Our data is divided into different subsets which were in most cases compiled by one single author. These subsets can correspond to both language families or dialectal varieties. Therefore, the subsets do not always correspond to a language family in the strict sense.

The word "concept" may be misleading in this context, since not all words in a given multiple sequence alignment necessarily denote the same concept. Nevertheless, it is often possible to assign a common concept to a given MSA, and it is a convenient way to search for specific subsets of the data.

Percentage Identity is a standard measure for the diversity of a given multiple sequence alignment (MSA). There are different ways to measure the PID of a given MSA. Here we used the most common approach which divides the number of identical positions in an alignment by the sum of the number of aligned positions and the number of internal gap positions. See Raghava & Barton (2006) for details.

Use this field to narrow down the number of alignments to browse by typing in the name of the language in which you are interested.

Check this box if you want to select only those alignments in which metathesis occurs.