If you use these data please cite

• the original source

  MultiTree: A digital library of language relationships. Institute for Language Information and Technology: Ypsilanti, MI. 2013.

• the derived dataset using the DOI of the particular released version you were using

MultiTree represents the most complete collection of language relationship hypotheses

This dataset is licensed under a CC-BY-SA-4.0 license

From the former MultiTree website:

MultiTree provides a unique approach to historical linguistic research, representing the most complete collection of language relationship hypotheses in a user-friendly, visually-appealing, and interactive format. Not only is it fun and informative, but it is a useful resource that gathers scholarly work and makes it accessible to academics and the public alike.

MultiTree is also an innovative tool for typological analysis, especially among lesser-known languages. It facilitates interdisciplinary collaboration with linguists to reach more accurate conclusions about human language, culture, and history.

The trees in MultiTree are intended to be faithful representations of their sources, but sometimes it can be difficult to capture a scholar's intentions in a graphical representation. Whenever possible, editors have added comments to disambiguate or clarify their interpretations. However, it is always recommended that users refer to the original source for a better understanding of the scholar's hypothesis.

MultiTree aims to collect as many hypotheses about language relationships as possible so that users may compare them. Inclusion of a tree does not indicate validity of the scholar's hypothesis or acceptance by the academic community.

Regarding contact languages (creoles, pidgins, mixed languages) and language isolates Although isolates have no known genetic affiliation, and the origins of contact languages are heavily contested, they have been included in the MultiTree database in order to make information about them available to scholars and to accurately represent whatever hypothesis the original scholar is making. "Trees" that include these languages do not reflect genetic affiliation unless this was the intention of the author.

• 2005-2009 National Science Foundation
• 2012-2015 National Science Foundation, No 1227106
• 2014-2015, National Science Foundation, No 1519050

The following CLDF datasets are available in cldf:

• CLDF Generic at cldf/Generic-metadata.json