A different look: comparative rock-art recording from the Torres Strait using computer enhancement techniques.

Brady, Liam M.

Abstract: In 1888 and 1898, Cambridge University's Alfred C Haddon made the first recording of rock-art from the Torres Strait islands using photography and sketches. Systematic recording of these same paintings and sites was carried out from 2000 to 2004 by archaeologists and Indigenous Torres Strait Islander and Aboriginal communities as part of community-based rock-art recording projects. Computer enhancement techniques were used to identify differences between both sets of recordings, to reveal design elements that Haddon missed in his recordings, and to recover images recorded by Haddon that are today no longer visible to the naked eye. Using this data, preliminary observations into the antiquity of Torres Strait rock-art are noted along with recommendations for future Torres Strait region rock-art research and baseline monitoring projects.

**********

Cultural materials from the Torres Strait islands in tropical far northeast Queensland reveal a visually striking repertoire of highly visible, decorative objects, such as turtle-shell masks, headdresses, carved and incised drums and tobacco pipes, sculpture, and canoes. Many of these objects were collected or recorded in the late 1800s and early 1900s by collectors, explorers, anthropologists, and government residents in the islands, and now reside in museums and personal collections all over the world. More recently, a contemporary Torres Strait artistic tradition has emerged that focuses on ancestral subjects in the form of painting and lino prints. This tradition of painting has become internationally famous through recognised artists such as Alick Tipoti, Dennis Nona, Billy Missi, David Bosun, and Victor Motlop (e.g. Mualgau Minaral Artist Collective 2001), and thus continues the distinctive Torres Strait artistic traditions. While description and analysis of these more visible forms of Torres Strait artistic expression are evident in the literature from the region (e.g. David et al. 2004b; Farr 1987; Fraser 1978; Kaus 2004; Moore 1984, 1989; Mosby and Robinson 1998), one form of artistic expression from the Torres Strait islands--rock-art--largely has been overlooked until recently.

AC Haddon, the noted Cambridge anthropologist, made the first recording of rock-art in Torres Strait in the late 1800s. Little rock-art research was undertaken after that time, most of it carried out sporadically by amateur rock-art recorders, anthropologists, and travel writers. (1) Two aspects of the results generated from a systematic rock-art recording project carried out across the Western and Central islands in 2000-04 are discussed here: (1) comparison of the first rock-art recordings from Torres Strait (Kirriri and Pulu) undertaken by Haddon through photography and sketches with those same pictures and sites recently redocumented using computer enhancement techniques; and (2) preliminary data concerning the antiquity of Torres Strait rock-art. Comparative analysis of individual pictures and sites is used to illustrate the usefulness of computer enhancement as a tool to assess the deterioration, recording and preservation of rock-art, and the recovery of faded or deteriorated pictures. Temporal data from recent archaeological research are used to shed some light on the antiquity of Torres Strait rock-art.

[FIGURE 1 OMITTED]

Torres Strait, north-eastern Queensland

The Torres Strait islands (Figure 1) are scattered across a 150km stretch of water between Cape York and south-western Papua New Guinea. Geographically, the islands are divided into three main groups: Western, Central, and Eastern. The geological makeup of the Western and Central islands is granitic (Badu Suite of granite) and high acid volcanic rocks (Torres Strait Volcanics) (Von Gnielinski et al. 1997), while the Eastern islands are made up of acid volcanic rocks (Maer Volcanics) (Willmott et al. 1973).

The islands are home to Islander and Aboriginal communities--Saltwater People--who are renowned as some of the world's most marine-oriented peoples (McNiven 2003; Sharp 1992, 2002). This maritime orientation is reflected in subsistence pursuits (e.g. dugong and turtle hunting) and worldviews which, in most cases, reference the sea. The geographical location of Torres Strait--between the hunter-gatherer realm of Aboriginal Australia and the horticultural realm of Melanesia--has placed the Indigenous inhabitants in a unique position that straddles both cultural realms. While previous discussion has focused on Torres Strait as either a bridge or barrier to natural and cultural diffusion (Walker 1972), Torres Strait Islander and Aboriginal peoples have developed many cultural traditions that are uniquely Torres Strait in character (e.g. rituals and artistic conventions).

