Vinča sign corpus (Vinca.xml, from Winn 1981)
The only open machine-readable corpus of Vinča signs: 591 inscribed objects from Turdaș, Vinča, Banjica, Divostin and three minor sites, encoded in XML from Winn's 1981 catalogue by Wu, Solman, Linehan and Sproat. 804 sign tokens, 185 sign types (Winn's numbers), object class and reading order where known. Apache 2.0. Part of Sproat's bundle of eight non-linguistic symbol corpora.
Vinča signsVersion 1.0Licence Apache-2.0Updated 31 Oct 2026Use it
import pandas as pd url = "https://mysteriesdept.com/datasets/DAT0DEDVGHHT4NE/vinca-sign-corpus-vinca-xml-from-winn-1981/v1.0.txt" tokens = pd.read_csv(url) tokens.head()
tokens <- read.csv("https://mysteriesdept.com/datasets/DAT0DEDVGHHT4NE/vinca-sign-corpus-vinca-xml-from-winn-1981/v1.0.txt")
head(tokens)SELECT * FROM read_csv_auto('https://mysteriesdept.com/datasets/DAT0DEDVGHHT4NE/vinca-sign-corpus-vinca-xml-from-winn-1981/v1.0.txt') LIMIT 10;curl -H "Authorization: Bearer $MYSTERIES_TOKEN" -O https://mysteriesdept.com/datasets/DAT0DEDVGHHT4NE/vinca-sign-corpus-vinca-xml-from-winn-1981/v1.0.txt
Used by
0methods
0runs, with their results
0research sessions