We started out the previous chapter by asking how frequency of experience, memory and attention link together. It has long been known that ‘experience is the stuff of which all learning is made’ (Monsell 1991: 149), and linguists have been keen to collect frequencies of occurrence that provide a rough estimate of the differential amount of linguistic experience an average adult will have had. But the ubiquity of frequency effects does not imply that frequencies should be considered explanations or answers in their own right. Predictability-based approaches to language are one answer to this question, and in Chapter 8 we reviewed work relying on Bayesian belief updating.