Kennen Sie schon … PHOIBLE?

PHOIBLE ist eine Samm­lung pho­nol­o­gis­ch­er Dat­en:

PHOIBLE is a repos­i­to­ry of cross-lin­guis­tic phono­log­i­cal inven­to­ry data, which have been extract­ed from source doc­u­ments and ter­tiary data­bas­es and com­piled into a sin­gle search­able con­ve­nience sam­ple. Release 2.0 from 2019 includes 3020 inven­to­ries that con­tain 3183 seg­ment types found in 2186 dis­tinct lan­guages.

A bib­li­o­graph­ic record is pro­vid­ed for each source doc­u­ment; note that some lan­guages in PHOIBLE have mul­ti­ple entries based on dis­tinct sources that dis­agree about the num­ber and/or iden­ti­ty of that language’s phonemes.

Two prin­ci­ples guide the devel­op­ment of PHOIBLE, though it has proved chal­leng­ing both the­o­ret­i­cal­ly and tech­no­log­i­cal­ly to abide by them:

1. Be faith­ful to the lan­guage descrip­tion in the source doc­u­ment (now often called ‘doculect’, for rea­sons indi­cat­ed above)
2. Encode all char­ac­ter data in a con­sis­tent rep­re­sen­ta­tion in Uni­code IPA

In addi­tion to phoneme inven­to­ries, PHOIBLE includes dis­tinc­tive fea­ture data for every phoneme in every lan­guage. The fea­ture sys­tem used was cre­at­ed by the PHOIBLE devel­op­ers to be descrip­tive­ly ade­quate cross-lin­guis­ti­cal­ly. In oth­er words, if two phonemes dif­fer in their graphemic rep­re­sen­ta­tion, then they nec­es­sar­i­ly dif­fer in their fea­t­ur­al rep­re­sen­ta­tion as well (regard­less of whether those two phonemes coex­ist in any known doculect). The fea­ture sys­tem is loose­ly based on the fea­ture sys­tem in Hayes 2009 with some addi­tions drawn from Moisik & Esling 2011.

How­ev­er, the final fea­ture sys­tem goes beyond both of these sources, and is poten­tial­ly sub­ject to change as new lan­guages are added in sub­se­quent edi­tions of PHOIBLE.

The data set also includes addi­tion­al genealog­i­cal and geo­graph­i­cal infor­ma­tion about each lan­guage from Glot­tolog.

The PHOIBLE project also inte­grates the the­o­ret­i­cal mod­el of dis­tinc­tive fea­tures from an extend­ed phono­log­i­cal fea­ture set based on Inter­na­tion­al Pho­net­ic Alpha­bet (Asso­ci­a­tion 2005) and on Hayes 2009. This is accom­plished by cre­at­ing a map­ping rela­tion­ship from each IPA seg­ment to a set of fea­tures (Moran 2012). In this way, the IPA is a piv­ot for inter­op­er­abil­i­ty across all resources in PHOIBLE because their con­tents are encod­ed in Uni­code IPA.

For a detailed descrip­tion of PHOIBLE, see Moran 2012. For exam­ples of some of the research we are doing with PHOIBLE, see: Moran et al. 2012, Cysouw et al. 2012, McCloy et al. 2013 and Moran & Blasi, Cross-lin­guis­tic com­par­i­son of com­plex­i­ty mea­sures in phono­log­i­cal sys­tems, forth­com­ing.

How to use PHOIBLE

Users can browse or search PHOIBLE’s inven­to­ries by click­ing on the tabs „Inven­to­ries“, „Lan­guages“ or „Seg­ments“ above. Data can be down­loaded by click­ing the down­load but­ton . If you use PHOIBLE in your research, please cite appro­pri­ate­ly, fol­low­ing our rec­om­mend­ed cita­tion for­mat.

ent­deckt auf Twit­ter