VINKO is a spoken corpus based on crowd-sourced audio recordings that has been designed to provide relevant linguistic information about the minority languages and dialects spoken in the area between Innsbruck and the Po Valley. The corpus contains audio recordings from local languages and varieties spoken in the regions Trentino-Alto Adige/Südtirol, Veneto and Friuli-Venezia Giulia, with particular focus on the so-called 'language contact' between Germanic (Cimbrian, Mòcheno, Tyrolean, Saurano, and Sappadino) and Romance (Ladin, Trentino and Veneto dialects). The data collection took place from June 2017 to May 2023.
Part of the recordings can be visualized on the open-access section “Listen and Explore” of the AlpiLinK project web page, which can be used by participants and anyone interested in exploring the collected data.
URL: https://alpilink.it/vinko/
Contact: vinko@ateneo.univr.it
PROJECT OF EXCELLENCE DiLLS: Digital Humanities applied to Foreign Languages and Literatures 2018-2022, University of Verona Project code: CUP n. B31I18000250006 Funding organization: Ministero dell’Università e della Ricerca;
Project PRIN 2015 “TREiL” (Technologies for Research and Education in Linguistics) 2017-2020, University of Trento (CIMeC/DIPSCO) Project code: Prot. 2015MNX5ZE - CUP n. E72F16001620001 Funding organization: Ministero dell’Università e della Ricerca;
Project AThEME (Advancing the European Multilingual Experience) Project grant: n. 613465 Funding organization: European Seventh Framework Programme for Research, Technological Development and Demonstration (EU)
File structure under each language variety is identical and organized as follows:
VINKO
|--- README.txt
¦
+--- cim_mhn_plo_zah
¦ +--- cim
¦ |--- S0027_cim_U0056.flac
¦ |--- W0098_cim_U0056.flac
¦ |--- T0101_cim_U0056.flac
¦ |...
¦ +--- plo
¦ |--- ... equivalent to "cim"
¦ +--- mhn
¦ |--- ... equivalent to "cim"
¦ +--- zah
¦ |--- ... equivalent to "cim"
¦
+--- lldan_lldfa_lldfo
¦ +--- lldan
¦ |--- S0027_lldan_U0056.flac
¦ |--- W0098_lldan_U0056.flac
¦ |--- T0101_lldan_U0056.flac
¦ |...
¦ +--- lldfa
¦ |--- ... equivalent to "lldan"
¦ +--- lldfo
¦ |--- ... equivalent to "lldan"
¦
+--- lldba_lldgh
¦ +--- lldba
¦ |--- S0027_lldba_U0056.flac
¦ |--- W0098_lldba_U0056.flac
¦ |--- T0101_lldba_U0056.flac
¦ |...
¦ +--- lldgh
¦ |--- ... equivalent to "lldba"
¦
+--- tir_S01 (from S0001 to S0089)
¦ +--- tir
¦ |--- S0001_tir_U0392.flac
¦ |--- S0001_tir_U0360.flac
¦ |--- S0089_tir_U0391.flac
¦ |...
+--- tir_S02 (equivalent to "tir_01", from S0090 to S0159)
+--- tir_T01 (from T0101 to T0313)
¦ +--- tir
¦ |--- T0101_tir_U0392.flac
¦ |--- T0101_tir_U0360.flac
¦ |--- T0308_tir_U0391.flac
¦ |...
+--- tir_T02 (equivalent to "tir_01", from T0314 to T0403)
+--- tir_W01 (from W0394 to W0449)
¦ +--- tir
¦ |--- W0394_tir_U0392.flac
¦ |--- W0394_tir_U0360.flac
¦ |--- W0448_tir_U0391.flac
¦ |...
+--- tir_W02 (equivalent to "tir_01", from W0450 to W0464)
¦
+--- tre_S_T
¦ +--- tre
¦ |--- S0115_tre_U0077.flac
¦ |--- S0115_tre_U0169.flac
¦ |--- T0316_tre_U0607.flac
¦ |...
