About the Concepticon

A paper introducing our Concepticon will be presented at LREC 2016 [PDF]. For more details, see below.

Concept List and Concept Sets

A concept list is a selection of meanings that is deemed interesting by some scholars to compare lexemes between languages. There are very many different reasons why a particular meaning might be included into such a list, and we do not have any preference here for a particular set. The only goal we have here is to link meanings that are found in more than one list, with the goal to be able to compare various datasets, collected on the basis of different concept lists. In practice, we take any concept list, and reduce it to the main information as found in a particular source. Typically, a concept list will have concept labels in one or a few widespread languages, either major scientific languages (like English, Russian, or Spanish) or major languages from the region in which the data is collected (i.e. Hausa for the Chadic list from Kraft (1981)). Furthermore, most concept lists have some kind of numerical identification (ID), sometimes simply an ordering number, which we will also include. Any other information that we consider important will also be extracted from the sources (e.g. semantic field from the World Loanword Database, WOLD).

Most importantly, every concept in every concept list is linked to a concept set, i.e. a set of concepts sharing the same definition. Depending how one looks at it, it is either very hard to define meanings, or very easy. It is very easy, because just anybody can stand up and propose whatever definition s/he wants to define in whatever way deemed interesting. It is very hard to actually come up with definitions that are useful for widespread application across many different languages. In our Concepticon, we link concept sets by assigning simple relations like "broader" and "narrower". Yet even these simple relations can get very complex, as you can see from the network that shows the major kinship relations which are linked to the concept set "SIBLING" at the start page. If no suitably defined concept set exists, we simply add a new one. The combined list of all concept sets is our "Concepticon" in the sense of Poornima and Good (2010).

To each concept set, we add a rough gloss, but this is not supposed to be taken as the definition, just as a convenience abbreviation that offers further clarification as to what concept we try to describe. An attempt to give a more precise definition of each concept set is made by taking definitions from different sources, but also by adding them ourselves, if no valid definition is available. For convenience, we also include semantic fields from the World Loanword Database (extended by us for new meanings that are not included there) and ontological categories. The ontological categories are not supposed to be cross-linguistically comparable, but only a help to better identify the meaning, and as a way to order the different meanings. For many concepts, additional meta-data, including links to BabelNet and OmegaWiki, are offered, will be expanded in future versions.

Contributing

Our Concepticon is an ongoing community effort. Read our contribution guidelines to find out how to become involved.

Acknowledgements

The development of the Concepticon in its current version was supported by the DFG research fellowship grant 261553824 Vertical and lateral aspects of Chinese dialect history (JML) and the ERC starting grant 240816 Quantitative modelling of historical comparative linguistics (JML, MC). As part of the CLLD Project and the Glottobank Project, the work was further supported by the Max Planck Society, the Max Planck Institute for the Science of Human History, and the Royal Society of New Zealand (Marsden Fund grant 13-UOA-121, RF). All support is gratefully acknowledged.

Many people helped us in many ways in assembling the data. They pointed us to missing lists (M), provided scans (S), digitized data (D), typed off and corrected concept lists (C), provided translations (T), linked concept lists (L), or gave important advice (A). For all this help, we are very grateful and express our gratitude to

  • Alexei Kassian (D),
  • Andrew Kitchen (D),
  • Damian Satterthwaite-Phillips (D),
  • Frederike Urke (CLS),
  • Harald Hammarström (DMS),
  • Julia Fischer (SDT),
  • Lars Borin (DL),
  • Martin Haspelmath (AD),
  • Nicholas Evans (A),
  • Paul Heggarty (D),
  • Robert Blust (D),
  • Sean Lee (D),
  • Sebastian Nicolai (CMS),
  • Thiago Chacon (MSD), and
  • Viola Kirchhoff da Cruz (CLS).