The Plymouth Colony Archive Project


Death's Head, Cherub, Urn and Willow

By James Deetz and Edwin S. Dethlefsen


Originally published in
Natural History Vol. 76(3) 1967, pp. 29-37.

dividing bar

An unusual kind of archeological detective work yields new insights into the spread of culture through colonial New England. It also tests the science of archaeology for accuracy.

Enter almost any cemetery in eastern Massachusetts that was in use during the seventeenth and eighteenth centuries. Inspect the stones and the designs carved at their tops, and you will discover that three motifs are present. These motifs have distinctive periods of popularity, each replacing the other in a sequence that is repeated time and time again in all cemeteries between Worcester and the Atlantic, and from New Hampshire to Cape Cod.

The earliest of the three is a winged death's head, with blank eyes and a grinning visage. Earlier versions are quite ornate, but as time passes, they become less elaborate. Sometime during the eighteenth century -- the time varies according to location -- the grim death's head designs are replaced, more or less quickly, by winged cherubs. This design also goes through a gradual simplification of form with time. By the late 1700's or early 1800's, again depending on where you are observing, the cherubs are replaced by stones decorated with a willow tree overhanging a pedestaled urn. If the cemetery you are visiting is in a rural area, the chances are quite good that you will also find other designs, which may even completely replace one or more of the three primary designs at certain periods. If you were to search cemeteries in the same area, you would find that these other designs have a much more local distribution. In and around Boston, however, only the three primary designs would be present.

If you were to prepare a graph showing how the designs change in popularity through time, the finished product might look something like three battleships viewed from above, the lower one with the bow showing, the center one in full view, and the third visible only in the stern. This shape, frequently called a "battleship-shaped" curve, is thought by archaeologists to typify the popularity career of any cultural trait across time. Prepared from controlled data taken from the Stoneham cemetery, north of Boston, where the style sequence is typical of the area around this eighteenth-century urban center of eastern Massachusetts, the graph below shows such a curve.


Click image to view stylistic sequence of gravestones from a cemetery in Stoneham, Massachusetts

Click to View Figure 1

It is appropriate here to interrupt and pose the question: why would an archaeologist study gravestones from a historic period?

Whether archaeology can be considered a science in the strict sense of the word is much debated. One of the hallmarks of scientific method is the use of controls in experimentation that enable the investigator to calibrate his results. Since archaeology deals largely with the unrecorded past, the problem of rigorous control is a difficult one. Much of modern archaeological method and theory has been developed in contexts that lack the necessary controls for precise checking of accuracy and predictive value. For this reason, any set of archaeological data in which such controls are available is potentially of great importance to the development and testing on explanatory models, which can then be used in uncontrolled contexts.

For a number of reasons, colonial New England grave markers may be unique in providing the archaeologist with a laboratory situation in which to measure cultural change in time and space and relate such measurements to the main body of archaeological method. All archaeological data -- artifacts, structures, sites -- can be said to possess three inherent dimensions.

A clay pot, for example, has a location in space. Its date of manufacture and use is fixed in time, and it has certain physical attributes of form. In a sense, much of archaeological method is concerned with the nature and causes of variation along these dimensions, as shown by excavated remains of past cultures.

The spatial aspect of gravestones is constant. We know from historical sources that nearly all of the stones in New England cemeteries of this period were produced locally, probably no more than fifteen or twenty miles away; an insignificant number of them came from long distances. This pattern is so reliable that it is possible to detect those few stones in every cemetery that were made at a more remote town. Once placed over the dead, the stones were unlikely to have been moved, except perhaps within the cemetery limits.

Needless to say, the dimension of time is neatly and tightly controlled. Every stone bears the date of death of the individual whose grave it marks, and most stones were erected shortly after death. Like the spatial regularity, this temporal precision makes it possible to single out most of the stones that were erected at some later date.

Control over the formal dimension of gravestone data derives from our knowledge of the carvers, who, in many instances, are known by name and period of production, and who, even if anonymous, can be identified by their product with the help of spatial and temporal control. Thus, in most cases stones of similar type can be seen to be the product of a single person, and they reflect his ideas regarding their proper form.

Furthermore, it is known that the carvers of the stones were not full-time specialists, but rather workers at other trades who made stones for the immediate population as they were needed. We are dealing, then, with "folk" products as is often the case in prehistoric archaeology.

