the nucleated form. Topsfield, Boxford, Beverly, Wenham, Manchester, Danvers and Reading were all dispersed settlement villages, while only Andover was based on the nucleated form. In many cases, these villages, like Middleton, were simply outlying neighborhoods of the Town of Salem. Then, as in many cases in Massachusetts and Connecticut, "Salem's settlers were simply too widely dispersed from homesteading on great lots . . . not to split into more manageable social, religious and (eventually) political units--villages" (Wood 59). Thus their form was not specifically designed.
Another study conducted by Wood demonstrated the fallacy inherent in the notion that the typical vernacular architectural landscape of eighteenth-century colonial New England consisted of two-story homes of the type that have most commonly been preserved. Though it is readily apparent that surviving houses would represent the higher end of the economic scale, the conventional view has held that these houses predominated throughout colonial New England. This has led to grossly exaggerated notions of the architecture of the region, such as Embury's contention that in eastern Massachusetts, "there was practically no building of down right ugliness and there were a very great many of consummate beauty" (185). Such misconceptions were based on ideas stemming from the practice of "the evolutionary concept of history so popular in the early twentieth century" (Garvan 117). The conventional view was Frederick Kelly's "logical evolutionary sequence," based on the study of house plans (cited in Morrison 22). Kelly held that in Connecticut "the two-room plan prevailed up to 1650; that from 1650 to 1675 it shared honors with the added lean-to type; and that from 1675 to 1700 the original lean-to became general" (Kelly cited in Morrison 22). This idea was widely accepted for all of New England except the most western edges.
But, as Wood's careful study of th...