Faculty of Language, Literature and Humanities - Corpus Linguistics and Morphology


BeMaTaC – A deeply annotated multimodal map-task corpus of spoken learner and native German


BeMaTaC uses a map-task design, where one speaker (the instructor) instructs another speaker (the instructee) to reproduce a route on a map with landmarks. Before the recordings, both participants fill out a detailed metadata survey and give their consent to the release and usage of these data along with the recordings on a separate form.

The recordings take place in a sound booth with the speakers sitting opposite each other. The speakers cannot see each other and are thus unable to communicate non-verbally. During the actual task, there are no other people present in the sound booth. Two separately placed microphones ensure the clarity of both speakers on the recordings. (In current releases, however, only one of the recordings is actually used.)

A sandwich board is placed between the speakers, serving both as a screen and as a surface for the maps. The instructee, who has to draw the route according to the instructor's instructions, is handed a thick black marker pen. One of the recording devices is placed looking over the instructee's shoulder, recording a video of the map and the drawing hand.

Technical details:

  • Olympus LS-11 (for the instructor in the L1 subcorpus)
  • Olympus LS-20M (in all other cases)
  • Audio: 16-bit linear PCM, stereo, 44100 Hz, WAVE format
  • Video: H.264/MPEG-4 AVC, 1280x720, QuickTime format


L1 subcorpus:

  • Native speakers of German

L2 subcorpus:

  • Native speakers of English
  • Advanced learners of German, approximately CEFR level C1/C2
  • Not bilingual with German as one of the native languages
  • Not raised in a German-speaking country or in a German-speaking environment

The exact nature of the task is not disclosed beforehand, however, participants are usually aware that the study is language-related. Participants do not receive monetary compensation or course credit.

L2 participants know that the study is conducted in German. A language proficiency test (onDaF C-Test) takes place after the recordings or at a later date.


The two participants are randomly assigned the roles of instructor and instructee. After taking their positions, the participants are introduced to the sound booth and to their tasks. To ensure consistency, the task description is provided in written form.

Instructor's task: (in German for all participants)

Ihre Aufgabe besteht darin ein Gespräch zu führen, in welchem Sie Ihrem Gegenüber die in Ihre Karte eingezeichnete Route erklären. Ihr Gegenüber hat als Aufgabe, diese Route auf ihrer bzw. seiner Karte zu reproduzieren.

Ihr Gespräch wird dabei mit Tonaufnahmegeräten aufgezeichnet.

Instructee's task: (in German for all participants)

Ihre Aufgabe besteht darin ein Gespräch zu führen, in welchem Ihnen Ihr Gegenüber eine Route erklärt, sodass Sie diese auf der Ihnen vorliegenden Karte einzeichnen können.

Ihr Gespräch wird dabei mit Tonaufnahmegeräten aufgezeichnet. Darüber hinaus wird eine Videoaufnahme angefertigt, welche jedoch ausschließlich Ihre Karte sowie Ihre zeichnende Hand aufnimmt.

The experimenters are available to answer any resulting or remaining questions before the recordings begin. After the task's conclusion, the roles are reversed. Both participants are again handed their respective task descriptions.

The task descriptions are deliberately vague in terms of details (e.g. whether or not the maps' landmarks may be named) in order to elicit a conversation as natural and spontaneous as possible.


The maps each contain nine landmarks, originally chosen to study regional variation in German. The maps are printed in color and in A3 format.

There are two sets of maps, one for each role allocation. Each set consists of the instructor's map, which includes the route, and the instructee's map, which only includes the start and end markers. In addition, one landmark differs between the two maps, this introduces an element of surprise and forces the participants to adopt a different strategy.

After the first task's conclusion, the roles are reversed and a second set of maps is used. The sequential arrangement is fixed.

First set:

map 1 instructor map 1 instructee
instructor instructee

Second set: (after role reversal)

map 2 instructor map 2 instructee
instructor instructee

Last update: 07 May 2014