Part I
Collecting in Its Institutional Context
CHAPTER 1
Collections as Artefacts
The Making and Thinking of Anthropological Museum Collections
Leonn Satterthwait1
Collections are artefacts—artefacts created by the bringing together, in some fashion, of sets of other artefacts.2 As artefacts, collections have a kind of coherence, a kind of integrity, as singular entities even though they are made of physically separate things (much as a dry-stone wall can exist as an ‘artefact’ even though its individual constituents are not physically bound together and can exist as separate entities). The associations that link the individual elements of a collection give the collection a presence in the world, an actuality, which extends beyond the existence of the individual elements that constitute it.3 Viewed from this perspective, the individual items contained in a collection are the collection’s physical constituents, constituents linked together by associations that extend beyond the mere physical existence of the collection’s separate components and that make the collection, quite literally, something that is more than the sum of its parts.
Of course, the individual items constituting a collection have an inherent interest and worth of their own and, in fact, frequently bear great and profound significance. But giving attention to collections as complete entities—as artefacts in their own right—shifts consideration of collections to another level of thought and provides insights that are not available from looking at collection constituents singly, as informative and useful as the latter might often be.
The idea of collections as artefacts and its implications are explored here. As artefacts, collections have formal properties from which much can be learned once they are identified and defined. Furthermore, there are many kinds of collections and these can come into being—that is, take on an existence as artefacts—in a number of ways. The processes involved can enhance understanding not only of collections and collection formation processes4, but also of material culture processes more generally. Also, collection formation is, ultimately, a sociocultural process, one entailing, among other things, both the cognitive and physical manifestation of an important cultural phenomenon: the making of categories. Finally, the perspective considered here raises questions about the very notion of a ‘collection’. The concept is, in fact, quite problematic.
As a focal point for these issues, and to give these reflections on collections and their making and meaning a tangible point of reference, attention will be given to the character of a collection of Australian Aboriginal items collected by Walter E Roth in Normanton in the southeast of the Gulf of Carpentaria, northern Queensland, in the 1890s. However, the more general points made, and more importantly, the perspective developed, are applicable to any collection, including collections of natural history specimens.
The Nature of Collections
The notion of collections as artefacts applies no matter how collections were assembled and how systematic or not the collectors were in their collecting activities. It also applies at several levels to the items held in museums—from distinctively identified sets of objects (the Donald Thomson Collection, the Winterbotham Collection) to assemblages of items of particular types (baskets, boomerangs) or from particular localities (Milingimbi, Mornington Island), to the entire holdings of a museum. The last, of course, represent artefacts constructed of collections of collections.
As constructions, collections have formal and material attributes of various kinds, as any artefact does. Of particular importance, however, is the fact that, deliberately and systematically assembled or not, collections are structured. This structure can be characterised in several ways: the types of objects represented in the collection; the materials present; the manufacturing techniques embodied in the objects; the functions served by the objects; the contextual associations the objects have; the places from which the objects were obtained; and so on.
What is important, however, is not just the kinds of things represented in a collection (no matter what attributes are used to establish this), but also their proportions. The latter is especially crucial; two collections could conceivably contain exactly the same kinds of objects, but differ considerably in character because they contain these objects in greatly differing proportions—a difference laden with implications. At any rate, when they are taken in conjunction with one another, types of collection constituents and their frequencies are probably the most salient and fundamental expressions of a collection’s distinctive character and of the way in which the collection is structured.5
A further significant consideration is that there are both positive and negative dimensions to the structure of a collection. The structure of a collection is defined as much by what it could contain, but does not, as it is by what it does contain. It is the configuration of the boundary between what a collection contains and what it could but does not, and the tension between the two, that gives to each collection its unique character as a collection and imparts dynamism to collections, to the differences between them and to the processes by which they came into being.
As artefacts, collections are amenable to consideration from the same perspectives and analytical approaches applicable to any item of material culture.6 However, it is the difference between what a collection contains relative to what it does not that is one of the most revealing attributes of a collection as a structured entity, as an artefact. But how can this difference be pinned down? How can the shape of the boundary between the actual and the possible be delineated in the case of any particular collection?