+--- tre_W
¦ +--- tre
¦ |--- W0001_tre_U0077.flac
¦ |--- W0001_tre_U0169.flac
¦ |--- W0025_tre_U0607.flac
¦ |...
¦
+--- vec_S01 (from S0004 to S0037)
¦ +--- vec
¦ |--- S0004_vec_U0307.flac
¦ |--- S0004_vec_U0515.flac
¦ |--- S0006_vec_U0449.flac
¦ |...
+--- vec_S02 (equivalent to "vec_S01", from S0037 to S0051)
+--- vec_S03 (equivalent to "vec_S01", from S0052 to S0070)
+--- vec_S04 (equivalent to "vec_S01", from S0071 to S0079)
+--- vec_S05 (equivalent to "vec_S01", from S0080 to S0088)
+--- vec_S06 (equivalent to "vec_S01", from S0098 to S0111)
+--- vec_S07 (equivalent to "vec_S01", from S0112 to S0119)
+--- vec_S08 (equivalent to "vec_S01", from S0120 to S0129)
+--- vec_S09 (equivalent to "vec_S01", from S0130 to S0140)
+--- vec_S10 (equivalent to "vec_S01", from S0141 to S0147)
+--- vec_S10 (equivalent to "vec_S01", from S0148 to S0153)
+--- vec_T01 (from T0101 to T0104)
¦ +--- vec
¦ |--- T0101_vec_U0307.flac
¦ |--- T0101_vec_U0515.flac
¦ |--- T0102_vec_U0449.flac
¦ |...
+--- vec_T02 (equivalent to "vec_T01", from T0105 to T0110)
+--- vec_T03 (equivalent to "vec_T01", from T0111 to T0115)
+--- vec_T04 (equivalent to "vec_T01", from T0116 to T0202)
+--- vec_T05 (equivalent to "vec_T01", from T0203 to T0302)
+--- vec_T06 (equivalent to "vec_T01", from T0302 to T0305)
+--- vec_T07 (equivalent to "vec_T01", from T0306 to T0309)
+--- vec_T08 (equivalent to "vec_T01", from T0310 to T0314)
+--- vec_T09 (equivalent to "vec_T01", from T0315 to T0403)
+--- vec_W01 (from W0465 to W0474)
¦ +--- vec
¦ |--- W0465_vec_U0307.flac
¦ |--- W0466_vec_U0515.flac
¦ |--- W0467_vec_U0449.flac
¦ |...
+--- vec_W02 (equivalent to "vec_W01", from W0475 to W0484)
+--- vec_W03 (equivalent to "vec_W01", from W0485 to W0493)
+--- vec_W04 (equivalent to "vec_W01", from W0494 to W0501)
¦
+--- Metadata
¦ |--- Users.ods
¦ |--- Sentences.ods
¦ |--- Tales.ods
¦ |--- Words.ods
¦
+--- Images
|--- 001_image.png
|--- 002_image_IT.png
|--- 002_image_DE.png
|...As can be seen, the VinKo Corpus consists of:
1 readme file: contains the main information about the corpus and the VinKo project
1 Metadata folder: contains tables with relevant linguistic information as well as sociolinguistic information about speakers;
1 Images folder: contains the pictures employed as visual context for the morphology section;
34 Audio folders containing raw audio recordings collected from speakers. The audio files were organized based on language variety and stimuli sentence. For ease of dowloading, multiple folders (max size 2GB) were created for Veneto and Tyrolean. In these subfolders, the audio files were split along stimulus IDs (see Sentences, Tales, and Words tables).
The audio files collected in this repository come from two versions of the VinKo web platform which have been implemented over the years: version 1 covers the two-year-span 2017/2018, whereas version 2 covers from 2019 to 2023.
