Metaphorical structure is often easier to expose by starting from source-domain language rather than from target-domain lexemes. Many target domains cannot be thoroughly investigated by searching a corpus for collocates of target lexemes, because not all target language occurs near source-domain triggers. How readily metaphorical structure can be exposed therefore depends on the relationship between source-domain language and the structural character of the target domain. Instead of relying on pre-specified semantic fields, corpus collocation patterns are used to gather candidate metaphor triggers.