The islands are linked with each other, and the Papuan and Australian mainlands, through complex customary exchanges, oral traditions, and kinship. Historical alliances exist between local communities that have created an ongoing state of 'enmity and amity' (McNiven 1998). The most detailed anthropological research carried out in Torres Strait was by AC Haddon and his Cambridge University team of researchers in 1898, while archaeological research has a much shorter tenure--beginning in the early 1970s and continuing sporadically until 2000 when community-based research programs began.

AC Haddon and early rock-art recording

Haddon recorded rock-art using two techniques: sketches and photographs. The former involved simple drawings of images in his notebooks and sketchbooks (e.g. Haddon 1888), while the latter must have been considered slightly more labour-intensive given the bulky nature of photographic equipment in the late 1800s. During Haddon's three visits (1888-89, 1898, and 1914) he amassed a considerable archive of photographs (approximately 500) related to his Tortes Strait research. These are now housed at several museums and institutions around the world, although the largest collection remains at the Cambridge University Museum of Archaeology and Anthropology (Philp 2004:90). (2) The photographic equipment used during his first visit to the islands consisted of bulky, 120 x 165 mm, dry-gelatin, half-plate glass negatives (Philp 2004:94). However, in 1898, specialised photographic equipment was employed, more flexible and sophisticated than that used in 1888, and demonstrating the 'importance the Expedition placed on the production of visual evidence' (Elizabeth Edwards 1998:108). While images of Islanders dressed in both traditional and colonial attire, cultural objects, activities (turtle-butchering, cooking etc.), ceremonies, and landscapes (among others) clearly dominate his collection of photographs, he also included some photographs of rock-art from sites that he visited at Kirriri and Pulu. While many of these photographs have experienced damage over time, such as cracked glass plates or crystallisation on the glass plates, the images Haddon that recorded are still mostly preserved and visible.

Torres Strait rock-art and computer enhancement

The Torres Strait Rock-Art Recording Project (2000-04) consisted of a series of community-based, systematic rock-art recording projects undertaken jointly between Monash University academics and Tortes Strait region Islander and Aboriginal communities. Research aims and objectives were co-devised by the Indigenous communities and researchers to address questions that interested both parties (Brady 2005). Rock-art sites were recorded from eight Western islands, one Central island, and one site from the tip of Cape York. Site types consisted of exposed boulders (paintings found on rock-walls, usually protected by a shallow overhang), boulder complexes creating rockshelters (paintings located on rock-walls and ceilings), and shallow caves cut into hillslopes (paintings located on rock-walls). A total of 983 determinate paintings (pictures that could be formally identified according to their formal attributes) and 298 indeterminate images (images unable to be identified or categorised as a result of heavy deterioration/damage or fading) were documented from 56 sites, the majority (97.4%) painted in monochrome red. A unique dichotomy exists in Tortes Strait region rock-art: paintings have thus far only been documented in the Western and Central islands, while carvings have only been recorded in the Eastern Islands (McNiven and David 2004). (3)

Given the harsh coastal regime of the Torres Strait region many rock-paintings have deteriorated from extensive water and salt damage, leaving some images nearly invisible to the naked eye. Although computer enhancement techniques have been known for over two decades (e.g. Dickman 1984; Rip 1983, 1989), it is only recently that inexpensive software programs such as Adobe Photoshop and Corel Paintshop have made it possible for the enhancement of images to form part of standard rock-art recording methodologies. Computer enhancement of faded or deteriorated rock-art was first tested and carried out on Torres Strait rock-art in 2000 at the Kabadul Kula rock-art site at Dauan in the Top Western islands (McNiven et al. 2000; McNiven et al. 2002a; McNiven et al. 2004). Results generated from this project indicated that computer enhancement was required to make a more complete inventory of a site's rock-art record since many pictures could only be recovered using this technique. As a result, the systematic application of computer enhancement techniques was undertaken on every picture documented during the Torres Strait

