Browse results

Hadas Yeverechyahu


The paper presents consonant co-occurrence restrictions in Hebrew, focusing on the influence of the similarity factor. A lexical analysis of Hebrew verbs reveals tendency to avoid similar, close consonants, by showing a highly significant correlation (p<0.0001) between co-occurrence of C1-C2 sequences in the lexicon and similarity factors (based on Frisch et al.’s 2004 model for similarity, adjusted to Hebrew). In other words, the more two consonants are similar to each other, the smaller their chances are to co-occur as C1-C2 in a Hebrew verb. In addition, a major role of place of articulation is observed, such that consonants that share major place of articulation are less likely to co-occur. However, the highly significant correlation between co-occurrences and similarity factors suggests that not only major place of articulation affects the restrictions; otherwise we would wrongly predict no effect in non-homorganic pairs.

Noam Faust


Modern Hebrew is written with the traditional Hebrew orthography, which contains several symbols that refer to guttural sounds. However, the pronunciations corresponding to these symbols in Modern Hebrew are not phonetically guttural. This paper is an exhaustive survey of these realizations. It shows that in many cases, there are reasons to think that even though no sound is produced, there is an underlying segment in the position of the historical guttural, and this segment behaves in a predictable manner. That said, alongside this general pattern, there are some effects related to historical gutturals that must be regarded as morpheme-specific, as well as some idiosyncrasies of the different original gutturals.

Evan-Gary Cohen


The phonology of loanwords often differs from the phonology of native words in various aspects. These differences are evident in the prosodic structure and even the segmental inventory. The differences between the loanword and native phonology, however, are not necessarily stable, and it is often the case that what originated as phonological structures in loanwords which were illicit in the native vocabulary eventually overrode the native norm, bringing about diachronic change to the phonology of the native words. Hebrew is no exception in this respect. The stress system of loanwords differs from that of native words, with the latter’s system undergoing changes inter alia due to the effect of loanwords (e.g. ante-penultimate stress, immobile stress patterns). The licit syllable structure inventory of native Hebrew words has been expanded to include loaned structures (e.g. complex codas, triconsonantal structures), and the phonemic inventory of Hebrew now includes several consonants originating in loanwords (e.g. ʒ and d͡ʒ).

Roey J. Gafter


This paper surveys current research on the sociophonetics of Modern Hebrew, meaning the research of phonetic variation in Hebrew speech that is socially conditioned, or interpreted as socially meaningful. The paper discusses recent methodological and theoretical advances in sociophonetic research on production and perception, and illustrates how these have been implemented in Hebrew and influenced our understanding of Hebrew sociolinguistics. It further highlights a number of key sociolinguistic variables that have received the most attention in quantitative research on segmental variation: the pharyngeal segments (ħ) and (ʕ), the Hebrew rhotic (r), the glottal fricative (h), and the diphthong (ej). The paper concludes with a discussion of future directions and additional variables of interest which have the potential to advance the growing field of Hebrew sociophonetics.

Modern Hebrew stress

Phonology and phonetics

Outi Bat-El, Evan-Gary Cohen and Vered Silber-Varod


The paper provides a comprehensive description of the phonology and phonetics of Hebrew stress. The distribution of the stress patterns draws a categorial distinction between verbs and nouns, and enhances the typologically uncommon disparity between the most common pattern (final stress) and the default pattern (penultimate stress). As the acoustic studies reveal, the main cue for Hebrew stress is duration, though the duration contrast is eliminated between a phrase final unstressed syllable and the preceding stressed syllable. A second important result of the acoustic studies is that there is no evidence for secondary stress.

Noam Faust, Evan-Gary Cohen and Outi Bat-El

Avi Mizrachi


In Hebrew, consecutive obstruents that differ in voicing can be produced with the same voicing. This process is regressive as the first obstruent assimilates to the obstruent following it. It is also an optional process as it is dependent on the rate of speech and the formality of the utterance (Bolozky 1978, 1997). Using elicited data, an acoustic study on intervocalic obstruent sequences in Hebrew (Mizrachi 2016) has recently shown that assimilation-to-voiceless is more frequent than assimilation-to-voiced. Moreover, not only does devoicing assimilation occur more often, but it also occurs progressively—at least in terms of vocal fold vibration—in more cases than suggested in previous work.

Daniel Asherov and Evan-Gary Cohen


In this paper, we provide a detailed description of the phonetic inventory of Modern Hebrew. We systematically review the phonetic contrasts that distinguish among consonants and vowels, and highlight cases of inter-speaker variation. The contrasts are illustrated with data from an ultrasound tongue imaging study of a native speaker of Modern Hebrew. We provide tongue shape comparisons based on the ultrasound recordings, as well as present acoustic data in form of spectrograms and amplitudes. We also occasionally provide quantitative data from previous studies.

Evan-Gary Cohen, Lior Laks and Carmen Savu


This paper investigates manner variation of Israeli Hebrew rhotics with respect to two factors: prosodic position and speaker gender. An acoustic experimental study shows that although the Hebrew rhotic phoneme tends to be a dorsal approximant, it is significantly more likely to undergo fortition in onset position. This fortition is a result of target overshoot, the rhotic subsequently being produced with a greater degree of constriction than that which would have resulted in an approximant, subsequently surfacing as a stop, a fricative, a tap or a trill. Furthermore, in onset position, female speakers show more variation and produce fewer approximants than male speakers.

Shmuel Bolozky


Connected speech is a natural, continuous stream of sounds, as in normal conversation, in which natural phonetically-motivated processes optimize articulation (through assimilation, elision, etc.), but are still constrained by the need to maintain morpho-semantic transparency at the receiving end. This paper discusses connected speech in Israeli Hebrew, based mostly on data from a spoken corpus. The paper starts with consonant assimilation (voicing assimilation and total assimilation), and then concentrates on various reduction phenomena: vowel elision or reduction, elision of consonants (with or without adjacent vowels), and the “expanded” elision of multiple segments and even complete morphemes and words. As long as the entity targeted is frequent enough, it can easily be reconstructed from the context by the hearer. We proceed to discuss prosodic phenomena in connected speech: pre-tonic lengthening, whose scope seems to be expanding today, and rhythmic secondary stress alternation.