The problem is that, with rare exceptions, collections represent samples derived in largely unknown ways from unknown universes.7 If insights are to be derived from how collections are constructed and how they are structured, it is necessary to compare what a collection contains to what it lacks. This requires identifying the omitted— somehow giving a presence to the non-present—and this is not an easy feat. This is one case in which the absence of evidence—that is, the absence of elements, in type and frequency, of the original artefactual universe from which a collection was extracted—is in fact evidence.
There are several approaches one could take to this problem. Here a distinction can be made between information that might be obtained about a collection and what is absent from it, and information that might be obtained from a collection. The first corresponds to information obtainable from sources external to the collection itself, such as published accounts, archival documents, photographs, analogies with the material culture of other regions or communities, comparisons with other collections, the collector’s notes and diaries, and, most important of all, the people who created and used the objects in the first place.
The second kind of information, on the other hand, corresponds to that which is derivable, or at least inferable, from the items in the collection themselves. All material culture objects are woven into an extensive web of associations and relationships8, including those with other objects, and if these associations with other objects can be determined, it becomes possible to infer what is not present in a collection from what is. Such associations commonly relate to the relationships occurring between objects during their manufacture or in the course of their use or while together in storage.
These kinds of associations among objects can often be inferred by reasonable supposition or from detailed physical examination of the attributes of the objects present. For example, the presence of a paddle in a collection implies the use of canoes, even though the latter might not be represented in the collection. Charring and attrition of wood on the side of a central Australian spearthrower implies its use in generating fire by friction, and hence the presence of other fire-making accoutrements, such as split sections of timber, wedges to hold the splits open and dung for use as tinder. Again, although the spearthrower might have been collected and the other things not, the features of the spearthrower itself suggest their existence.
Some objects are also made for use with other objects; in fact, they often can only be used in this way. Such objects are coupled in use, functionally linked9, and the presence of one such object implies the existence of the other.10 Most Australian spears, for instance, were and are made to be thrown with a spearthrower. The presence of one, either a spearthrower or a spear (particularly a spear with a hole or depression in the proximal end of its shaft for articulating spear and spearthrower when projecting the spear), implies the existence, even if it is not represented, of the other.
Some possible approaches to determining what might be absent from a collection through physical examination of the items present entail the application of analytical methods more commonly associated with archaeological investigations. Included here would be detailed studies of use-wear patterns and detection and identification of the residues that are invariably transferred from one object to another when the two come into contact.11 Shell disks strung to make a necklace, for instance, imply the use of a grinding implement to shape the disks and a drill for perforating them, and use-wear and residue analyses might provide information about the manufacturing equipment used, even if they were not collected with the necklace. The application of such techniques to non-lithic objects in ethnographic collections is only in its infancy.12
Of course, all of these approaches can be and, in some cases, must be, used in combination, and that is the way in which the structure of a collection is best delineated and information drawn then from it. The following is provided as an example of such an attempt, as well as a stimulus for consideration of broader issues relating to collections and how they can be conceptualised.
Walter E Roth’s Normanton Collection
Mention was made above of the fact that all material culture objects, museum collections included, are woven into an extensive web of multidimensional associations and interrelationships. In fact, it is these associations and relationships that give items of material culture their anthropological relevance. The objects themselves, the objects severed from their human connections, would have little allure for anthropologists and archaeologists. In this regard the study of material culture, including museum collections, is ultimately about relationships, not objects per se.
This is reflected in the phrase ‘material culture’, which connotes a relationship between, on the one hand, the material, the objects that people make, use or to which they attach meaning (without necessarily modifying them in any way), and on the other hand, the culturally constituted ideas and behaviours with which they interrelate.
The totality of the web of interrelationships—sociocultural, environmental, spatial and temporal—in which material culture objects are embedded can be regarded as constituting their context (‘context’, after all, derives from the Latin ‘contexto’: ‘to weave together’). The question is, as artefacts, to what extent do collections reflect in their content and structure the contexts from which they were originally derived?
A case in point is the set of 103 items of Aboriginal material culture obtained by the well-known early Australian collector and ethnographer WE Roth during a series of visits to Normanton from 1895 to 1898, prior to and at the beginning of his appointment as Protector of Aborigines for the Northern District of Queensland (see Chapter 6). These objects range from fire-making implements, through string bags, necklaces and message sticks, to spe...