Armenian linguistics

Armenian is an understudied Indo-European language with an odd mix of typological properties. I focus on the various ways morphology and phonology interact within the word. I mostly focus on cyclic phenomena in Armenian, with some excursions into neighboring topics.

For fun, here’s:

General descriptions

I’ve worked on the following more descriptive and more comprehensive works:

Iranian Armenian (Parksahayeren)

Dolatian, Hossep, Afsheen Sharfizadeh, and Bert Vaux (2023) “A grammar of Iranian Armenian: Parksahay or Iranahayeren”. In Languages of the Caucasus. Berlin: Language Science Press. [link, draft, slides]

And also:

  • Slides about the language (presented at UCLA 2022)
  • Slides about documenting the language (presented at UC Berkeley 2023)

Translation of Adjarian 1911 (Armenian dialectology)

A translation about Armenian dialects[book, pdf, slides].

  • Dolatian, Hossep (2024) ” Adjarian’s Armenian dialectology (1911): Translation and commentary”. In Languages of the Caucasus. Berlin: Language Science Press.

Artsakh Online Dictionary

I’m helping to coordinate the digitization of a dictionary of Karabakh Armenian (Սարգսյան – 2013 – Ղարաբաղի բարբարի բառարան). The project is par of the Learn4Artsakh initiative. If you can read Armenian and to want volunteer, please contact me.

Software and data

Morphological analyzer of Western Armenian

Feel free to use our morphological analyzer, made with Apertium [paper, slides, repo]. Some useful tools that you can use are a sentence analyzer and a paradigm generator.

ReRooted: Speech corpus of Syrian Armenian refugee testimonials

The ReRooted Archive is an archive of spoken testimonials by Syrian Armenian refugees. The original archive contains almost 80hrs of transcribed speech. I’m slowly cleaning up transcripts from those recordings [GitHub]. The goal is to make them the archive be suitable as a speech corpus for work on linguistics and NLP. If you’d like to volunteer and help out, please contact me.

Speech corpus of Armenian question-answer dialogues

I helped make a dataset of question-answer pairs for an ongoing intonation study.

Automatic Speech Recognition for Armenian

Bits of the above audio files have been used to make an ASR model for Armenian.

  • Malajyan Arthur, Victoria Khurshudyan, Karen Avetisyan, Hossep Dolatian, and Damien Nouvel. 2024. Bi-dialectal ASR of Armenian from Naturalistic and Read Speech. In Proceedings of the 3rd Annual Meeting of the Special Interest Group on Under-resourced Languages @ LREC-COLING 2024, pages 227–236, Torino, Italia. ELRA and ICCL. [slides]

 

Morphophonology

Cyclic Lexical Phonology of Armenian

My main focus is on documenting the lexical phonology. Armenian has many phonological processes which distinguish between stem-level and word-level morphology. One rampant process is destressed high vowel reduction which shows…

  1. unbounded cyclic application
  2. stem- vs. word-level distinctions
  3. sensitivity to sublexical prosodic constituents, the Prosodic Stem
  4. monotonic variation in application across lexemes and strata

Over the years, I’ve presented and published the data at different venues:

  • [1-2] CLS paper and slides
    Dolatian, Hossep (2018) “Armenian stress: A case for (prosodic) stems.” Proceedings from the Annual Meeting of the Chicago Linguistic Society, vol. 53, no. 1, pp. 39-53. Chicago Linguistic Society, 2017
  • [3]  PLC42 paper (link) and slides
    Dolatian, Hossep (2019) “Cyclicity and prosody in Armenian stress-assignment.” University of Pennsylvania Working Papers in Linguistics: Vol. 25 : Iss. 1 , Article 10.
  • [1-4] an NLLT paper (link, preprint), based on Ch2 of my dissertation
    Dolatian, Hossep (2021) “Cyclicity and prosodic misalignment in Armenian stems. Natural Language & Linguistic Theory 39, 843–886 (2021). https://doi.org/10.1007/s11049-020-09487-7

Bracketing paradoxes in Armenian compounds

Armenian has a simple compounding rule, yet compounds exhibit a bracketing paradox when pluralized. The plural suffix displays two phonologically-conditioned allomorphs: -er after monosyllables, -ner after polysyllables. Although polysyllabic, endocentric compounds are paradoxically pluralized by counting only the second stem. Exocentric compounds are transparently pluralized as polysyllabic bases with -ner. A lot of additional factors confuse this simple picture. I document these factors and show that they require a cyclic approach to compound bracketing paradoxes that is sensitive to semantic heads, morphological heads, and prosodic heads such as the Prosodic Stem.

