The ZISA project (Zweitspracherwerb italienischer, spanischer und portugiesischer Arbeiter, "Second Language Acquisition by Italian, Spanish, and Portuguese Workers") was a research programme initiated in 1974 at the University of Wuppertal, directed by Jürgen Meisel with students Manfred Pienemann and Harald Clahsen, and continued through the late 1970s and early 1980s. It remains the empirical foundation on which Processability Theory was eventually built.
ZISA followed 45 adult naturalistic learners of German with Italian, Spanish, and Portuguese L1s, migrant workers who had received no formal German instruction. The design combined a large cross-sectional study with a longitudinal sub-study of twelve learners recorded over roughly two years. Data came from informal conversational interviews, transcribed and coded for a wide range of morphosyntactic features.
The finding that changed the field was a five-stage developmental sequence for German word order that learners moved through in the same order, regardless of L1:
Each stage was a prerequisite for the next. Instruction could not skip learners ahead; learners who were pushed beyond their current stage either avoided the structure or produced fossilised approximations.
ZISA made three moves that shaped the next thirty years of SLA:
Meisel, Clahsen and Pienemann's (1981) multidimensional model was the first theoretical consolidation of the ZISA findings; Pienemann and Johnston (1987) extended the approach to English as a second language; and Processability Theory (Pienemann 1998) generalised it into a cross-linguistic theory of how the L2 processor drives developmental sequences.
ZISA's design is strong for morphosyntactic sequencing and weak for lexis, phonology, and pragmatics. The learners were adults in naturalistic settings; the project did not test instructed learners, young learners, or typologically distant L1s. Subsequent research has extended the approach across those dimensions, sometimes modifying the original claims, more often reinforcing the core finding that developmental sequences exist and resist simple pedagogic override.