Paper Short Abstract:

The game of Tig, and its counting out rhymes, showcase the Irish Schools’ Collection and the value of a fully transcribed archival collection. Using text-mining and NLP to map and compare rhymes offers new understanding of counting out, bringing it from archival text toward performance and play.

Paper Abstract:

The chasing game variously known as “Tig,” “Tag,” or “Tib” by English speakers is one of the simplest and most widespread children’s games. In 1937, when the student collectors of the Irish National Folklore Commission’s Schools’ Collection were charged with documenting various folklore categories, including “Games I Play,” they produced over a thousand descriptions of Tig. While the game itself is simple, unpacking its play reveals a complex linking of physical, linguistic, and intellectual components, including the oft-studied counting-out rhyme. Thanks to the team at the NFC, duchas.ie, and the crowd-source work of Meitheal Duchas, the collection has been fully digitally transcribed. Now, computational methods from natural language and text processing can be applied to the rhymes, allowing for categorization and comparison. Utilizing and adapting state-of-the-art methods, such as phonetic embeddings, shifts the object of study from the text itself back towards structure, sound, and playground performance. Comparing the many different rhymes reveals geographic distribution, gendered patterns of play, and the ways children collect and utilize a repertoire of rhymes. The computational visualizations provide a broad view of patterns, without either abstracting the performances, or choosing a select few to stand as representative examples. This study furthers the act of un-writing by reimagining the corpus as a landscape of tradition, shaped by geography and structural similarity, rather than presenting only the linear text. Finally, the work highlights the sometimes-overlooked genre of children’s games, by focusing on the oral formulas and poetics of play.