Other cultural dimensions can also be controlled in the gravestone data with equal precision, and with the addition of these, the full power these artifacts as controls becomes apparent: probate research often tells the price of individual stones; status indication occurs frequently on the stones, as well as the age of each individual. Since death is related to religion, formal variations in the written material can be analyzed to see how they reflect religious variations. Epitaphs provide a unique literary and psychological dimension. Spatial distributions can be measured against political divisions. In short, the full historical background of the seventeenth, eighteenth, and nineteenth centuries permits both primary and secondary control of the material, and with the resulting precision, explanations become quite reliable.

With such controls available to the archaeologist, the pattern of change in colonial gravestone design and style can be used with great effect to sharpen our understanding of cultural process in general. To return to the battleship-shaped curves to the left, what does this mean in terms of culture change? Why should death's heads be popular at all, and what cultural factors were responsible for their disappearance and subsequent rise of the cherub design? The most obvious answer is found in the ecclesiastical history of New England. The period of decline of death's head's coincides with the decline of orthodox Puritanism. In the late seventeenth century, Puritanism was universal in the area, and so were death's head gravestones. The early part of the eighteenth century saw the beginnings of change in orthodoxy, culminating in the great awakenings of the mid-century. In his recent, excellent book on the symbolism of New England gravestones, Graven Images, Allan Ludwig points out that the "iconophobic" Puritans found the carving of gravestones a compromise. While the use of cherubs might have verged on heresy, since they are heavenly beings whose portrayal might lead to idolatry, the use of a more mortal and neutral symbol -- a death's head -- would have served as a graphic reminder of death and resurrection.

Given the more liberal views concerning symbolism and personal involvement preached by Jonathan Edwards and others later in the eighteenth century, the idolatrous and heretical aspects of cherubs would have been more fitting to express the sentiment of the period. It is at this point that available literary controls become valuable. Each stone begins by describing the state of the deceased: "Here lies" or "Here lies buried" being typical early examples. Slowly these are replaced by "Here lies [buried] the body [corruptible, what was mortal] of." This slightly, but significantly, different statement might well reflect a more explicit tendency to stress that only a part of the deceased remains, while the soul, the incorruptible or immortal portion, has gone to its eternal reward. Cherubs reflect a stress on resurrection, while death's heads emphasize the mortality of man. The epitaphs that appear on the bottoms of many stones also add credence to this explanation of change in form over time. Early epitaphs, with death's head designs, stress either decay and life's brevity:

    My Youthful mates both small and great
    Come here and you may see
    An awful sight, which is a type of which
    you soon must be.
or a Calvinistic emphasis on hard work and exemplary behavior on the part of the predestined:

    He was a useful man in his generation, a lover of learning, a faithful servant of Harvard College above forty years.
On the other hand, epitaphs with cherub stones tend to stress resurrection and later heavenly reward:

    Here cease thy tears, suppress thy fruitless mourn
    his soul -- the immortal part -- has upward flown
    On wings he soars his rapid way
    To yon bright regions of eternal day.

The final change seen in gravestone style is the radical shift to the urn and willow design. It is usually accompanied by a change in stone shape; while earlier stones have a round-shouldered outline, the later stones have square shoulders. "Here lies the body of" is replaced by "In memory of" or "Sacred to the memory of," quite different from all earlier forms. The earlier stones are markers, designating the location of the deceased or at least a portion of him. In contrast, "In memory of" is simply a memorial statement, and stones of this later type could logically be erected elsewhere and still make sense. In fact, many of the late urn and willow stones are cenotaphs, erected to commemorate those actually buried elsewhere, as far away as Africa, Batavia, and in one case -- in the Kingston, Massachusetts, cemetery -- "drowned at sea, lat. 39 degrees N., long. 70 degrees W." The cultural changes that accompany the shift to urn and willow designs are seen in the rise of less emotional, more intellectual religions, such as Unitarianism and Methodism. Epitaphs change with design and in the early nineteenth century tend more to sentiment combined with eulogy.

This sequence of change did not occur in a vacuum, unrelated to any cultural change elsewhere; indeed, the sequence of three major types also takes place in England, the cultural parent of the Massachusetts colony, but about a half century earlier. Thus cherubs have become modal by the beginning of the Georgian period (1715), and urns and willows make their appearance, as a part of the neoclassical tradition, in the 1760's. In fact, the entire urn and willow pattern is a part of the larger Greek Revival, which might explain the squared shoulders on the stones -- a severer classical outline.

