Acquisition processes in maximally diverse languages: min(d)ing the ambient language

Chintang grandmother with child

Children can learn any language. This is astonishing given how different languages are. The main question driving the ACQDIV project [ˈækdɪv] is therefore: How is acquisition possible in spite of the great diversity found in the languages of the world?
Here is an example. The sentence 'Mummy, cook rice for me!' is expressed as follows in Chintang, a small Tibeto-Burman language which forms part of the ACQDIV corpus and which is spoken in the east of Nepal: Ama, kok thuktabidahã!

This simple sentence differs from the English version in many respects:

  • The noun ama does not only mean 'mother' but 'my mother' - it contains a prefix a- that indicates the possessor.
  • The object noun kok precedes the verb ("rice cook").
  • There is no word corresponding to me. Person and number of the beneficient as well as of the cook are coded in the verb (-ahã).
  • There is no word corresponding to for. This is also included in the verb form (-bid).

Nevertheless, children have no difficulties in learning this language or any other of the approximately 7000 languages spoken around the world. How is this possible?

In trying to answer this question, ACQDIV analyses data from ten languages taken from five maximally diverse language clusters. Taking less well-known languages into consideration is crucial for our enterprise: research on language acquisition is heavily biased towards comparatively big European languages, which are all very similar to each other, so data from these languages do not help much in answering the question how children deal with diversity.


  • The ACQDIV team will visit LuCiD at the University of Manchester for a collaborative workshop on the acquisition of morphology (October 2-3). Various team members will give talks.

    More …

  • Damián Blasi​, Jekaterina Mažara​ and Sabine Stoll have organized a workshop at the 50th SLE titled "First language acquisition in the languages of the world: Differences and similarities".

    More …

  • Sabine Stoll, Jekaterina Mažara, Damián Blasi and Balthasar Bickel will present a talk titled "Children use statistical learning in naturalistic language acquisition: evidence from a large longitudinal corpus of a polysynthetic language" at "Interdisciplinary Advances in Statistical Learning" (Bilbao, June 28-30).

    More …

  • Sabine Stoll has been invited to present the ACQDIV project at the ERC's 10th anniversary celebration at CERN on July 6. This is an important networking event which will be attended by the Swiss ERC grantees and Swiss science stakeholders such as the presidents of the universities and the federal institutes of technology or representatives of the SNSF.

    More …

  • Sabine Stoll will present a talk titled "The quest for species-wide acquisition mechanisms: a radical diversity approach" at the workshop "Key questions and new methods in the data sciences" (organized by the MPI for Psycholinguistics, Nijmegen, at Berg en Dal, June 15-18).

    More …