One of the greatest pleasures for many collectors of African art is being able to trace the history of a piece that they own. Fortunately, the provenance of a large number of masks of the Dan-Guere family in private collections can be learned. Many were originally gathered by one medical missionary, Dr. George Harley, and for quite a few of these masks, a date may be obtained that indicates when they were brought to the United States or Europe. Some can even be traced back to the village of origin, and the names and functions that were supplied to Dr. Harley can be determined. This article provides a guide for documenting some of the many Harley masks in private collections and honours his important contribution to private and museum collections and to our knowledge of African art.
On 30 January 1926, after arriving in Monrovia, Liberia, Dr. George Harley and his bride Winifred travelled for two weeks to their post in Ganta, in the Mano country in the interior, adjacent to the Dan region. The Harleys did not match the stereotype of missionaries—rather than pursuing single-mindedly the conversion of Liberians to Christianity, they developed many other interests. They were dedicated to improving the physical health of the inhabitants. Studying the village life and local customs, the couple became deeply involved in serious studies of native medical practices, local plants, and the secret societies of the region. They mapped parts of the country and collected plants, reptiles, chimpanzee skulls, and other material for various scientists in the United States.
Rather than burning the masks that were so common in the area, the Harleys collected them, apparently gathering over 1,000 full-sized masks from the Dan-Guere region. A few hundred went to museums, and the majority are now in private collections. The number of miniature ma masks collected and placed in private hands was even greater.
At the time of the Harleys' arrival, the Poro society, in whose ceremonies most of the masks were used, was faced with Liberian government efforts to outlaw it. Although efforts were later undertaken to revive it, today it is simply deteriorating as Western ideas penetrate the bush. Few high-quality, old masks now come out of the region in which Dr. Harley collected.
The Harleys witnessed much of this change, for they did not leave Liberia permanently until 1960 when Dr. Harley retired. Even during this 34-year period, their absences from Liberia were infrequent and brief. Their three children were born in Ganta—one died there, at the age of four, from quinine tablets he swallowed.
Although Dr. Harley was not a professional anthropologist or art historian, he had received considerable training in these areas. His interest in the secret societies and their masks owed much to the visit of Dr. and Mrs. George Schwab to Liberia in 1926. The Schwabs, Presbyterian missionaries in the Cameroons, who had long collected data on the local ethnic groups as an avocation, had been in contact with Ernest Hooten, the head of Harvard's anthropology department. They introduced Dr. Harley to Hooten, and thus began a long association with Harvard.
In 1930-31, on his first furlough, Dr. Harley enrolled as a graduate student at Harvard and took courses in anthropology as well as in tropical medicine. Back at his missionary post, he held his first Harvard assignment, as a Field Associate, to collect data on comparative physical characteristics of the various ethnic groups. On the second furlough, in 1937-38, Dr. Harley wrote Native African Medicine (1941), his PhD thesis for the Hartford Seminary Foundation, and Notes on the Poro in Liberia (1941). Later, a second monograph, Masks as Agents of Social Control in Northeast Liberia (1950), was produced—the last two books are devoted to the role of the masks he collected. In fact, his research is responsible for much of what we now know about the use of masks in the Dan regions. In 1944, Dr. Harley organised the material the Schwabs had collected in 1926 into Tribes of the Liberian Hinterland (Schwab 1947). Although Dr. Harley is listed only as 'Editor', a fact that bitterly disappointed him, he apparently did a great deal of work on the book.
The collection of over 1,000 masks would be important enough to those interested in the art of the Dan regions, but most of those collected in the early years, until about 1946, also came with a great deal of documentation—Dr. Harley acquired the pieces from people closely connected with them or familiar with their use. Unfortunately, in many cases, the documentation and even the date of sale have since been separated from the masks. After about 1950, the commercial trade became the dominant source of masks, and most came with no data as to origin or use—it is still possible, however, for many owners of these masks to determine the date when they were sent to Europe or America.
Dr. Harley’s data are particularly important because they were collected at a time when many informants’ memories reached into the period before the Poro society (and other secret societies) had begun to undergo profound change. Data were collected in several ways, but much of Dr. Harley’s general knowledge of the Poro was obtained from two informants who worked with him regularly. One died mysteriously—Mrs. Harley suspects the death was from poisoning for revealing secrets of the Poro. That opinion is shared widely in the region. The second informant initiated Dr. Harley’s mask collecting in 1932 or 1933 by bringing him some secret articles he had inherited and wanted to be rid of. The information on individual masks was acquired at the time they were purchased. Whoever brought a mask, which was often in a sack by night, was taken to a backroom in the Harley home and questioned on its use. No doubt sellers sometimes fabricated a story to satisfy the curious doctor, but from his years of living with the Mano, his description of the Poro and of the role masks played in the Mano and Dan regions must certainly be reasonably accurate.