Thus far we have been discussing formal change through time, and some of the fundamental causes. We have seen that New England is changing in harmony with England, with an expectable time interval separating the sequences. But we have not identified the relationship of all of this to archaeological method.

The battleship-shaped curve assumption is basic to many considerations of culture process in general and to such dating methods as seriation. Seriation is a method whereby archaeological sites are arranged in relative chronological order based on the popularity of the different types of artifacts found in them. The approach assumes that any cultural item, be it a style of pottery or a way of making an arrowhead, has a particular popularity period, and as it grows and wanes in popularity, its prevalence as time passes can be represented graphically by a single peaked curve. Small beginnings grow to a high frequency of occurrence, followed in turn by a gradual disappearance. If such an assumption is true, it follows that a series of sites can be arranged so that all artifact types within them form single peaked curves of popularity over time. Such an arrangement is chronological, and tells the archaeologist how his sites relate to one another in time.

By plotting style sequences in this manner in a number of cemeteries, we find that the assumption, not previously measured with such a degree of precision, is a sound one: styles do form single peaked popularity curves through time. By adding the control of the spatial to the form-time pattern explained above, we gain a number of understandings regarding diffusion -- the spread of ideas through time and space and how this, in turn, affects internal change in style. In looking now at the three dimensions we will see that all of the secondary cultural controls become even more important.

The style sequence of death's head, cherub, and urn and willow design is to be found in almost every cemetery in eastern Massachusetts. However, when we inspect the time at which each change takes place, and the degree of overlap between styles from cemetery to cemetery, it becomes apparent that this sequence was occurring at a widely varying rate from place to place. The earliest occurrence of cherubs is in the Boston-Cambridge area, where they begin to appear as early as the end of the seventeenth century. Occasional early cherubs might be found in more distant rural cemeteries, but in every case we find them to have been carved in the Boston area and to be rare imports from there. The farther we move away from the Boston center, the later locally manufactured cherubs make their appearance in numbers. The rate at which the cherub style spread outward has even been approximately measured, and shown to be about a mile per year. It is not common in archaeology to make such precise measurements of diffusion rate -- the usual measurements are cruder, such as hundreds of miles in millenniums.

We can view Boston and, more significantly, nearby Cambridge as the focus of emphasis of Puritan religion and inquire what factors might contribute to the initial appearance of cherubs and the change in religious values in this central area. We have noted that the change had already been accomplished in England by the early eighteenth century, so that when the first cherubs begin to appear in numbers in Cambridge, they were already the standard modal style in England. While cherubs occur in Boston, they never make a major impression, and as many death's heads as cherubs are replaced by the urn and willow influx.

On the other hand, in Cambridge cherubs make an early start and attain a respectable frequency by the late eighteenth century. Although they never attain a full 100 per cent level there, as they do in most rural areas, they do at least enjoy a simple majority. When the cherub stones in Cambridge are inspected more closely, we find that roughly 70 per cent of them mark the graves of high status individuals: college presidents, graduates of Harvard, governors and their families, high church officials, and in one case, even a "Gentleman from London." From what we know of innovation in culture, it is often the more cosmopolitan, urban stratum of society that brings in new ideas, to be followed later by the folk stratum. If this is true, then the differences between Boston and Cambridge indicate a more liberal element within the population of Cambridge, reflected in the greater frequency of cherub stones there. This is probably the case, with the influence of the Harvard intellectual community being reflected in the cemetery. It would appear that even in the early eighteenth century, the university was a place for innovation and liberal thinking. Cambridge intellectuals were more likely to be responsive to English styles, feelings, and tastes, and this could well be what we are seeing in the high number of cherub stones marking high status graves.