Rock-Art Recording Project. Of the 983 determinate pictures identified, a total of 113 pictures (11.5%) had their identification changed from 'indeterminate' to 'determinate' using the computer enhancement technique (Table 1). Rock-art discussed in this article was digitally recorded using Nikon Coolpix 950 (Pulu), and Nikon Coolpix 995 (Kirriri) cameras. Documented paintings, along with high-resolution scans of Haddon's black-and-white photographs, were downloaded onto Adobe Photoshop 7.0 where the following interchangeable tools were used in various combinations to rotate or saturate colours during the enhancement of images: (4)

* saturation

* selective colour

* curves

* hue

* levels

* brightness

* contrast

* channel mixer

After successful image enhancement, rock-art designs can also be converted to black-and-white images for publication. Using the Magic Wand tool, a sample colour from the rock-art design is selected (e.g. red) and the program selects and displays all of the same colours from the selection in the image and converts them to black. The converted image is shifted to the foreground, while a white background is created to highlight the converted picture. The resulting image is the final converted black-and-white digitally enhanced image.

Torres Strait rock-art: late 1800s vs. 2000-04 Kirriri 1

The first, in 1888, rock-art site recorded in Torres Strait was that at Kirriri. The site, a rock-shelter created by small granite boulders leaning against one another, is located within woodland approximately 35 m up from a track. The former chairman of the Hammond Island (Kirriri) Council, Mr Joseph Sabatino, informed John Singe (nd:11) (a former schoolteacher stationed on Thursday Island) that there were more paintings at the site; however, the rocks they were painted on have broken off and become buried. A range of cultural materials have been noted at the site, including shell fragments, ceramics, fish bones, and a hearth. In 1996, Ian McNiven was informed by John Singe that 'half a baler shell containing remnants of red ochre and most likely represent[ing] a pigment container' was found on the ground surface of the site, but this has been lost (McNiven and David 2004:215). Additionally, Haddon (1904a:358) noted that the site lay close to crevices where skulls were found.

[FIGURE 2 OMITTED]

Haddon took two photographs of Kirriri's rock-art; both are in very poor condition with crystallisation of the glass plate obscuring much of the photographed area. Upon close inspection, a 'hammerhead shark' and sections of two 'canoes' are evident (Figure 2). In addition to the photographs, Haddon reproduced several images as sketches in publications (e.g. Haddon 1901, 1904b) (Figure 3). The images documented by Haddon consist of: a 'hammerhead shark', two 'canoes', an anthropomorph, 'dugong', and 'turtle'. He noted that the painted animals were 'representing totems (hammer-headed shark, turtle, dugong, and so on)' (Haddon 1901:185), while the canoes 'were supposed to illustrate the canoes in which the spirits paddle about on calm nights when they want to catch turtle, dugong, or fish' (Haddon 1901:185).

[FIGURE 3 OMITTED]

The first systematic recording of this site was undertaken in 2003 by Brady with members of the Kaurareg community (Brady 2004). (5) The site features some patchy lichen growth on the inner and outer walls of the rock-shelter, while several mud-wasp nests obscure some of the paintings and granular disintegration is apparent. The paintings are in a relatively good condition, with some images clearly visible from a distance, while several others were not discernable to the naked eye and required computer enhancement. A total of 26 determinate pictures were documented--only two of which required computer enhancement in order to be labelled as a determinate image (an anthropomorph and a notched-rectangle shape), while several others had their design elements clarified using this technique. Of the six pictures reproduced by Haddon (through photographs and sketches) the 'hammerhead shark', and both 'canoes' remain the clearest images on the rock-face (visible from a distance), while the 'anthropomorph', 'turtle', and 'dugong' became visible upon a much closer inspection of the rock-face. Computer enhancement was used to reveal additional design elements on the 'turtle' and two 'canoes' (the remaining three images were enhanced by computer to increase their clarity since they were also faded, but not invisible to the naked eye).

