Káldi, Levente Madarász & Anna Babarczy, the authors of ‘Contextual triggers of the Hungarian pre-verbal focus structure: A guided production study’, used an online sentence-creation production experiment to find out what kind of contextual factors facilitate the use of preverbal focus in Hungarian. The novel experiment asked participants to create sentences that fit given contexts, by clicking on words that were randomly listed in a cloud. They found that if the context supports identification and contrast, the use of preverbal focus is facilitated. In addition, they also found that in non-contrastive contexts, preverbal focus was more likely to be used when the context evoked explicit or implicit sets, as opposed to not evoking any set, while in the contrast conditions, the presence of sets evoked did not have an effect on the outcome.
Hungarian backness harmony is investigated in the paper ‘Testing variability effects in Hungarian vowel harmony’, by Fanni Patay, Ágnes Benkő, Ágnes Lukács, Péter Rebrus & Miklós Törkenczy. This paper reports experimental work testing the behavior of harmonically mixed stems (containing a back and some neutral vowels) when it comes to vacillation with the front or back alternants of the same suffix. Using a sentence completion task, the team collected information about the variants of existing words combined with four different suffixes (the dative, the instrumental, the elative and the allative suffixes), from 21 native speakers. The results of the statistical analysis confirm the findings of earlier corpus studies and also reveal that the height, the number, the quality and the order of neutral vowels in a stem all play a role in determining a stem’s harmonic behavior. In addition, it also came to light that the four harmonic suffixes do not show uniform behavior, contrary to the majority of earlier accounts of Hungarian vowel harmony.
In the paper ‘With or without the definite article: On the syntax of anaphoric possessor strategies in Hungarian’, György Rákosi investigates the use of the definite articles in possessive constructions with a nominative possessor, reporting the results of a questionnaire survey with 149 Hungarian native speakers. Unlike pronominal nominative possessors, which require a definite article, reflexives and reciprocals show variation among them when it comes to the use of the definite article. The primary reflexive maga is preceded by a definite article, while the complex reflective önmaga as well as the reciprocal egymás preferably occur without the definite determiner. The author’s analysis of these facts follows accounts that treat the definite article as something that delimits the binding domain in noun phrases.
Ádám Szalontai & Balázs Surányi report an experimental investigation of the word order of given constituents (givenness defined as familiar topicality) in the postverbal field, in the paper ‘Word order effects of givenness in Hungarian: Syntax or prosody?’. Using a two-alternative forced-choice experiment in which participants had to decide between a topic–verb–new–given vs. topic–verb–given–new order, they found that the immediately post-verbal position of the given phrase is preferred above the other order. The findings of this paper also show that the new > given order in the postverbal field is not categorically ruled out, though. The authors propose a prosodic account of the observed preference for the given > new order, with the hypothesis that syntactically both types of orders are well-formed. They suggest that textually given elements may be deaccented, and that the end of the clause is aligned with prosodic prominence. This kind of prosodic markedness triggers the word order preference observed.
Krisztina Szécsényi & Tibor Szécsényi put forward a new proposal for the nature of Hungarian agreement patterns with objects in ‘Object agreement and locality in Hungarian: Infinitival complement clauses, second position objects and accusative adjuncts’. They argue that the specific form of person agreement that obtains with 2sg or 2pl objects in the presence of a 1sg subject, the so-called -lak agreement, is an operation distinct from regular object agreement (in definiteness) with a DP-type argument of the verb. Evidence comes from a systematic look at verb types with one or both of these agreement possibilities, which reveals that there is a set of verbs that show no regular object agreement, as they do not allow for accusative complements, but they allow -lak agreement, such as készül ‘prepare’, fél ‘be afraid’ and igyekszik ‘do one's best’. It is shown that -lak agreement is independent of regular object agreement and the availability of accusative case, and that both agreement phenomena are cyclic in nature but in distinct ways.
Irene Vogel analyzes Hungarian word stress in a large, systematically collected corpus in the article ‘Fixed stress as phonological redundancy: Effects on the production and perception in Hungarian and other languages’. Hungarian, being a language with fixed stress that allows for no exceptions is brought in comparison with languages with other stress systems concerning stress production. Statistical assessment of the clarity and the manifestation of stress and its individual acoustic properties shows that Hungarian speakers do not produce a very strong difference between stressed and non-stressed syllables in non-focal words, indicating that the redundancy of predictable stress affects the manifestation of stress. The paper further argues that the nature of predictability plays a role: compared to the results of the same study on two other languages with predictable stress, namely Turkish and Arabic, it emerges that Hungarian stress has the least clear manifestation of all, and that less clear manifestation only occurs in fixed stress languages (Turkish and Hungarian) but not in languages with a predictable but variable stress (Arabic). These findings are different from what we see in stress perception (failing to perceive stress), as in this area only stress predictability is crucial, irrespective of whether stress is fixed or variable.
In their article ‘(Non-)exhaustivity in focus partitioning across languages’, Malte Zimmermann, Joseph P. De Veaugh-Geiss, Swantje Tönnis & Edgar Onea report three parallel experimental studies on focus partitioning constructions in German, English and Hungarian, using an incremental information retrieval paradigm, involving sentence verification and falsification tasks that aimed to test for the existence and the nature of the exhaustivity inference. The experiments showed that exhaustivity inferences are not obligatory with focus-background clefts and definite pseudoclefts, and they are not obligatory in Hungarian sentences with preverbal focus either, across experiments and participants: while there is a rather robust exhaustivity inference under falsification, this inference is not mandatory under verification. This shows that the exhaustivity inference is not a cross-linguistically robust, interpretable feature of focus, contrary to claims found in some previous theoretical works on the topic, and in line with recent proposals that derive the exhaustivity inference as a pragmatic effect.
As the above quick overview of the volume's contents indicates, this publication is about a large range of grammatical phenomena, investigated from different perspectives. In addition to contributing to outstanding issues in their research fields and without exception reflecting on the feasibility of earlier analyses, all contributions are also uniform in that they have a very strong empirical basis: they contain novel observations based on large amounts of data, in many cases experimentally obtained in a controlled manner, using recently developed techniques and procedures. The volume is extremely uniform both when it comes to the high-quality scholarship that the individual papers report about, as well as when it comes to its style and presentation. I have only counted ten small typos on the 233 pages of the volume (none of these causes any problems for understanding), which attests to the extremely careful editorial process the papers have undergone.
In summary, the 16th volume of the Approaches to Hungarian series is a highly welcome addition to the growing body of works dealing with grammatical phenomena that are related to or inspired by scholarship on Hungarian. The implications of the findings of the papers clearly go beyond the study of a single language, as they enhance our knowledge of grammatical phenomena in many languages and open new perspectives on a broad array of long-standing questions.
