Review of ‘The Rock Paintings of Arnhem Land, Australia: Social, Ecological and Material Culture Change in the Post-Glacial Period’ by Darrel Lewis

26th May 2014

‘The Rock Paintings of Arnhem Land, Australia: Social, Ecological and Material Culture Change in the Post-Glacial Period’ by Darrel Lewis, 1988, Oxford: BAR International Series 415, 426 pp. ISBN 0-86054-532-6 (pbk)

Review by Kim Sales

This BAR monograph deals with the chronology and styles of Arnhem Land rock paintings. It is based surprisingly, considering the amount of research undertaken (sixteen years of fieldwork) and the breadth of problems addressed, on a BA honours thesis, and sets out to assess previous work done in the region and provide a reinterpretation and social explanation of the data.

The book is divided into two parts. The first evaluates the methodologies, chronologies and nomenclatures used by two previous principle researchers in the area Brandl and Chaloupka. Then changes in material culture items depicted in the art are used to divide the corpus into four art periods and to name these. The second part critically examines Chaloupka’s chronological model and questions his assertions that the earliest Arnhem Land art style predates 20,000 BP, and that the x-ray art style with its emphasis on aquatic fauna documents the appearance of estuarine conditions in the region c. 9000 – 7000 BP Lewis provides an alternative chronological model for his art period, Interpreting change in terms of differing social strategies adopted in response to postglacial ecological shifts. However, the major volume of the monograph consists of illustrations, both photographs and, line drawings – a welcome visual presentation of a visual medium This allows other researchers to evaluate that first level of rock art analysis, interpretation of what exactly it is being depicted, and gives them access to the data for further analyses.

Much of this book is devoted to a detailed assessment an ultimate rejection, of Chaloupka’s identification and dating of Arnhem Land art styles. Lewis notes that Chaloupka’s methodology is to identify superimposition sequences, acceptable perhaps for polychrome over monochrome paintings, but much more controversial when it comes to assessing relative overlay of monochrome (here red ochre) motifs. Numerous examples are needed, but are apparently rare in Arnhem land; where good examples exist the resultant maze of red lines is too complex to be convincingly separated out; there is no way to assess temporal distance between superimposed paintings, and appearance of relative ‘freshness’ has more to do with pigment and matrix chemical composition that relative age. Lewis concludes that ‘… determining the sequence of overlays through visual means becomes difficult and unreliable’ (p.5), a technical point which does need emphasis, but disappointingly he does not consider the notion that overlay may be a deliberate, conventional strategy of motif and idea association implying cognitive continuity rather than temporal separation. Chaloupka’s methodology is also criticized for inconsistency, for example the lack of clarity over whether identified ‘styles’ are situated chronologically before or after their contents were analysed.

Inconsistency is also the major reason for rejecting Chaloupka’s nomenclature. Styles are labelled variously by their subject matter, the form of their subject matter, their presumed relative age or combination of these factors In addition some of the labels are value-laden, for example the so-called ‘decorative x- ray’ implies loss of interest in internal anatomy’, yet this is not borne out by contemporary Aboriginal informants.

Given that Chaloupka’s chronology is questionably based on his subjective assessment of superimpositions, and that his styles are defined using inconsistent variables, it becomes difficult to accept his dating scheme. Nevertheless Lewis does examine further Chaloupka’s correlation between well-documented prehistoric environmental changes in Arnhem Land from pre-estuarine, through estuarine, to freshwater conditions, and his art styles, and rejects each for varying reasons. For example, the presumed earliest ‘handprints and grassprints style’ dated by Chaloupka to anywhere between 35,000 and 18,000 BP, is criticized by Lewis because it lacks adequate criteria for designation as a style (technique and presumed superimposition are used), it assumes that economically important species necessarily form the subject matter of art, and it is shakily identified with grinding mortars dated archaeologically to 18,000 BP Similarly, Chaloupka’s later style which depicts species introduced to the region as sea levels rose and is associated with the development of estuarine conditions, is questioned. Firstly the ‘estuarine’ species referred to all also spend parts of their lifecycles in freshwater habitats, and so were liable to have been known to local inhabitants long before the onset of estuarine conditions. Secondly, the underlying assumptions that environ- mental and cultural change are congruent, and that art represents a literal record of prehistoric life, are rejected.

Lewis fairly convincingly demolishes Chaloupka’s model of Arnhem Land art styles and chronology, and then proceeds to set up his own schema. His methodology consists of analysing human figures and their associated material culture items, particularly hunting and fighting weapons. Artefacts are identified, associated figures thus grouped. The characteristics of other motifs found in association with these figures are noted and used to assign motifs unassociated with artefact-wielding figures to the initial groupings. This procedure would seem to be one which would obscure anything other than the most blatant intra-group stylistic variability, and which weighs artefact-type disproportionately. On the assumption that different artefacts can be temporally ordered, an idea unlikely to receive much criticism from mainstream archaeology, Lewis places his groupings In chronological order, identifying four main periods in Arnhem Land art. These are labelled the Boomerang. the Hooked Stick and Boomerang, the Broad Spearthrower and the Long Spearthrower Periods. In the ethnographic present in Arnhem land hand-thrown spears and boomerangs are not manufactured or used as weapons, whereas spearthrower technology is. So, following Brandl’s reasoning, Lewis places his periods in chronological order, with the spearthrower periods the most recent, and boomerang figures the earliest. However it is interesting to note that his line drawings of Boomerang period figures never unequivocally depict the boomerang as weapon. Instead we have boomerangs wielded by running or dancing anthropomorphs, decked out in costume and headdress (e.g. ‘dancing skirts’) which belles literalist, practical interpretation of these as hunting scenes, and depicted at least twice in what are seen as ‘ritual’ scenes (Figs 36 and 37) analogous to the rare usage of boomerangs in Arnhem Land in the ethnographic present.