The 'turtle' documented by Haddon consisted of a plan drawing of an outlined 'turtle' (oval shape) with a 'head', two front and rear 'flippers' and a 'tail' (Figure 4). In 2003, this image was heavily deteriorated and could barely be identified with the naked eye. Computer enhancement of the picture was used to reveal the full clarity of the image, but also revealed a thin red line bisecting the longest section of the oval-shaped 'shell' of the 'turtle'--a minor, yet structurally significant design element used in the depiction of a 'turtle'.

The first 'canoe' sketched by Haddon reveals an indeterminate section at the left-hand side of the 'hull' (Figure 5). Computer enhancement revealed the indeterminate section to be a faded, slightly upturned, rectangular shape. Haddon's sketch features a crescent-shape at the right-hand side of the 'canoe hull'; that has now been covered over by a white fungus, along with a near-vertical line extending upwards from the 'hull'.

The second 'canoe' also features an indeterminate section at the left-hand side of the image (Figure 6). Two notable features recovered using computer enhancement can be added to Haddon's sketch. First, while a central column with four slightly angled [??]-shapes (identified as decorative 'flags'--a common decoration on Torres Strait canoes (6) extending from the top is easily discernable, a second, smaller column, also with four slightly angled [??]-shapes is located at the extreme left edge of the 'canoe'--an area where Haddon's sketch appears less clear. Second, Haddon's sketch of the 'canoe hull' is incomplete. The left side of the 'canoe' is unfinished--no doubt due to the deteriorated nature of that end of the painting. Computer enhancement has revealed the complete outline of the 'canoe hull', complete with angled lines ('grass fringe') extending from the bottom of the 'hull'.

Pulu

In 1898, Haddon visited the kod ceremonial site complex at Pulu Islet, the location for important ritual and ceremonial activities such as war dances, death dances, initiations for young men, preparation of heads obtained during battle, hunting rituals, and the production of rock-art (Haddon 1904c:3-5; McNiven and Feldman 2003). The three rock-art sites that he recorded there, along with several other previously unrecorded sites, were re-recorded in 2001 by Ian McNiven, and Bruno David, with members of the Goemulgal community (David et al. 2003; McNiven et al. 2002b). Computer enhancement was subsequently carried out on the recorded images in 2002 and 2003.

[FIGURE 4 OMITTED]

Pulu 1. Also known as Menguzi Kula, this site is a large boulder located at the edge of the kod on the beach near the high-water mark. Two panels of paintings, containing both old and recent rock-paintings, were recorded in 2001. The boulder features block collapse, granular disintegration, and large patches of thick black lichen. Haddon (1904c:4) documented 'some nearly effaced paintings in red of various animals and other objects, such as a cassowary, spoon-bill (tapur), curlew (karuri), crayfish (kaiar), canoe and dugong platform', however he did not provide reproductions of any of these images. He also recorded a series of charcoal hand-stencils in a photograph. He noted (1904c:4) that the stencils were made by:

 ... placing the outstretched palm and fingers on the rock, and
 spitting powdered charcoal mixed with water round the hand. The
 handprint thus appears light against a black background. I was
 informed the men used to sit on a mat under the shelter of a stone
 as a protection from the sun, and then might make the pictographs,
 which I understood had no serious meaning.

Systematic re-recording of this site in 2001 yielded thirteen red paintings (seven determinate and six indeterminate). Although computer enhancement of the boulder in 2001 failed to reveal any evidence of the charcoal hand-stencils documented by Haddon, enhancement of Haddon's 1898 image of the boulder was able to reveal a total of six hand-stencils on the rock-face (McNiven et al. 2002b) (Figure 7). Additionally, of the four zoomorphs noted by Haddon (three 'birds'--'cassowary', 'spoon-bill', 'curlew', and a 'crayfish'), only one--a heavily deteriorated, in-filled 'bird' with an oval-shaped 'body', triangular 'tail', reverse L-shape 'leg', and a -shaped 'neck' and 'beak'--could be identified at the site (Figure 8). Based on the picture's form, it is most likely Haddon's 'curlew' or 'spoon-bill'. The 'canoe' and an extended X-structure (possibly representing a dugong hunting platform) were also recovered using computer enhancement (Figure 9). These images (along with a third picture--an in-filled non-geometric shape) were obscured by thick black lichen growth and invisible to the naked eye. Given the effects of coastal damage and lichen intrusion, it is assumed that the remaining images that Haddon recorded have disappeared entirely.

