PHONOLEX is basically a list of canonical pronunciations (most of them in German SAM-PA) but entries may be extended by syntactic, semantic or other markers as well as by lists of empirically found pronunciation variants together with their source and word count.
In the following example the first entry stems from a list of fully inflected words created by the University of Saarbrücken (OR marker), the syntactic word class is nomen (CL marker) and the pronunciation was created automatically using the text-to-phoneme system P-TRA (TP marker). This entry contains no empirical pronunciations because it is based on a generative linguistic system.
The remaining three entries all stem from the Verbmobil project (OR marker) and have therefore no syntactic word class and the canonical pronunciation has been created manually (TP marker). Empirical pronunciations found in the corpus using the automatic segmentation method MAUS are listed below the canonical pronunciation.
"Ubernachtungsgeldes CL:nom OR:sb TP:ptra Qy:b6naxtUNsgEld@s * "Ubernachtungskosten OR:vm TP:manu Qy:b6n'axtUNs#k"Ost@n y:b6naxtUNskOst@n 1 VM MAUS y:b6naxtUNskOsn 1 VM MAUS * "Ubernachtungsm"oglichk OR:vm TP:manu Qy:b6n'axtUNs#m"2:klICk y:b6naxtUsm2:klICk 1 VM MAUS * "Ubernachtungsm"oglichkeit OR:vm TP:manu Qy:b6n'axtUNs#m"2:klICkaIt y:b6naxtUNsm2:klICkaIt 1 VM MAUS y:b6naxtUNsm2:kICkaIt 1 VM MAUS *