This paper is concerned with the lexical-syntactic intensification (of adjectives) in Romansh and puts a specific emphasis on intensifiers that have been borrowed from (Swiss) German (e.g. quei ei krass tup,this is very stupid’). Intensifiers scale other elements upwards or downwards (Kirschbaum, 2002; Paradis, 1994). It is well known that Romansh varieties are influenced by contact to Swiss German on multiple linguistic levels (Solèr, 1997; Weinreich, 2011 [1951]). Thus, it is to be expected that certain innovations to the Romansh varieties’ intensification systems are contact-induced (for evidence on Ladin see Fiorentini & Sansó, 2017).

Using corpus linguistic and variationist methods, we explore which (Swiss) German intensifiers are to be found in Romansh and how they relate to “native” intensifiers (e.g. whether there are differences in the syntactic contexts or in the expressions that are typically modified). We draw on three (non-tagged) sub-corpora: a) Swiss SMS-corpus (29,473 Romansh tokens) (Stark et al., 2009–2014), b) Swiss WhatsApp-corpus (262,689 Romansh tokens) (Stark et al., 2014–2020) and c) our own corpus including transcripts of semi-structured interviews (25,000 tokens). We base our corpus-query on a list containing more than 50 frequently occurring (Swiss) German intensifiers based on Pheiff (submitted) to which we added Romansh equivalents (in all idioms) and examples found in the idioms’ grammars (e.g. Spescha, 1989).

Preliminary results show that echt, huere, krass and voll are among the most frequently occurring intensifiers borrowed from (Swiss) German. We find further parallels such as the Graecism mega featuring as prominent intensifier both in the Romansh and (Swiss) German intensification system (Pheiff, submitted). We discuss these parallels and the question as to why only some intensifiers are borrowed while others are not. This research will contribute to an empirical basis for language instruction: Intensification is a language universal that requires specific instruction (Lorenz, 1999).