Many uses of the masks that the Harleys documented can probably no longer be confirmed, as the roles have simply disappeared. However, later studies have verified some of the earlier Harley information. One type of mask appears over and over with a similar description of function and appearance—Thompson tells of the kao gle mask he saw in 1967 much as Dr. Harley earlier described the klua ge, the chimpanzee mask that teaches social order by acting out social disorder. Himmelheber identifies this type mask as kagle and also provides a very similar report of its function. Another, practically identical account of the gue ke zu appears with a photograph of a similar mask in the envelope of a record made in the Dan region in 1965 and 1967 (Office de Radiodiffusion-Télévision Française).
Observers agree on names and roles for other masks. Both Harley and Thompson identify the highest power in the mask hierarchy as residing in the go gle. For both researchers, go gle was the “leopard spirit.” However, Thompson pictures a mask identified as go gle that is very different from those Harley illustrates. There are numerous other accounts of masks with similar names, features and functions.
It is also probable that the descriptions of some masks' usage are only partial. In fact, Harley in his notes to a group of masks he sold in 1946 treats the descriptions of a particular type of mask as cumulative. For a single go gle mask, he refers in his personal correspondence to the descriptions of others with the same name and features. Apparently, informers usually described only some of the uses of the masks, and the whole picture emerges when the accounts from several informants are combined for masks of the same type and name. The difficulty is that a single mask can have several names, as others have pointed out.
Many other of Harley's findings have been confirmed. Himmelheber, Becker-Donner (in Kunst und Handwerk in Nordostlich Liberia, 1940) and Harley all point out that some masks were not worn; some of the masks did not even have holes in the periphery. Himmelheber and Becker-Donner also agree with Harley that some of the masks were portraits. In fact, Harley is often able to name the person depicted. The mask in the middle and right images below is identified as a portrait in Harley's own notes entered in his copy of Masks as Agents of Social Control in Northwest Liberia. Perhaps in the Mano region, portraiture is more common than in the Dan, if Harley's contention is correct that masks played a role in ancestor cults there before the arrival of the Poro.
No doubt, the contention of many authors is true that the form of a particular mask does not necessarily determine its function. However, the consistent relation between physical characteristics and functions for a number of masks suggests that function and form are, at least in some types of masks, closely linked.
Before World War II, most of the masks collected by the Harleys were apparently sold to the Peabody Museum at Harvard, although the museum’s correspondence with Harley indicates that even in those days a very few were sold to private collectors. The Peabody acquired about 300 from 1932 to 1940, and another 45 in 1948 (a few of these were traded to the Denver Art Museum). During this period and later, Dr. Harley gave some masks to acquaintances. Many of these may never return to collectors or museums; some have been destroyed. In my efforts to trace one such mask, I was told by the recipient that “knowing the sad story that went with the mask, I kept it for a period of time out of courtesy to the Harleys, then destroyed it.”
By 1946, a significant number of masks were going into private hands—to dealers or directly to collectors. Prices had climbed above the $5 or so typically asked in the 1930s (one was sold for $50) to an average $10 to $20 in 1946, to around $25 to $50 for reasonably good masks in the 1950s, and up to $150 to $200 for a small number that Dr. Harley particularly admired. A private market for the masks was developing in the United States. In fact, correspondence from this period shows that few museums were willing to pay the going market price. By the 1950s, dealers and collectors were frequently contacting Dr. Harley for masks. Dozens were sold through dealers in Boston, Connecticut, and New York, and a few masks even made their way to California. Some now belong to the Peabody Museum of Yale University, which acquired them from the widow of Ralph Linton, an early customer of Dr. Harley.
A number of the pieces sold after the war had been “restored” by Dr. Harley. New jaws were carved when the old ones were missing, and small chips and breaks were repaired. A few masks were completely refinished. One dealer, according to Harley’s correspondence, was reworking the restorations after he received a mask. He used lacquer thinner and earth to dull the new finish. Dr. Harley’s offer to one New York dealer of more restoration work and a label, “repaired by G. W. Harley,” was, not surprisingly, rejected. In spite of the cool reception from dealers and collectors, restoration was apparently undertaken all through this period. A major motivation, it seems, was the challenge Dr. Harley found in trying to reproduce the native dyes.