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

Pulu 2. Also at the kod ceremonial site complex, although located away from the high-water mark. The site consists of a 'very small rock-shelter created by a large boulder balancing on a large, flat-topped outcropping boulder ... The shelter created under the overhang protects a near vertical (slightly backsloping), flat rock surface' (McNiven et al. 2002b:89). Damage consists only of small amounts of granular disintegration. Haddon (1904c:5) reproduced the paintings from this site as sketches and has identified the images as a group consisting of 'two muri [spirits] dancing, while a third beats a drum' (Figure 10). Rerecording in 2001 revealed no visible damage to the paintings--all images remain clearly visible from a distance--and computer enhancement could not identify design elements additional to Haddon's 1898 recording.

[FIGURE 7 OMITTED]

[FIGURE 8 OMITTED]

Pulu 3. Also located at the kod and forming part of a large granite boulder complex. The boulder complex consists of a large granite boulder positioned on top of a massive boulder with a flat surface, while two other smaller boulders are located at the base of the large top boulder. Four panels of paintings were recorded in 2001: one on the perched boulder, one each on the two smaller boulders at the base of the perched boulder, and one on the basal boulder. The paintings are protected by a large overhang on the large perched boulder. The large upper boulder is covered in black lichen and patchy white fungus except in the area occupied by panel 1; panels 2 and 3 on the two smaller boulders are heavily weathered and suffer from granular disintegration but lack any black lichen growth; and panel 4 is also nearly entirely covered in black lichen with the painting undertaken in a small area relatively free of lichen growth. Haddon (1904c:5) reproduced a total of three paintings as sketches: two anthropomorphs (identified as muri), and 'one pictograph, according to Gizu, represents a waterspout (baiu), which is the harpoon of the muri ... one man who probably did not know, called it a centipede (sag)'. Additionally, one 1898 photograph depicts the clearly visible 'waterspout'.

[FIGURE 9 OMITTED]

[FIGURE 10 OMITTED]

In 2001, a total of ten pictures (seven determinate and three indeterminate) were systematically documented from the site. Both anthropomorphs remain clearly visible, although initial sketches (by Bruno David) from the 2001 recordings, and subsequent computer enhancement in 2002 detected a distinctive Torres Strait-style 'headdress' extending from the head of one anthropomorph (a single curved line with a series of smaller downward curved lines extending upwards at approximately 45 degrees and spaced at intervals) (Figure 11 upper right).

The 'waterspout' remains visible from a distance and is in good condition. However, a detailed sketch by Bruno David in 2001 revealed several design details not recorded in Haddon's original sketch, while subsequent computer enhancement was also able to clarify many of these 'new' design details and reveal further, deteriorated elements of the image. Comparison between the 1898 and 2001 recordings illustrates two key design details missing in Haddon's sketch (Figure 12, Figure 13a). First, while the 1898 sketch depicts a near-horizontal line extending to the left from the mid-point of the central line of the 'waterspout', sketches and computer enhancement have revealed an indeterminate design drawn directly above the near-horizontal line, and a design similar to the 'headdress' recorded on the anthropomorph from the same site (an upward curving line extending from the upper right-side of the indeterminate image; eight short angled lines extend along the entire length of the curved line) extending from the upper right side of the indeterminate image. Additionally, a thin, slightly wavy line extends from the bottom of the indeterminate image to the upper tip of the 'waterspout'. Second, a series of short, rayed lines are recorded extending from the top of the 'waterspout', while thin parallel lines are drawn alongside the central line of the upper half of the 'waterspout' (Figure 13b). Further higher-resolution digital photography and computer enhancement may be able to shed more light on the indeterminate design drawn above the near-horizontal line.

