The authors of the work to be discussed in this chapter all to a greater or lesser extent reject the argument that the crucial changes in the Western family over the past 500 years can be deduced from a limited range of demographic sources. Indeed the dominant figures seem to take the extreme view that a proper understanding of family history requires us to be mainly concerned not with stability or change in structure but with changes in meanings, not with ‘the family as a reality but the family as an idea’ [8: 7] or as Shorter puts it: ‘Many constellations of sentiment… are possible within any given structure, and because the crisis of the family today is a crisis of emotion – of attachment and rejection – it is incumbent upon the family historian to trace the tale of sentiments’ [2: 79].

The literature in this area is dominated by four general surveys of at least a major part of our period: Ariès's Centuries of Childhood [8], Snorter's Making of the Modern Family [2], Stone's Family, Sex and Marriage in England 1500–1800 [4] and (though his work has wider relevance) Flandrin's Families in Former Times [7], (Note also the many parallel approaches in [9] and [10].) Though I shall introduce other material, much of this chapter is concerned with a review of these four works. At a superficial level there are many differences between their analyses; yet their methodologies, interpretations and broad conclusions have much in common and they also share many weaknesses.