For a long time, he kept some of the masks he particularly liked, eventually selling these early pieces in the later years. The dates in the series that follow later in this article can only give a latest possible date for collection. In fact, some of the pieces sold in the 1960s were collected in the 1930s or 40s, as is evidenced by some items acquired by the College of William and Mary in 1965, but which had been published in 1947. The correspondence indicates that by 1962, when he had retired to Virginia, Dr. Harley was ready to sell everything he had kept. He was eager again to place some of the remaining masks in museums and to provide some sort of fitting memorial to his work. Relations with Harvard’s Peabody Museum had cooled. In retrospect, the early prices paid by the Peabody seemed very low, and Dr. Harley thought his pieces and name had become rather lost in the large museum. Duke University, his alma mater, agreed to provide a special room that would be named after him and thus was able to hold off other contenders. Dr. Harley agreed to save a significant number of pieces for them, once they began the long process of acquiring the needed money. The acquisition was completed in 1965. Meanwhile, the Smithsonian Institution, the College of William and Mary, and Lincoln University (Pennsylvania) obtained small groups of masks. Ironically, to most collectors and scholars, it is the Harley collection at Harvard that is well known and often studied.
By 1966, when Dr. Harley died, virtually all the old masks were gone. Most of the few remaining items were donated by Mrs. Harley to Lincoln University.
Most of the masks that Dr. Harley collected before the war are from the Liberian Dan (Gio), the Mano, and the Geh ethnic groups. That is, they come from the region close to Ganta, where the mission station was located. After the war, the collections began to include a significant number of masks that would now be identified as Guere—some are described by Dr. Harley as Kran, others as Tchien. He gathered a smaller number of Konor masks and Dan masks in the western part of the Ivory Coast, and occasionally acquired a few Bassa, Kpelle (Guerze), Toma (Buzi), Gola, and Vai pieces.
The large number of masks collected from a relatively small region, often with the place of origin specified, provides some evidence of the style regions within the Dan-Guere area. The masks confirm the broad thrust of the classification system published by Vandenhoute in 1948, if one allows for some masks that might have been traded away from their point of origin (and maybe for some inaccurate data accompanying the masks). Not surprisingly, the Harley collections include few masks of what Vandenhoute labels as Northern (more accurate would be North-eastern) Dan style. In all the Harley mask collections and photos I have examined, there is not one mask of the wide-eyed “runner” type so frequently seen in French collections that draw heavily on the Northern Dan of the Ivory Coast.
Some of the details of the Vandenhoute classification system are inconsistent with the Harley data. For example, not all the Southern (actually Southwestern) Dan masks have median strips on the forehead, as Vandenhoute suggests. In fact, there are numerous well-documented exceptions in the Harley collection. Nor are the deep, shiny black patinated masks exclusively Northern, although they are certainly less common in the Southern group; the Mano ‘portrait’ mask shown above illustrates a Southern mask with such a patina. Similarly, the “beak and horn” styles apparently extend outside the regions indicated by Vandenhoute. The Harley notes indicate that horned masks came from Dan villages far into the Southern-style region, but their small numbers could possibly be explained by village trade. The beaked masks provide a more notable exception to Vandenhoute’s rules—their places of origin seem to spread through-out the Mano and Gio region, although the masks of the Southern regions are rather different from those of the Northern Dan.
The Harley masks from the Southern Dan regions stand in sharp contrast to the Dan masks that appear typically in French collections. Although exceptions will probably exist to the following guidelines, which are based on my examination of a large number of the Harley masks and a review of other documented masks, the rules may still assist collectors in locating the region from which a mask of the Dan family originated.
The Southern-style usually has a fairly sharp division between the planes of the forehead and those of the lower part of the mask. The transition between the upper and lower parts often consists of a rather sharp line through slit eyes or directly above round eyes. The planes of the forehead may form a rather noticeable angle with the planes of the lower part. Alternatively, the forehead may consist of a jutting, rounded form, rather than flat planes.
In contrast, the Northern style is usually characterised by rather smooth, rounded surfaces. Slit eyes are more common in the Southern style, while round eyes appear to be more frequent in the Northern style. When round eyes appear in Southern masks, they are likely to be tubular and projecting. Although Northern masks may have ringed, round eyes, the rings do not usually project far from the mask, and in many cases, the eyes are also larger. The wide, round-eyed masks of the “fire runner” mask are completely absent in the Southern style. The round eyes of Northern Dan masks are often set in a roughly circular depression. This method of making the transition from the forehead to the lower part of the face is absent in Southern Dan masks, which are more likely to use intersecting planes, as previously described. The Southern masks may also have a well-defined horizontal intersection at the cheekbone level.