[FIGURE 11 OMITTED]

Discussion

Comparison between Haddon's recordings in the late 1800s and those using digital technology undertaken between 2000 and 2004 reveal that, in some instances, noticeable differences can be observed in the detail of documentation of specific design elements, and in the preservation of some images.

Haddon's 1888 and 1898 photographs have provided a benchmark for rock-art comparisons. With one exception, all images photographed by Haddon were still easily visible during the recent project, indicating that they are not experiencing any major deterioration (apart from natural coastal weathering processes, e.g. minor salt- and water-damage). However, failure to recover the charcoal hand-stencils from Pulu 1 (after extensive surveying and computer enhancement of the rock-face) suggests that charcoal pigment deteriorates at a much faster rate than ochre in Torres Strait. The durability of red ochre compared with other pigments used in the production of rock-art has been recognised elsewhere (e.g. Chaloupka 1993; Chippindale and Tacon 1998; Clarke and North 1991).

Application of computer enhancement techniques has revealed design elements not recorded in Haddon's sketches. Also, charcoal hand-stencils observed by Haddon at Pulu in 1898 can be clearly illustrated by computer enhancement of his 1898 photograph. Enhancement was used to retrieve a more 'complete' recording of images (e.g. 'canoes' and 'turtles' at Kirriri), and revealed previously undetectable design elements of specific images. These observations suggest that integration of early rock-art recordings with computer enhancement methods have significant potential in the baseline monitoring and management of rock-art sites.

[FIGURE 12 OMITTED]

In only one case could paintings documented by Haddon in the late 1800s not be recorded by the recent project. At Pulu 1, the 'cassowary', 'crayfish', and either the 'spoonbill' or the 'curlew' could not be identified. This site is located near the high-water mark and has experienced the greatest deterioration of any of the sites mentioned; extensive lichen intrusions across much of the boulder probably also have obscured or damaged the other paintings noted by Haddon.

What do these results indicate about the rate of deterioration of rock-art in Tortes Strait, and the antiquity of rock-art on the islands? Comparisons between Haddon's recordings and those carried out over the last few years suggest various rates of deterioration. Some images clearly visible over 100 years ago are still visible today, while others such as those from Pulu 1--located in closer proximity to water than other sites at Pulu--have suffered more rapid deterioration. Our research also suggests that other factors may be responsible for deterioration rates, such as the geological substrate on which paintings are executed (Brady 2005, 2006; McNiven and David 2004). Paintings recorded by Haddon, McNiven and others (2002b), and Brady reveal minimum dates of 103 years for Pulu and 115 years for Kirriri (Brady 2005)--relatively recent in the context of Tortes Strait archaeology.

[FIGURE 13a OMITTED]

Data concerning the antiquity of Torres Strait rock-art is scarce. Indirect dating provides minimum ages for paintings and sites; however, there is a complete lack of absolute dating of paintings and sites. For example, excavation by David and others directly below a rock-art panel at the well-protected inland site of Turao Kula at Mua revealed in-situ powdered ochre, which, they suggested, was a by-product painting there: this dated between 1400 and 1850 AD--at least 150 years ago (David et al. 2004a). This panel was also documented in the 1940s by travel writer Kylie Tennant (1959:193) who observed 'a long-nosed man climbing a coconut tree. There was a row of coconut trees gracefully drawn and true, not just indications, but verifiable coconut trees'; she also noted that the site exhibited the 'clearest painting we had so far seen'. In 2001, about 60 years later, the paintings were no longer visible to the naked eye and could only be seen using computer enhancement--leading David and others (2004a:169) to suggest that this panel was probably executed sometime during the 19th century.

[FIGURE 13b OMITTED]

Oral traditions can be used in establishing the antiquity of Torres Strait rock-art. Of the Somerset rock-art site at Cape York, Nonie Sharp (1992:35) remarked that 'No more drawings have been made on the rock faces of the cave' since the arrival of John Jardine in 1864--the beginning of the European occupation of Somerset. Cessation of painting probably coincided with the decrease in the Aboriginal population as a result of the massacres by the European invaders (Sharp 1992:35), suggesting that the paintings are at least 140 years old.