  1. Paper in Morphology (link, preprint) and slides based on Ch3 of my dissertation
    Dolatian, Hossep (2021) “The role of heads and cyclicity in bracketing paradoxes in Armenian compounds.” Morphology 31, 1–43. https://doi.org/10.1007/s11525-020-09368-0
  2. ConCALL-4 proceedings on optionality in the bracketing paradox (paper, link)
    Dolatian, Hossep (2022) “Variation in a bracketing paradox: A case study in Armenian compounds”. In Proceedings of the 4th Conference on Central Asian Languages and Linguistics (ConCALL-4) (ed. Öner Özçelik and Amber Kennedy), page 95-108

Affix mobility in Armenian dialects

A handful of endangered Armenian dialects show affix mobility. This is when an affix is either a prefix or suffix based on lexical or phonological properties of the base. Based on fieldwork with Nikita Bezrukov, we show that certain Armenian display a mobile affix which is conditioned by phonological, syntactic, and prosodic factors. The data is a potentially extreme case of affix mobility and it is a problem for the Lexical Integrity Condition and other issues in affix allomorphy.

  • [5] PLC paper (link) and AIMM4 slides
    Bezrukov, Nikita and Hossep Dolatian (2020) “Mobile Affixes Across Western Armenian: Conflicts Across Modules,” University of Pennsylvania Working Papers in Linguistics: Vol. 26 : Iss. 1 , Article 6.

Cyclicity and affix truncation in passive verbs

Morphophonological relationships tend to show cyclic depnedences between bases and derivatives. Armenian passives at first seem like a counter-example because passive morphology looks root-based while their phonology is based on active verbs. I resolve this contradiction by using affix truncation. The analysis explains paradoxes in passive phonology.

  • NLLT paper (preprint) and slides
    Dolatian, Hossep (2023) “Cyclic residues of affix deletion in Armenian passive stems”. In Natural Language & Linguistic Theory. https://doi.org/10.1007/s11049-023-09586-1

Schwa epenthesis in orthography and morphophonology

Armenian has simple syllable structure, but the orthography under-represents a substantial number of schwas that are spoken but not written. I argue that as a spelling-pronunciation rule, schwa epenthesis is triggered by directional syllabification. Based on the interaction between epenthesis, cyclicity, and registers, I argue that schwa epenthesis participates in a degree of parallelist phonologically-conditioned allomorphy.

  • PDA paper (link) and slides
    Dolatian, Hossep (2023) “Isomorphism between orthography and underlying forms in the syllabification of the Armenian schwa.” Phonological Data and Analysis, 5(4), 1–48. https://doi.org/10.3765/pda.v5art4.68 Data

Morphology of irregular stress

I document how irregular stress is assigned by prestressing suffixes in Western Armenian. We find that irregular stress is systematically by regular stress under further affixation or enclisis. The end-result is that we argue for cophonologies to get full coverage, with ongoing work for alternatives.

Output-sensitive and long-distance readjustment rules

In Western Armenian, the theme vowel -i- undergoes a `simple’ transformation to the -e- theme vowel. The triggers have two flavors. One is output-sensitive and phonologically-conditioned: the theme vowel changes when unstressed. The other flavor is morphological and long-distance: the past morpheme triggers the readjustment rule. Crucially, the past morpheme can be non-adjacent to the theme vowel, all the way on a separate word.

Syntax-sensitive phonology of floating segments

In Iranian Armenian, the perfective converb suffix ends in a floating or ghost consonant: /-e(l)/ or /-e(ɻ)/. What triggers the consonant to dock or delete depends on both the following auxiliary. Surprisingly, the auxiliary doesn’t have to be adjacent to the suffix, just has to c-command the suffix. This means that docking is a long-distance rule of syntax-sensitive external sandhi.