Median strips on the forehead appear more frequently on masks of the Southern style. I have never seen double median strips on a mask of Northern provenance, although they appear on quite a few Southern ones. Nor have I ever seen a Southern mask with the median line indicated by the meeting of two surfaces of the forehead. Relief scarifications around the sides masks appear more characteristic of Northern masks—in the few instances of scarifications on Southern masks, the markings are usually incised.
The retrousse nose, with wide, flared nostrils, and plaited fibres, forming a "U" on the forehead and dropping to the sides of the mask, appear to be unique to masks of the Northern Dan, as are the fibre headdresses that cover the top of the wearer's head. On the other hand, a simple rolled coiffure curved over the top of the mask, often decorated with cowries or beads, appears to be more frequent on Southern masks. These coiffures may be made of fibre or of "country-cloth" rolled on a fibre core.
Interestingly, none of the Harley masks I have seen are of one type that belongs to the style Vandenhoute labels as nuclear Guere-Ouobe, a very expressive, aggressive mask type. As those pictured by Vandenhoute are attributed to villages in the eastern Guere area, this kind of mask apparently did not usually reach Ganta.
Much needs to be done with the chronology of Dan masks, little is known about how they have changed over time. However, some interesting observations can be made from a quick examination of the Harley masks.
First, almost all the masks acquired before 1937 were blackened—a few were so completely covered with encrustation that the colour was no longer easily visible. By the 1950s, a significant number of the masks collected were no longer predominantly black. Many of these had apparently been blackened in the past—and a number no doubt pre-date some of those collected before 1937—but the frequent recoating that was so obvious in the masks collected earlier had not been continued. Increasingly, there were masks that apparently had never been blackened, but quite a few of these were from tribes hardly represented in the earlier groups (particularly the Guere). In addition, the colour of the masks' interior tended to change. Few of the later masks had the signs of kola nut sacrifice on the inside, so common on the masks collected earlier.
Second, there was some shift in the use of metals. Although aluminium appeared on the early masks—in fact, aluminium beads were already being made locally in the 1930s—the metal appeared much more frequently on those collected later. In the masks collected in the early years, teeth were more commonly of bone, human or animal teeth, broken pottery, or iron, later, aluminium was commonly used. In many cases, oversized holes in those masks suggest that aluminium teeth had been substituted for the original teeth of other materials. On the other hand, the use of brass or copper for outlines of eyes or for median strips on the forehead, although never common, appears not to have changed much in its frequency of use over the period.
Third, certain types of masks decreased in frequency. Go gle masks became increasingly rare. The group of Harley masks sold in 1946 included two such masks, which were probably collected before the war—in the pictures of close to 200 masks sold in the 1950s, there is but one glo gle mask. Beaked masks also became less frequent. "Hornbilled masks," with separate nose and beak, were rare throughout the period and much sought after by collectors according to Dr. Harley's correspondence. Later groups had several hornbills labelled as "recent" or "copies," apparently carved on order to supply this demand but rarely an old one.
Fourth, not surprisingly, there was a decline in the quality of the average mask over the period. Although the Harley collection cannot be taken as a completely random sample of masks available at the time, apparently they are reasonably representative. According to Mrs. Harley, her husband rarely destroyed a mask (unless it was exceptionally poor) and, until the commercial trade dominated, he rarely turned down one offered to him. The masks sent in 1937 to Harvard’s Peabody Museum (excluding those marked as having been carved on Harley’s order) apparently represented what he had gathered to that date, with the exception of a few pieces he decided to retain for himself. But even a number of these are illustrated in his publications. The 1952 series apparently represented that period reasonably well. Although there are some very fine examples in this series, the average quality is already below that of the 1937 group, even if one eliminates those masks in the 1952 series that Harley labelled as "recent." In the later groups, the deterioration in average quality is even more marked, although a fine mask certainly appears from time to time.
Most of the masks coming out of Liberia in recent years derive from peoples different from those whose masks Harley collected. Authentic old Mano and Liberian Gio masks now rarely appear in Monrovia. However, examples of Bassa and Grebo masks such as those which have come on the market in some quantity since 1968 are hardly represented in the Harley group. Similarly, Harley collected few of the masks of the Gola, Vai, and Toma, although old examples have appeared on the market very recently. The distance of these groups from Ganta may account for the shortage of such masks in the Harley collections. In addition, the exhausted supply of the masks of the better-known Dan-Guéré groups has probably led traders in recent years to search more vigorously for masks of the other tribes.