Introduced into Cambridge and Boston by a distinct social class, the cherub design slowly begins its diffusion into the surrounding countryside. Carvers in towns farther removed from Cambridge and Boston -- as far as fourteen miles west in Concord -- begin to change their gravestone styles away from the popular death's head as early as the 1730's, but fifty miles to the south, in Plymouth, styles do not change until the fifties and sixties and then in a somewhat different cultural context. We find, however, that the farther the cemetery is from Boston, and the later the cherubs begin to be locally manufactured, the more rapidly they reach a high level of popularity. The pattern is one of a long period of coexistence between cherubs and death's heads in the Boston center, and an increasingly more rapid eclipsing of death's heads by cherubs in direct proportion to distance, with a much shorter period of overlap. One explanation is that in towns removed from the diffusion center, enforcement of Puritan ethics and values would lessen, and resistance to change would not be so strong. Furthermore, revivalism and the modification of orthodox Puritanism was widespread from the late thirties through the sixties in rural New England, although this movement never penetrated Boston. Such activity certainly must have conditioned the rural populace for a change to new designs.

We have, then, a picture of the introduction of a change in the highly specific aspect of mortuary art, an aspect reflecting much of the culture producing it. We see the subsequent spread of this idea, through space and time, as a function of social class and religious values. Now we are in a position to examine internal change in form through time, while maintaining relatively tight control on the spatial dimension.

One significant result of the use of gravestone data with its accompanying controls is the insight it provides in matters of stylistic evolution. The product of a single carver can be studied over a long period of time, and the change in his patterns considered as they reflect both ongoing culture change and his particular manner of handling design elements. The spatial axis extending outward from Boston shows not only systematic change in major style replacement rates but also a striking pattern of difference in style change. We find that in many cases, the farther removed we become from Boston, the more rapid and radical is change within a given single design. This has been observed in at least five separate cases, involving a number of the styles of more local distribution; we can inspect one of these cases closely, and attempt to determine some of the processes and causes of stylistic evolution.


Click image to view some gravestone locations and movements of carvers

Click to View Figure 2

The design in question is found in Plymouth County, centering on the town of Plympton. Its development spans a period of some seventy years, and the changes effected from beginning to end are truly profound. Death's heads occur in rural Plymouth County, as they do elsewhere in the late seventeenth century. However, in the opening decade of the eighteenth century the carver(s) in Plympton made certain basic changes in the general death's head motif. The first step in this modification involved the reduction of the lower portion of the face, and the addition of a heart-shaped element between nose and teeth. The resulting pattern was one with a heart like mouth, with the teeth shrunken to a simple band along the bottom. The teeth soon disappear entirely, leaving the heart as the sole mouth element. This change is rapidly followed by a curious change in the feathering of the wings.


Click image to view evolution of styles in Plympton cemetery and photographs of sample gravestones

Click to View Figure 3

While early examples show all feather ends as regular scallops crossing the lines separating individual feathers, shortly after the first changes in the face were made, every other row of feather ends had their direction of curvature reversed. The resulting design produces the effect of undulating lines radiating from the head, almost suggesting hair, at right angles to curved lines that still mark the feather separation. These two changes, in face and wing form, occupy a period of 35 years from 1710 through 1745. During the later forties this development, which has so far been a single sequence, splits into two branches, each the result of further modification of wings. In the first case, the arcs marking feather separations are omitted, leaving only the undulating radial lines. Rapid change then takes place, and soon we are confronted with a face surmounted by wavy and, later, quite curly hair. The heart mouth has been omitted. We have dubbed this style "Medusa." In the second case, the separating lines are retained, and the undulating lines removed; the result in this case is a face with multiple halos. At times, space between these halos is filled with spiral elements, giving the appearance of hair, or the halos are omitted entirely. The heart-shaped mouth is retained in this case and modified into a T-shaped element.

Both of these styles enjoy great popularity in the fifties and sixties, and have slightly different spatial distributions, suggesting that they might be the work of two carvers, both modifying the earlier heart-mouthed design in different ways. Yet a third related design also appears in the forties, this time with tightly curled hair, conventional wings, and a face similar to the other two. Although this third design seems to be a more direct derivative of the earlier death's head motif, it is clearly inspired in part by the Medusa and multiple halo designs. This tight-haired style has a markedly different spatial distribution, occurring to the west of the other two, but overlapping them in a part of its range. Of the three, only the Medusa lasts into the seventies, and in doing so presents us with something of an enigma. The final form, clearly evolved from the earlier types, is quite simple. It has a specific association with small children, and has never been found marking the grave of an adult, and rarely of a child over age five.

