This study is conducted using the NSM (Natural Semantic Metalanguage) methodology, which seeks to explicate complex language-specific concepts into configurations of simple universal concepts (Goddard, 2011; Goddard & Wierzbicka, 2014a; cf. Ye, 2013). The study has three main dimensions. It begins by turning the lens of NSM semantic analysis onto a set of words that are central to the “discourse of the unexpected” in English: surprised, amazed, astonished and shocked. By elucidating their precise meanings, one can gain an improved picture of the English folk model in this domain. A comparison with Malay (Bahasa Melayu) shows that the “surprise words” of English lack precise equivalents in other languages (cf. Goddard, 1997). The second dimension involves grammatical semantics, seeking to identify the semantic relationships between agnate word-sets such as: surprised, surprising, to surprise; amazed, amazing, to amaze. The third dimension is a theoretical one, concerned with the goal of developing a typology of “surprise-like” concepts. It is argued that adopting English-specific words, such as surprise or unexpected, as descriptive categories inevitably leads to conceptual Anglocentrism (Wierzbicka, 2014). The alternative, non-Anglocentric strategy relies on components phrased in terms of universal semantic primes, such as ‘something happened’ and ‘this someone didn’t know that it will happen’, and the like.