At Dauan, traditions involving the Kabadul Kula rock-art site shed light on the antiquity of some of the site's rock-art. Margaret Lawrie (1970) recorded an oral tradition that described how some paintings were made by raiders from Kiwai Island in the Fly River estuary who voyaged to Dauan on a head-hunting expedition. The Kiwai arrived secretly at night in many canoes and, while waiting for the dawn, they drew pictures with red ochre (parma) on the boulder now known as Kabadul Kula (Dauan 1). No mention is made of which paintings were made on the boulder or whether or not the Kiwai superimposed their paintings over others. This account provides evidence for the creation of at least some of Dauan's painting before the 1880s since, as McNiven (et al. 2004:247) noted, headhunting raids had ceased in this area by that time.

The limited temporal data (all gathered using indirect dating techniques) available for Torres Strait rock-art points to a relatively recent age for many of the documented images. Future research needs to be directed to absolute dating in order (1) to gain a more holistic understanding of the antiquity of the region's rock-art assemblage, and (2) to understand how rock-art from Torres Strait fits into the occupational trends already established for Western and Central Torres Strait. For example, recent research at the Pulu ceremonial kod site complex indicates that a particular ritual site type (dugong bone mound) began to be constructed sometime between 1540 and 1830 AD (McNiven and Feldman 2003:184). These dates indicate the antiquity of a particular ritual site type at a specific part of the islet. Whether this ritual activity extended to rock-art production is unknown, but absolute dating may be able to place the Pulu rock-art in the wider context of occupation and ritual use of the kod complex (Brady 2005:199).

Conclusion

Comparative rock-art research has revealed variable deterioration of the first recorded rock-paintings from Torres Strait. Whereas charcoal hand-stencils have completely disappeared, most paintings are still, to some degree, visible to the naked eye. Computer enhancement has proven to be a useful technique in this comparative exercise by illustrating how it can be used in the recovery of faded or deteriorated design elements--some of which were unable to be recorded in the late 1800s, yet could be recovered over 100 years later using computer enhancement. Recovery of deteriorated design elements contributes to the preservation and conservation of Indigenous visual cultural heritage; however, research into the age of Torres Strait rock-art--a question asked by many members of Torres Strait Islander and Aboriginal communities--remains to be undertaken. Computer enhancement can provide indirect and imprecise clues, but only absolute dating will be able to provide a clearer picture of the antiquity and history of Torres Strait rock-art.

ACKNOWLEDGMENTS

This research could not have been undertaken without the support of the Islander and Aboriginal communities who have been kind enough to invite me, and fan McNiven, and Bruno David who invited me to work with them in documenting their rock-art. In particular I would like to thank community members who assisted with rock-art recording: (at Kirriri) Roney Wasaga, Billy Wasaga, Greg Tom, Jerry Wasaga, Pando Wasaga, Sam Wasaga; (at Pulu with Ian McNiven and Bruno David) Cygnet Repu, Thomas Mene, Aaron Whap, and Tony Yellub. Special thanks also to the Goemulgaw Kod, and the Buugar Council of Elders on Mabuyag, and the Kaiwalagal Aboriginal Corporation, Kaurareg Elders Corporation, and the Kaurareg Land Trust in south-western Torres Strait for advice and guidance on all cultural matters. Anita Herle (Cambridge University Museum of Archaeology and Anthropology) kindly gave permission to reproduce Haddon's historical photographs. Funding for fieldwork was generously provided through an AIATSIS research grant to LMB and Roney Wasaga in 2003, and a Coastcare Australia grant to the Mabuyag community in 2001. Thanks to my anonymous referees and to Graeme Ward for their comments and suggestions, and to Ian McNiven and Bruno David for kindly providing comments on an earlier draft of this article.