The carver of the fully developed Medusa was probably Ebenezer Soule of Plympton; a definitive sample of his style is found in the Plympton cemetery. Normal Medusas, except for the late, simple ones marking children's graves, disappear abruptly in the late sixties. In 1769, and lasting until the eighties, stones identical to Soule's Medusas, including the simple late ones, appear in granite around Hinsdale, New Hampshire. Fortunately, a local history has identified the carver of some of these stones as "Ebenezer Soule, late of Plympton." This alone is of great interest, but if Soule did move to Hinsdale in 1769, who carved the later children's stones in Plymouth County? As yet, no answer is known.

This development raises two interesting considerations. First, we see that a style, the Medusa, which had been used for the general populace, ends its existence restricted to small children. This pattern has been observed elsewhere, with children’s burials being marked by designs that were somewhat more popular earlier in time. In other words, children are a stylistically conservative element in the population of a cemetery. While no clear answer can be given to this problem, it may well be that small children, not having developed a strong, personal impact on the society, would not be thought of in quite the same way as adults, and would have their graves marked with more conservative, less explicitly descriptive stones.

The second problem raised by the Medusas is their reappearance in Hinsdale. If, as archaeologists, we were confronted with the degree of style similarity seen between Hinsdale and Plympton in mortuary art, might we not infer a much greater influence than a single individual arriving in the community? After all, mortuary art would be about the only distinctively variable element in material culture over eighteenth-century New England, and such a close parallel could well be said to represent a migration from Plympton to Hinsdale. One man moved.

Placing this striking case of stylistic evolution in the broader context of culture change and style change in eastern Massachusetts, we find that it is paralleled by other internal modifications of death's head designs in other remote rural areas. The closer we move toward Boston, the less change takes place within the death's head design, and in Boston proper, death's heads from 1810 are not that different from those from 1710. Yet 1710 death's heads in Plympton and elsewhere had changed so radically by 1750 that it is doubtful that we could supply the derivation of one from the other in the absence of such an excellently dated set of intermediate forms. This difference in rate of change can be explained by referring back to the long, parallel courses of development of both death's head and cherub in the diffusion area's Boston center. However, culture change in the area of religion, marked by a shift of emphasis from mortality to immortality, probably generated a desire for less realistic and less grim designs on stones. Given this basic change in religious attitudes, what were the alternatives facing carvers in Boston as opposed to the Ebenezer Soules of rural New England? In Boston it was simply a matter of carving more cherub stones and fewer death's head stones; neither had to be altered to suit the new tastes.. The choice between cherub and death's head in Boston has been seen as ultimately a social one, and if there was a folk culture component within Boston, there was nothing but folk culture in the more democratic, less-stratified rural areas. With no one to introduce cherubs and to call for them with regularity in the country, carvers set to work modifying the only things that they had -- the death's head. The more remote the community, the later the local cherubs appear, diffusing from Boston, and the more likely the tendency to rework the common folk symbol of skull and wings. Thus we get Medusa and haloed T-mouthed faces populating the cemeteries of Plymouth County until cherubs finally appear. Even then, the waning popularity of the death's head in this area might be more the result of Soule's exit than their unsatisfactory appearance compared to the new cherubs.

Only a few applications of gravestone design analysis have been detailed here. A three-year program is presently under way through which we hope to pursue numerous other aspects of this fascinating study. There is a large and important demographic dimension to these data since precise date of death is given, as well as age at death, patterns of mortality and life expectance through time and space can be detailed. The results of this work, in turn, will add a biological dimension of style to the cultural one described above. Studies of diffusion rate, and its relationship to dating by seriation will be continued. Relationships between political units -- counties, townships, and colonies -- and style spheres will be investigated to determine how such units affect the distribution of a carver's products. Finally, a happy by-product will be the preservation on film of over 25,000 gravestones, a vital consideration in view of the slow but steady deterioration these informative artifacts are undergoing.

Aside from the value of this work to archaeology and anthropology in general, one final comment must be made. Compared to the usual fieldwork experienced by the archaeologist, with all of its dust and heavy shoveling under a hot sun, this type of archaeology certainly is most attractive. All of the artifacts are on top of the ground, the sites are close to civilization, and almost all cemeteries have lovely, shady trees.


dividing bar


View Photographs of Sample Gravestones

Mortuary Studies PageProject Home Page

Last Modified: December 14, 2007



© 2000-2007 Copyright and All Rights Reserved by
Patricia Scott Deetz and Christopher Fennell