  • Paper (prep) and slides
    Dolatian, Hossep, Afsheen Sharfizadeh, and Bert Vaux (in prep) “Non-adjacency in the phonosyntax of Iranian Armenian”.

Reducing phrasal allomorphy to floating segments

The Armenian definite suffix displays allomorphy that’s conditioned by the phonological structure of its base, the following clitic, and the following word. Such phrase-level allomorph looks like outward-sensitivity. However, I analyze these facts in terms of latent segments, with data from multiple lects.

Morphology

Long-distance allomorphy in conjugation class

Armenian has a number of conjugation classes which display different types of long-distance allomorphy. I show that the allomorphy displays tier-based locality. Tier-based locality is an under-explored dimension of locality in morphological theory, but it is well-grounded in computational morphology.

  • Linguistic Inquiry paper (preprint, link) with Peter Guekguezian
    Dolatian, Hossep and Peter Guekguezian (2023) “Relativized locality: Phases and tiers in long-distance allomorphy in Armenia”. Linguistic Inquiry. 54 (3): 505–545. https://doi.org/10.1162/ling_a_00456

Morphomic patterns in verbal conjugation classes

I catalog the distribution of the aorist stem in Western Armenian verbs. We determine that it is a morphomic process that varies by conjugation class. It interacts with root suppletion and irregular morphology. It triggers or blocks different types of other morphological rules.

Tense and agreement morphology

I contrast two segmentation strategies for Armenian tense and agreement suffixes. We find that although verbal inflection is fusional on the surface, an agglutinative approach with zero morphs captures different types of morphological regularities, including syncretism and default behavior.

Structure of verbs and theme vowels

I look at the morphological role of theme vowels in Western Armenian verbs. We find that theme vowels are simultaneously conditioned by the root, a covert verbalizer, and voice assignment. But complications and mismatches arise in causatives, passives, compounds, and equipollent verbs.

  • Paper (link) and WCCFL poster with Peter Guekguezian
    Guekguezian, Peter Ara and Hossep Dolatian (2024) “Distributing theme vowels across roots, verbalizers, and voice in Western Armenian verbs”. Proceedings of the 39th West Coast Conference on Formal Linguistics, 313–321. Somerville, MA: Cascadilla Proceedings Project.

Analogical extension of irregular paradigms

I’ve been doing fieldwork on Iranian Armenian (parksahayeren, պարսկահայերէն) which shows a lot of complicated diachronic changes from standard Armenian. We mainly looked at how morphologically-conditioned allomorphs gained more general or more narrow contexts in Iranian. The changes were likely due to analogical extension.

Inheriting allomorphy in ordinals

Cross-linguistically, high ordinal numbers like 21st differ in whether they inherit the suppletive forms of lowering numbers like 1st. I go through a typology of such inheritance across Armenian varieties.  I find that there’s ample variation in inheritance patterns. These allomorphy patterns iconicly line up with rules that refernce linearity or constituency.

  • Paper (preprint, link)
    Dolatian, Hossep (2023) ”Fluctuations in allomorphy domains: Applying Stump 2010 to Armenian ordinal numerals”. Journal of Linguistics.

Stem-based constituents for allomorphy

With Nikita Bezrukov, we look at how the indicative morpheme in Armenian seems to have a convoluted set of conditions for its allomorphy. The morpheme checks that’s attached to a monosyllabic stem that’s not just the root but less than the word.

  • PLC poster
    Bezrukov, Nikita and Hossep Dolatian (2023) “Stuck with stems: Domains in phonologically-conditioned allomorphy in Armenian verbs”. Poster presented at PLC 47, University of Pennsylvania, March 19, 2023.

Phonology-phonetics

Ejectives in Eastern Armenian

In collaboration with Tabita Toparlak, we document the presence of final ejectives in Eastern Armenian. We collected acoustic and articulatory data using an EGG.

  • ICPhS poster and paper
    Toparlak, Tabita & Hossep Dolatian. 2023. Aerodynamics and articulation of word-final ejectives in Eastern Armenian. In Radek Skarnitzl & Jan Volín (eds.), Proceedings of the 20th International Congress of Phonetic Sciences, 3246–3251. Guarant International

Illustration of Armenian phonetics

In collaboration with Scott Seyfarth, Peter Guekguezian, Niamh Kelly, and Tabita Toparlak, we carried out an illustration on the phonetics of Western and Eastern Armenian, for the Illustrations of the IPA.

