Atles interactiu de l'entonació del català


The Interactive atlas of Catalan intonation website contains a whole series of oral materials in audio and video format intended to provide the basis for further in-depth studies of Catalan prosody. These materials are organized according to the standard geographical dialect areas designated for Catalan (see. maps of dialects):

  • Alguer Catalan
  • Balearic Catalan
  • Central Catalan
  • Northwestern Catalan
  • Northern Catalan
  • Valencian

We have collected data from 70 different Catalan locales from distinct dialectal areas. Regarding the selection of locales, it was decided to include: a) the capitals of each county or another representative town, and b) population centers that have particular dialectal characteristics.

In every locale we interviewed speakers of different age groups. In the intonation survey and in the Map Task dialogue corpus we interviewed women between the ages of 20 and 35 year with university or higher level education. As for the videotaped interview the speakers included men and women older than 60 years of age. Participants had been born and raised in the locale. A total of 274 speakers have been interviewed.

In terms of content, for each population center selected, data was obtained by means of three main methods:


A. Intonation survey

Examples of different types of utterance (47 contours per locale) were taken from among the responses to an intonation survey. This survey was based on that used by Prieto (2001) and is designed to evoke everyday situations. It is an inductive method in which the researcher presents the subject with a series of situations (such as “You go into a shop you have never been in before and ask the shop assistant if they sell sugar”) and then asks him or her to respond accordingly. This method is especially useful because it allows the researcher to obtain a wide range of intonation contours that are difficult to obtain with other methods.

The Atlas research group adapted the intonation survey to the following dialects:

The survey consisted of approximately 50 situations, each intended to elicit a particular type of utterance. Utterances were organized as follows:

  • Statements
  • Yes-no questions
  • Wh- questions
  • Echo questions
  • Imperatives
  • Vocatives

Subject surveyed were mostly women aged between 25 and 45 years (see “Subject characteristics”).

The intonation pattern for each utterance is described in a file which contains the fundamental frequency contour as measured using the Praat program, the audio file, a description of the melodic curve and an interpretation of the utterance.

Instructions for questionnaire elicitation

The objective of the guided questionnaire is to obtain spontaneous utterances though the oral presentation of contexts which induce different types of utterances and intentions. It is very important that the informants do NOT read the contexts or the possible answers (the answers that appear in the questionnaires are for the researcher's approximate reference and do not have to coincide with the spontaneous responses.) One decisive factor that should be noted before administering the questionnaire to informants is that it should be adapted to the specific dialect under investigation. By making this adaptation, lexical items and sayings that do not exist in that dialect will not appear in the questionnaire, ensuring that possible misunderstandings of the situation are avoided.
Before beginning the questionnaire the researcher should explain to the informant that he/she should attend to the situation and respond in the most natural, spontaneous way possible, as if the situation were actually happening at that moment. The ideal method is for the researcher to explain the discourse context to the informant in a very clear way,  assuring that the context is understood. The reading/explanation of the discourse context (without modifying the situation) guarantees that the same context is always maintained, and in this way it can be used for purposes of comparison with other interviews and across dialects. Another important factor is that when the informant produces each utterance, the researcher is checking to make sure that each utterance type/intention produced is in line with the desired type (interrogative, statement of the obvious, etc.). If the informant produces an utterance which is not of the type desired, the researcher must ask the informant to reformulate their response.


B. Map Task Dialogue Corpus

The Map Task is a validated technique in which two subjects cooperate to complete a specified task. It is designed to cause the subjects to produce particular interrogative patterns. Each of the two subjects has a map of an imaginary town marked with buildings and other specific elements such as fountains and monuments. A route is marked on the map of one of the two participants (see mapa-giver), and that person has the role of the instruction-giver. The version of the same map held by the other participant (see mapa-follower) differs from that of the instruction-giver in that it does not show the route to be followed. The second participant therefore has to ask the instruction-giver questions in order to be able to reproduce the same route on his or her own map.

Algherese Northern Valencian Rest of the dialects
mapa-follower1 mapa-follower1 mapa-follower1 mapa-follower1
mapa-follower2 mapa-follower2 mapa-follower2 mapa-follower2
mapa-giver1 mapa-giver1 mapa-giver1 mapa-giver1
mapa-giver2 mapa-giver2 mapa-giver2 mapa-giver2

71 records have been obtained thus far by means of this method and make up the Catalan Map Task Dialogue Corpus. An audio recording of the performance of the Map Task and its corresponding orthographic transcription (see transcription criteria) can be accessed for each locale. The participating subjects were women aged between 20 and 35 years with a secondary or university education.

This project is part of an international research program whose aim is to collect dialogue corpora from different languages following the Map Task methodology standard (vg. HCRC Map Task Corpus Project). Currently the map task corpora includes data from English, Swedish, Dutch, Italian, Japanese, and Portuguese.


C. Videotaped conversations

For each locale the Atlas group conducted interviews of 10 to 15 minutes. A representative segment of conversation has been chosen for each locale and posted on this website. The segment is available in both audio and video formats (although the video format has restricted access) and the corresponding orthographical transcription (see transcription criteria) is also provided.

The source for the data from the Balearic Islands is the Audio-visual archive of Catalan dialects of the Balearic Islands, a project directed by Jaume Corbera. The Archive contains audio-visual materials representing approximately 100 hours of conversation. They are for the most part conversations among groups of relatives or friends brought together by the researcher. Though participants knew they were being videotaped, the researcher him or herself did not actively participate, and the fact that the conversations took place within an informal group of friends or relatives facilitated the spontaneity and flow of speech. The people recorded for the Archives were 111 men and 73 women, all aged over 60 years.



  • All data, in different formats, are available for free through our website for academic and research purposes.




Atles interactiu de l'entonació del català
Dept. de Traducció i Ciències del Llenguatge
Universitat Pompeu Fabra. 08018 Barcelona

Citation: Prieto, Pilar & Cabré, Teresa (coords.) (2007-2012).
Atles interactiu de l'entonació del català
. Web page: <>.


With funding from

Laboratori d'Informació Geogràfica i Teletecció (LIGIT), UAB

Arxius d'àudio i video
© Arxiu audiovisual dels dialectes catalans de les Illes Balears