As the Harley masks in private collections have changed hands, much of the documentation has been lost. For example, the Mano 'portrait' mask shown earlier was purchased in a European gallery—neither the gallery nor the consignor knew it had been collected by Dr. Harley. For many masks, however, there is a possibility of re-establishing the link with Harley and, in some cases, of learning something of collection dates, place of collection, and data on name or use. Two keys exist—photographs and Dr. Harley’s markings inside the masks.
The two monographs written by Harley and the one by Schwab, all publications of Harvard’s Peabody Museum, illustrate more than 100 Harley masks. Contrary to common belief, Harley did not turn all the published masks over to the museum. Some of the approximately dozen and a half masks that were still in Harley’s hands at the time the three books were written have made their way to other museums—a number are still in private hands. In addition to the photographs in these three publications, I have photographs of some 200 masks that Dr. Harley offered for sale between 1950 and 1952.
Dr. Harley marked many, though apparently not all, masks with numbers. Unfortunately, in some cases, the numbers have been removed by dealers and collectors. Harley’s numbering system was not systematic over the years and is not yet completely understood, but the following series can be rather clearly identified:
The identification of masks in this article illustrate the methods that can be used to match the masks to the information available. The Mano hornbill mask shown earlier was acquired from the private collection of a dealer who remembered only that he had purchased it from Dr. Harley. It had appeared in Schwab’s Tribes of the Liberian Hinterland; it was identified as the “greater hornbill,’” and the text carried a description of its use. The number 68 pencilled inside the back of the mask matched the number indicated on the negative used for the publication (the negative was on file in Harvard’s Peabody Museum). The museum records covered number 67 (the “lesser hornbill”’), which was apparently collected with number 68, from Ziketa, a southern Mano village, near the Bassa area. Number 67 had entered the Peabody’s collection in 1937.
The above Mano mask was purchased in Paris, where it had been sent on consignment from England. No information at all came with the mask. However, it appeared to be one that was illustrated in Harley’s Masks as Agents of Social Control in Northeast Liberia (plate XIVd), and was thus one of his masks. A check of the original negative in Harvard’s Peabody Museum confirmed that it was indeed the same mask and that it had been photographed in Cambridge in 1944. There was no number visible. The only other information came from Harley’s own copy of the publication, where he had pencilled the word ‘‘portrait’’ next to the photograph.
The identification of the Dan (Gio) Go Gle ‘leopard mask’ illustrates another method of tracing a Harley mask. The mask was purchased from a New York dealer, who provided a list of its previous owners. The first collector was erroneously credited with obtaining the mask in the field. The white pencilled number 14 inside the mask was in Dr. Harley’s handwriting, and the mask matched the description of number 14 in Harley’s list of masks in the 1946 Series. That description indicated its place of origin and its use and gave the 1946 price of $10. Masks as Agents of Social Control provided a further description of the same mask. Thus, the mask was now identified as go gle mask from Tawie, Gio territory, collected before 1946 by Dr. Harley. His description of its use matched well the description other observers had provided for this type of mask.
The masks collected by George Harley provide a fertile field for more study, for they constitute much of the corpus of Dan-Guéré masks owned by collectors and dealers. The interested owner can, in many cases, identify Dr. Harley as the collector and establish some links with the collection date. He may learn more: the name of the mask, its use, and details of its place of origin. I will be glad to assist anyone who has a Dan-Guéré mask he suspects to be of Harley origin. I will compare a (non-returnable) photograph against the photographs I have of pieces offered for sale in the 1950s and will try to match numbers against the lists I have of pieces offered for sale in the various series. In this way, we can enlarge the confirmed corpus of the Harley collection, which is so valuable for its known dating and ethnographic information.
This article originally featured in African Arts 10, no. 2 (January 1977), a quarterly journal devoted to the plastic and graphic arts of Africa.
Lou Wells’ interest in African art has its origins in his driving a VW Beetle from Boston to Costa Rica (and back) in 1966. His encounter there with Pre-Columbian art led him to NYC galleries, at a time when African and Pre-Columbian were often shown together. An assignment in Liberia in 1967 did it—he realised he loved African art more than Pre-Columbian. Friendship with Rene Guyot in Monrovia provided training, and access to Harvard’s Peabody Museum publications and collections led him to track material collected by Dr. George Harley, a medical missionary in Liberia from 1926 to 1960. Gradually his interest spread to the rest of West and Central Africa. He taught at Harvard Business School from 1966 until 2012, with a research focus on foreign investment and development.