  • Seyfarth, Scott, Hossep Dolatian, Peter Guekguezian, Niamh Kelly, and Tabita Toparlak. 2023. “Armenian (Standard Eastern and Standard Western)”. Illustrations of the IPA. DOI: 10.1017/S0025100323000130 [paper]

Stress and intonation

Over the years, I’ve dabbled in the phonetics of stress and intonation with different folks. My first attempt at getting my feet wet:

  • Athanasopoulou, Angeliki, Irene Vogel, and Hossep Dolatian (2017) “Acoustic Properties of Canonical and Non-Canonical Stress in French, Turkish, Armenian and Brazilian Portuguese.” In INTERSPEECH, pp. 1398-1402. [proceeding]

More ToBi-style intonation annotation

  • Toparlak, Tabita and Hossep Dolatian (2022) “Intonation and focus marking in Western Armenian”. In Proceedings of the 4th Conference on Central Asian Languages and Linguistics (ConCALL-4) (ed. Öner Özçelik and Amber Kennedy), page 81-94. [poster, paper, link]

More phonetically-oriented data collection and interpretation.

  • Chakmakjian, Samuel and Hossep Dolatian (2022) “Speech corpus of Armenian question-answer dialogues.” [repo]
  • Chakmakjian, Samuel, Hossep Dolatian, and Stavros Skopeteas (2024) “Word stress and prosodic events in Eastern Armenian”. In Proceedings of Speech Prosody 2024, 1170-1174, [poster, paper, link]

Variation in high front vowel /ʏ/ իւ

With Samuel Chakmakjian, I’ve been looking into sociophonetic variation of the vowel /ʏ/ իւ. It has a lot of possible forms, and work is still in progress.

  • NWAV50 slides
    Chakmakjian, Samuel, and Hossep Dolatian (2022) “Variation in front round vowels across Western Armenians from Syria”.

Syntax-phonology interface

Assignment of nuclear or sentential stress

The syntax and prosodic structure Western Armenian is similar to those of Turkish and Persian, but with subtle differences in the placement of nuclear stress. I document these differences, and I develop a descriptive set of constraints to capture these differences.

  • ConCall-4 poster and proceedings (paper, link)
    Dolatian, Hossep (2022) “Interface constraints for nuclear stress assignment under broad focus in Western Armenian vs. Turkish and Persian”. In Proceedings of the 4th Conference on Central Asian Languages and Linguistics (ConCALL-4) (ed. Öner Özçelik and Amber Kennedy), page 59-80

Prosodically-conditioned extraposition

In collaboration with Katherine Hodgson and Tom Meadows, I looked at how relative clauses extrapose in Armenian. Extraposition is primarily triggered by the prosodic phrasing between a word and the main verb. That is, you extrapose the preverbal argument for prosodic optimality.

  • TU8 paper and LACIM slides
    Dolatian, Hossep & Tom Meadows. 2023. Prosodically-conditioned relative clause extraposition in Armenian. Proceedings of the Workshop on Turkic and Languages in Contact with Turkic 8. 44–58. http://dx.doi.org/10.3765/xw9svp45.

Elicitations for syntax-semantics

Although I am primarily a phonologist, I am very interested in providing data for any linguistic phenomenon on Armenian. I have acted as a consultant on the following works in the syntax-semantics of Armenian:

  1. Luisa Martí on the semantics and number marking of numeral-noun constructions: paper
  2. Yağmur Sağ on the semantics of number marking and classifiers: dissertation
  3. Alexandros Kalomoiros on the syntax of agreement in number-noun constructions and pseudo-noun incorporation: slides, SALT article.

Acquisition

Phonological data on child acquisition

With Hasmik Hovhannisyan, we looked at some child data on acquiring Armenian phonology.

  • TU8 paper and poster
    Hovhannisyan, Hasmik & Hossep Dolatian. 2023. Simplification principles and child language development in Armenian. Proceedings of the Workshop on Turkic and Languages in Contact with Turkic 8. 226–240. http://dx.doi.org/10.3765/zkvc5m80.