Classification
Wu's place within the greater scope of Sinitic varieties is less easily typified than protoypically northern Chinese such as Mandarin or prototypically southern Chinese such as Cantonese which are the two main subgroups of Chinese. Its original classification, along with the other Sinitic varieties, was established in 1937 by Li Fang-Kuei whose boundaries more or less have remained the same and were adopted by Yuan Jiahua in his influential 1961 dialect primer^ .
The sole basis of Li's classification was the evolution of Middle Chinese voiced stops. In the original sense, a Wu dialect was by definition one which retained voiced initials. This definition is problematic considering the devoicing process which has begun in many southern Wu dialects which are surrounded by dialects which retain the ancestral voicing. The loss of voicing in a dialect does not entail that its other features are going to suddenly become dramatically different from the dialects it has had long historic ties with. It furthermore would place Old Xiang in this category. So more elaborate systems have developed, but they still mostly delineate the same regions. So regardless of the justification, the Wu region has been clearly outlined, and Li's boundary in some ways has remained the de facto standard.
In Norman's usage, Wu dialects can be considered "central dialects" or dialects which are clearly in a transition zone containing features which typify both northern and southern Chinese. .
Read more about this topic: Wu Chinese