There are audio recordings in eight language varieties for three different levels of linguistic analysis: phonology, morphology, and syntax. The investigation of each linguistic domain involves a different type of linguistic stimulus: single words for phonology; short stories with a visual context for morphology; sentences for syntax. All linguistic domains are available in the different language varieties and they are comparable across these language varieties.
The audio file name always mentions the stimulus ID (e.g. S0027) followed
by the abbreviation of the language variety (e.g., cim) and ending in the
user ID (e.g., U0056). This means that audio file S0027_cim_U0056 is a
Cimbrian translation of stimulus S0027 by speaker U0056. The first letter
of the stimulus ID indicates the linguistic domain under investigation:
Some speakers recorded more than one audio file for the same stimulus. These files are reported in the corpus as follows: S001_cim_U0056a, S001_cim_U0056b.
This section investigates three main phonological phenomena across language varieties:
These phenomena have been investigated within specific contexts. This information is reported in the corresponding table in the metadata folder (Words.ods).
This section investigates the following morphological phenomena:
The specific linguistic variables investigated are reported in the corresponding table in the metadata folder (Tales.ods).
This section investigates the following syntactic topics:
Each of these broad topics includes the analysis of different linguistic variables which are reported in the corresponding table in the metadata folder (Sentences.ods).
This folder contains four tables with the relevant information about the speakers and the linguistic stimuli:
The speaker information includes:
The metadata for the phonology section includes the following information:
The metadata for the syntax section includes the following information:
The metadata for the morphology section includes the following information:
Vita Trentina (06/2017) https://www.vitatrentina.it/2017/06/29/parli-dialetto-fatti-sentire/
La Usc di Ladins (04/2021) https://www.lausc.it/valedes-ladines/fascia/15356-vinko-n-projet-per-valoriser-i-lengac-mendres
Di Sait vo Lusérn (04/2021) http://mediateca.istitutocimbro.it/applications/webwork/site_bibliocim/media/sait_vo_lusern/2021_04_02.pdf
Zimbar Earde (27/03/2021 - min 7:00) https://www.youtube.com/watch?v=U8rBYsAlpa0
RAI Radio1 Interview - Alla sorgente del sapere (10/2021) http://www.urly.it/3gva6
The collected files are raw audio data and some may be missing or empty. If you spot any inconsistency, error, or corrupted recording please contact us at vinko@ateneo.univr.it.
Upload of the audio data collected from January 2022 to May 2023. Number of total audio files from 63863 to 189679
Metadata folder > Users: new participants included (U0901-U1903)
Metadata folder > Sentences: fixed a typo for item S0031 (him > ihm)
Casalicchio/Cordin 2020, Kruijt 2022, Kruijt/Rabanus/Tagliani 2023, Kruijt/Cordin/Rabanus 2023, Rabanus 2023a, Rabanus 2023b, Tomaselli/Bidese 2023 added to 4. Additional information > Scientific publications.
ReadMe file updated to reflect changes in project URL (Description and 4. Additional Information) and Reference in 1. General, and changes in the folder structure of the corpus (3. Data structure).
Addition of Sappadino to the language varieties under investigation
Metadata folder > Words: phonology section now includes new items for Sappadino. For each word, the glosses are provided both in Standard German and Standard Italian.
Metadata folder > Sentences: fixed a typo for item S0031 (him > ihm)
Upload of the audio data collected from June to December 2021. Number of total audio files from 37806 to 63863.
Audio data split into multiple folders for Veneto and Tyrolean due to large amount of data (see #3 Data Structure for explanation about the split)
Metadata folder > Users: new participants included
Metadata folder > Users: column ‘note’ included - here you can find comments on participants and questionnaire administration, e.g., if the participant selected the wrong language variety
User U0648 has been removed from Metadata folder > Users as it contains old linguistic data without precise information about the speaker. The corresponding audio files can still be accessed in Audio folder > lldgh (Gardenese)
RAI Radio1 Interview added to 4. Additional information > Dissemination