[1] Moosmüller, Sylvia; Schmid, Carolin; Kasess, Christian H. (2016): Alveolar and Velarized Laterals in Albanian and in the Viennese Dialect. In: Language and speech 59 (4), 488–515.
[2] Moosmüller, Sylvia (2012): The roles of stereotypes, phonetic knowledge, and phonological knowledge in the evaluation of dialect authenticity. In: Silvia Calamai, Chiara Celata und Luca Ciucci (Hg.): Proceedings of the Workshop "Sociophonetics, at the crossroads of speech variation, processing and communication". Pisa, Italy: Edizione della Normale, 49–52.
[3] Ernst, Peter (2008): Die Tschechen in Wien und ihr Einfluss auf das Wienerische. Eine kritische Bestandsaufnahme. In: Marek Nekula, Verena Bauer, Albrecht Greule und Ingrid Fleischmann (Hg.): Deutsch in multilingualen Stadtzentren Mittel- und Osteuropas. Um die Jahrhundertwende vom 19. zum 20. Jahrhundert. Wien: Praesens-Verl., 99–107.

Although Austrian German has only a single lateral phoneme /l/ that is described as being alveolar apical, quite some variation has been reported for different local and social varieties. A specific variant of /l/ is the so called "Meidlinger L" named after a district of Vienna, which is typically associated with low-prestige Viennese dialect. Interestingly, while the "Meidlinger L" is frequently part of public debate in the media, little research has been done from a scientific point of view in defining this variant. Velarized and retroflex laterals have been described as variants for /l/ in Viennese dialect depending on phonetic context. Velarized variants may appear word-initially, after (post-)alveolar consonants and back vowels, whereas retroflex variants may appear after labial sounds [1]. While [2] reports that the velarized lateral leads to the identification of speakers as Viennese in general, some (e.g. [3]) suggest the retroflex as the “Meidlinger L”.
The goal of the present study is to provide an empirical basis to the question about the perception of lateral allophones in Austrian German with regard to associations with certain language varieties, and specifically address the association between specific lateral allophones and the concept of "Meidlinger L". The allophones under investigation are the alveolar apical lateral that matches the description of the Austrian standard, as well as the velarized and retroflex allophones since those variants have been suggested as underlying "Meidlinger L ".
Materials were 76 high-frequent German words with laterals in initial, medial, and final position, as singletons as well as in consonant clusters. All words were recorded by a trained phonetician, a female native speaker of (Eastern) Austrian German, who produced all words with the three lateral allophones. Audio recordings were made in a sound-attenuating booth. Concurrent ultrasound tongue imaging was used to monitor the intended articulation. All recordings were phonetically annotated and formant measures were extracted at the lateral midpoint to document the acoustic differences of the allophones. Table 1 (see attached ressource) presents the mean values of the first three formants grouped by allophone.
We will report the results of a perception experiment with approximately 35 participants. In the experiment, all words will be presented with each of the three allophones (i.e., 228 trials). The experiment will ask about the perceived language variety as "Standard German as in Germany", "Standard Austrian", "non-native German", "Viennese", “Meidlinger L”, or "some other Austrian dialect". Demographic questions about the listeners' origin and language use as well as their familiarity with the concept "Meidlinger L" will be asked. Since we will use the recordings without further manipulations except for amplitude normalization, an additional control experiment will assess the perceived naturalness of each production.
We expect that the participants will be able to successfully differentiate the three lateral variants and assign different associations. Velarized and/or retroflex variants are expected to be judged as Viennese or “Meidlinger L”, especially in those word positions where they have been in the literature. The experiment will show whether one of those variants is preferably perceived as "Meidlinger L".