Wierzbicka's "Primes and Universals" is a summary of her decades of research. For those who have missed out on her earlier works, here it is.
Wierzbicka has a very clever way of dealing with semantics. Her approach (i.e. the Natural Semantic Metalanguage or NSM) is both intuitively simple and intellectual profound.
While many old schools of semantics struggled with the definition of meaning, and resorting to various explanations (e.g. fuzzy nature of meaning etc) to account for their inadequacies, Wierzbicka provided a universal approach to language meaning via NSM. The NSM approach is singularly the only universal approach to meaning. All other approaches are inadequate (e.g. fuzzy set theory, "meaning = use" approach etc), insufficient (e.g. prototype approach, which can only apply to certain groups of words), or irrelevant (e.g. truth conditioning etc). However, it is interesting to note that although the NSM approach is so effective, Wierzbicka's works are not widely referred to in the works of other semanticists (or for the matter, other linguists) especially outside Australia. For example, John Lyons hardly referred to Wierzbicka in his latest "Semantics". Richard Hudson's 1995 "Word Meaning" made no mention at all. Without a universal approach to semantics, it is not surprising that in both cases, the concept of meaning has not been made clear, confusing even.
In the book, Wierzbicka first gives an account of the NSM approach with the list of proposed primitives. She then moves on to discuss the syntax of the NSM (i.e. combinability of the primitives). The remaining chapters are mainly on the explications of words and grammatical structures, with justifications.
Besides being intellectually appealing, Wierzbicka's approach is also accessible for someone with minimal experience with linguistics. One does not require great understanding of other disciplines (e.g. philosophy, syntax, traditional semantic approaches etc) to grasp Wierzbicka's approach.
On the other hand, I think the list of semantic primitives proposed is not the final one. (I mean, I think the principle is intuitively sound but the set of primitives need to be perfected. For example, I do not think "word" is a primitive.) I also think Wierzbicka needs to account for the fact that each primitive is a word (and sometimes 2 like "there is"). We need to see why there are no bound morphemes in the list. Also, I would be very interested to see how Wierzbicka explicates prepositions or postpositions.
I would highly recommend this book to old school semanticists, semantic students of all levels, and even for the interested layman, who will find some chapters very interesting.
I definitely look forward to more findings on the NSM approach to semantics.