About our research

The corpus used for the research consists of three-person conversations from the BEA spontaneous speech database (https://www.nytud.hu/dbases/bea/index.html). The existing annotation was supplemented with labelling of conversational phenomena (background channel signals, simultaneous conversations, speaker changes, etc.).A total of 100 conversation recordings were selected so that two speakers were the same for about half of the cases (the interviewer and the conversation partner, young ladies between 25 and 35 years old), furthermore, certain variables could be reduced, for example, when examining speech adaptation. We welcome researchers and doctoral students who wish to become involved in our interdisciplinary research - either in co-authorship with the research team or using the corpus.

Contact: tarsalgas.elemzes@gmail.com

Pausing strategies in conversations - functions and realizations (Horváth Viktória, Gyarmathy Dorottya, Krepsz Valéria, Gráczi Tekla Etelka, Hámori Ágnes)

10 conversations were selected from the Hungarian Spontaneous Speech Database for the study (the material was 131 minutes long). The results showed that silent pauses appeared more often in grammatically functional positions than within a phrase. Silent pauses occurred more frequently in narrative sections of the conversations than in dialogic sections. Furthermore, the duration of pauses increased towards the end of the sections parallel with the decrease of the speakers' activity. In addition, the duration of pauses were longer at the beginning of the conversational units than in the middle of these sections because of the speech planning processes. The half of the turn takings took place after silent pauses, mainly less than 600 ms, which marks the presence of extensive prediction in triadic conversations. 

The main characteristics of laughter in spontaneous conversations (Krepsz Valéria, Hámori Ágnes, Gyarmathy Dorottya, Horváth Viktória)

The aim of the research was to analyze the occurrence, types and realization of laughs in a 4.5-hour 3-person conversation (1700 data), using a qualitative methodology from a quantitative and pragmatic point of view. About half of the phenomenon occurred as a backchannel. The results proved that in Hungarian three-person conversations, the laughter that appears in different interactional situations is associated with different frequencies and significantly different durations.

Participant roles and the structure of conversation influenced the frequency of the analyzed phenomenon.



The realization of silent and filled pauses with regard to speech style (Gyarmathy Dorottya, Horváth Viktória)

Silent and filled pauses were analyzed in 10 narratives and 10 conversations. The material was 3 hours long and contained 2400 items of pauses. The results showed that the speech style has not influenced the frequency of pauses. However, the duration of pauses was longer in narratives than in conversations. Furthermore, the position of silent and filled pauses has affected their duration. Silent pauses were longer at phrase boundaries than within phrases, while filled pauses were longer between two silent pauses than filled pauses adhered to word(s). 


Conversation, turn-taking and discourse organization: new insights on the interface of phonetics and pragmatics in a Hungarian conversation. A pilot study (Hámori Ágnes, Horváth Viktória)

The aim of this paper was to analyse the organisation of one Hungarian conversation from a complex pragmatic and phonetic view, bringing together phonetic analytic techniques and methods of Conversation Analysis in the framework of a functional pragmatic approach. Research question was: what are the characteristics of the turns and turn-taking system and how phonetic features interact in them with pragmatic aspects. Furthermore, we investigated two less-discussed, but equally important topics as well: the presence and possible factors of extensive prediction in the turn-taking system and the connections of phonetic patterns with global discourse structure units. One recording from Hungarian Spontaneous Speech Database (BEA) was used for the study. Three female speakers participated in the conversation: their age and level of education were the same, and they have known each other for years. The material was manually annotated in Praat, the annotation contained the utterance and the word level of each speaker, the overlapping speech, the turns and turn-takings. On one hand, the frequency, the duration and main pragmatic characteristics of local discourse units were analysed. On the other hand, the global structure and timing of the discourse were analysed in terms of the narrative and interactive discourse periods. The results showed that both short pauses and overlapping speech have been common during the turn-takings. The result also shed lights on the fact that half of the turn-takings were carried out in less than 400 ms; this data refers to the extensive prediction of turn-taking based on syntactic, prosodic and pragmatic features of turns.

Variability of articulation rate in conversations (Kohári Anna, Krepsz Valéria, Bóna Judit, Gráczi Tekla Etelka, Gyarmathy Dorottya, Huszár Anna, Horváth Viktória)

10 three-participated conversations were selected from the Hungarian Spontaneous Speech Database in order to analyse the changes in articulation rate, particularly with regards to the structure of the conversation. The duration of inter-pausal units and turns, the number of syllables in inter-pausal units and turns, and the articulation rate were automatically extracted based on manually annotated labels. There were no significant differences in AR between the speakers, presumably due to their similar age. The majority of speakers produced IPUs with a somewhat faster articulation rate in the mainly narrative periods than in the mainly dialogic periods. We found that the articulation rate of the speakers, who were constant in each conversation, were more similar to their own articulation rate in other conversations than to the articulation rate of the other speakers in the same conversation. The global articulation rate seems to be a relatively constant characteristic of speech irrespectively of the partner and the situation. 

Backchannels in spontaneous conversations (Gyarmathy Dorottya, Krepsz Valéria, Dér Csilla Ilona, Hámori Ágnes, Horváth Viktória)

The aim of our study was the multi-dimensional analysis of backchannels in spontaneous conversations, firstly in Hungarian. 10 conversations (about 200 minutes) of the Hungarian Spontaneous Speech Database were analyzed with regards to the characteristics of backchannels (duration, density in different parts of the conversation, were they produced in the duration of the partner's silent pause or was it an overlapping). Results showed that backchannels occurred mainly in overlapping positions. The hummings occurred with the shortest duration. The results of complex (pragmatic and phonetic aspects) analysis provide new information on how BCs manage the system of conversations, mainly the realization of turn-takings.


Overlapping speech in tryadic conversations (Horváth Viktória)

Overlapping speech was analyzed with regards to their functional types, duration and occurrence in conversational units based on 10 recordings selected from Hungarian Spontaneous Speech Database. Results showed that overlaps were the most frequent in the case of backchanneling behaviour (realized with the shortest duration in this case). Overlapping speech occurred more frequently at the end of conversational units than at the beginning. 10% of overlaps were followed by turn-taking. Unsuccessful turn-takings were preceded by longer overlaps than successful turn-takings.

Temporal characteristics of silent pauses and breathing: the effects of speakers' age (Gyarmathy Dorottya, Krepsz Valéria) 

This present study investigates the relationship between the pattern of the audible respiration and the occurrence and the duration of silent pauses in two age groups according to their position and function in two speech genres (narrative and conversation). We analyzed how silent pauses and respiration patterns are influenced by their position in spontaneous speech and the age of the speakers based on 20 Hungarian conversations and narratives from young and older adult speakers. Results showed that the strategies of pausing are determined by their functions and speech genre. We empirically and statistically confirmed that the physiological necessity of breathing is subordinated to the speech planning processes.