In order to date these positioned periods Lewis looks at their characteristics and stylistic variability, correlating them with prehistoric climatic changes and their presumed social impact.

The Boomerang Period is characterized as highly formalised, widespread and dominated by a single style of depiction. Animated human figures with ceremonial accoutrements, boomerangs, single-piece uniserially barbed spears and hafted stone spears axes are common. This is interpreted as evidence that one widespread social identity existed at this time, with social and cultural stability, established tradition and a desire to express ‘sameness’, and hence reciprocity, in art. This style is said to occur plateau-wide, covering an area analogous to the size of the present day semi- arid and arid zone territories. Lewis concludes that the Boomerang Period was one when Arnhem Land in- habitants had large territories, emphasized links and reciprocity, and were living in a semi-arid climate as was extant during the glacial from c. 18,000 BP to c. 9000 BP.

In contrast. the Hooked Stick and Boomerang Period sees the development of regionalized human figure styles, with the persistence of a widespread element, the Rainbow Snake Complex of motifs. Here the unidentifiable (despite a chapter given over to that purpose) hooked stick and the multipronged multi- barbed spear join the artefact inventory, and confrontation scenes are depicted. The Rainbow Snake Complex motifs are assigned to the period on the (questionable) basis that one anthropomorph from this complex has a similar headdress to those worn by a few of the hooked stick wielding figures. The period is interpreted as one of social reorganization. development of smaller regional group identities Indicating mutual ‘differences’ yet bound by an overarching philosophy of conciliation as illustrated by the composite Rainbow Serpent motif. These changes occur because of demographic growth prompted by increased rainfall and land carrying capacity, together with the demands of environmentally-stressed coastal, near coastal and riverine populations following sea level rise c. 8000 BP

The Broad Spearthrower Period is characterized by the advent of unambiguous spearthrowers. both broad (which disappear later) and cylindrical (which continue through the next period), and the development of high degrees of stylistic variation without regionalization. To Lewis this indicates a period of rapid coastal change. Indeed turmoil. The paintings depict spears with hafted lanceolate spearheads reminiscent of stone spears points dated archaeologically to c. 5500 – 4000 B P.

The Long Spearthrower Period features a material culture assemblage paralleling that of the region’s ethnographic present, and the x-ray style is still produced today, albeit on bark. It is the most recent period, and reflects wetlands environment linked to the development of freshwater lagoons c. 2000 – 1000 BP

In this short space I cannot do justice to Lewis’ complex arguments, varied lines of evidence and sometimes frustrating admixture of critical analysis and interpretation. Nevertheless, although his arguments represent an advance, I have some misapprehensions about his model. For example. we may accept Lewis’ assumption that artefact change in art reflects temporal change, but this is still literalist and does not consider the possibility that artefact depiction has ideological and symbolic dimensions. and may not directly reflect contemporary assemblages. Similarly, the idea that art styles as indicative of shared conventions and concepts informs on social identity and affinity is quite orthodox. However the question of style, what it alludes to and how it is identified is inadequately discussed by Lewis. He purports to follow Hodder (1978) in seeing style as an expression of group identity, yet does not take cognisance of that author’s more recent position (Hodder 1982, 1986) that not all style is used to communicate ethnic identity (sameness v. difference), that style may mirror within-group strategies, even within household/ family strategies.

Lewis’ dating of his first two periods of Arnhem Land art is based on his interpretation of the spatial distribution and significance of styles within the periods I am not convinced that the existence of these styles has been adequately demonstrated. Analysis involving relevant, quantifiable, comparable criteria or variables, chosen for their potential to inform on degree of identity and affinity, is needed. This may not be an easy task, but is preferable to an intuitive assertion that various different styles exist.

Overall, this monograph is imaginative. audacious and ambitious. Lewis’ challenge to Chaloupka is well thought out and damaging. His interpretation of the cosmological significance of the composite Rainbow Serpent is fascinating, as are his speculations on the affinity of his Boomerang figures and the Bradshaw figures of the Kimberley. In addition he dares to move beyond description and taxonomy to provide a social interpretation of art change, even though his social adjustments have environmental spurs. However, his analysis does not include a clear demonstration of the existence of distinguishable styles within the corpus of red ochre paintings, exhibiting affinities or differences.

There is inadequate discussion of the theories dealing with style and its implications, despite the fact that his interpretation of stylistic variation is crucial to the model. This makes it difficult to accept his dating, although we may concur that his periods have validity. Future research can address this problem, and it does not detract from the fact that this is a stimulating book sure to provoke further debate.


Hodder, I. (ed.) 1978 The Spatial Organization of Culture. Duckworth: London

Hodder, I. 1982 The Present Past. B.T. Batsford Ltd: London.

Hodder, I. 1986 Reading the Past. Cambridge University Press: Cambridge.

Sales, K.
Review of 'The Rock Paintings of Arnhem Land, Australia: Social, Ecological and Material Culture Change in the Post-Glacial Period’ by Darrel Lewis
December 1990
Book Reviews
You must be a member to download the attachment ( Login / Sign up )