The shift (or smock, or chemise) was a basic linen undergarment worn by women throughout the early modern period. Some examples of seventeenth-century shifts survive in museums and collections, and you can find examples and details about a few of them in books like Patterns of Fashion 4 and Seventeenth-century Dress Patterns.
However, so far there is very little surviving physical evidence known to scholarship (i.e., surviving garments we know about – there might be more somewhere for researchers to discover one day!) about the style and construction of shifts during the period of the Restoration in England, i.e. the period dating from the return to the throne of Charles II to the later part of the century. After 1700, more evidence survives, and the construction and style of shifts from the 1700s (the eighteenth century) is much more well understood.
In order to reconstruct an idea of how a shift from the Restoration period would have been constructed, and the details of its design, modern day scholars and costume makers are forced to use other forms of research. To build up a picture, we can analysis large numbers of visual sources like portraits and drawings, which show parts of shifts that can be seen sticking out from underneath sleeves and necklines etc, or portraits of women in erotic styled ‘dishabille’ dress: the deliberately dishevelled ‘undress’ style of popular portraits like those of the King’s mistresses. These portraits often seem to show women wearing nothing but their shift, plus maybe a few interestingly draped fabrics, but we cannot be certain that the linen garments they are wearing were the same shift that would have been worn as a standard undergarment at the time. This is because paintings like these often depicted fancy dress, costumes and imaginary draperies that created a fantasy look, which was fashionable in portraiture of the time. So, it could have been real underwear, but equally, it could have been garments worn specially for the portrait, to look ‘dishevelled’ and alluring. (My personal opinion is that in most cases, these really are ordinary shifts, but without certainty it is important to be careful and remain objective).
The other method by which researchers and makers can build a picture of the Restoration shift is to use practical reconstruction and sewing experimentation, to literally attempt to recreate shifts that can be worn by modern wearers, and test out theories about their design and construction that way. To do this, we can use the ideas gathered from examining large numbers of the kinds of visual sources listed above, along with contextual written records (although, because the period we are looking at is pre-industrial, such written records are fairly sparse and not always that helpful) and other sources of clues like literature (novels, poems, plays etc) to design prototypes for possible shift patterns. We can then make them up in period accurate fabrics (as far as possible), and try wearing them under reconstructions of the kind of outer garments that we definitely DO know people were wearing at the time, and observe the results to see if they match our expectations.
Relevant images and written sources that give us clues about shifts during this period can be found in a very wide ranging variety of locations. I have brought some of them together below, in hopes of working towards building an accurate idea of the way some fashionable shift designs from the Restoration period may have been constructed.