REFERENCES

Brady, Liam M with the Kaurareg Community 2004 The Kaurareg Rock-art Recording and Archaeological Survey Project, Southern Torres Strait, Queensland, Report to the Australian Institute of Aboriginal and Torres Strait Islander Studies (AIATSIS), Canberra, Programme for Australian Indigenous Archaeology, School of Geography and Environmental Science, Monash University (Cultural Heritage Report Series 6).

--2005 Painting Patterns: Torres Strait Region Rock-Art, unpublished doctoral dissertation, Monash University.

--2006 'Documenting and analyzing rock paintings from Torres Strait, NE Australia, with digital photography and computer image enhancement', Journal of Field Archaeology 31(4): 363-79.

Chaloupka, George 1993 Journey in Time: The world's longest continuing art tradition: the 50,000-year story of the Australian Aboriginal rock art of Arnhem Land, Reed, Chatswood, NSW.

Chippindale, Chris and Paul Tacon 1998 'The many ways of dating Arnhem Land rock-art, north Australia,' in C Chippindale & PSC Tacon (eds) The Archaeology of Rock-Art, Cambridge University Press, pp.90-111.

Clarke, John and N North 1991 'Pigment composition of post-estuarine rock art in Kakadu National Park,' in C Pearson and BK Swartz Jr (eds) Rock Art and Posterity: Conserving, managing and recording rock art, Australian Rock Art Research Association, Melbourne (Occasional Publication 4), pp.80-7.

David, Bruno, John Brayer, Ian J McNiven and Alan Watchman 2001 'Why digital enhancement of rock paintings works: rescaling and saturating colours', Antiquity 75:781-92.

--Liam Brady, Ian J McNiven, Judith Fitzpatrick and Goemulgau Kod 2003 'Pulu Rock-Art Recording Project (Tortes Strait)', unpublished report to Coastcare Australia.

--Ian J McNiven, Louise Manas, John Manas, Saila Savage, Joe Crouch, Guy Neliman and Liam Brady 2004a 'Goba of Mua: archaeology working with oral tradition', Antiquity 78(299):158-72.

--Ian McNiven, William Bowie, Manual Nomoa, Peo Ahmat, Joe Crouch, Liam Brady, Michael Quinnell and Anita Herle 2004b 'Archaeology of Tortes Strait turtle-shell masks: the Badu Island cache', Australian Aboriginal Studies 2004/1:18-25.

Dickman, Jeffrey L 1984 'An image digitising and storage system for use in rock art research', Rock Art Research 1(1):25-32.

Edwards, Elizabeth 1998 'Performing science: still photography and the Torres Strait Expedition', in A Herle and S Rouse (eds) Cambridge and the Torres Strait: Centenary essays on the 1898 anthropological expedition, Cambridge University Press, pp.106-35.

Farr, Francine 1987 'Face and crocodile mask from Mabuiag', Art Tribal 1:17-22.

Fraser, Douglas F 1978 Torres Strait Sculpture: A study in Oceanic primitive art, Garland Publishing, New York and London.

Haddon, Alfred C 1888 Unpublished Tortes Strait fieldwork journal, HP papers, Cambridge University Library.

--1893 'The secular and ceremonial dances of Tortes Strait', Internationale Archiv Fur Ethnologie 6:131-62.

--1901 Head-hunters, Black, White and Brown, Methuen, London.

--1904a 'Magic and religion', in AC Haddon (ed.) Reports of the Cambridge Anthropological Expedition to Torres Straits. Volume 5. Sociology, Magic and Religion of the Western Islanders, Cambridge University Press. pp. 320-78.

--(ed.) 1904b Reports of the Cambridge Anthropological Expedition to Torres Straits. Volume 5. Sociology, Magic and Religion of the Western Islanders, Cambridge University Press.

--1904c 'Introduction', in AC Haddon (ed.) Reports of the Cambridge Anthropological Expedition to Torres Straits. Volume 5. Sociology, Magic and Religion of the Western Islanders, Cambridge University Press, pp.1-8.

--1904d 'Funeral ceremonies', in AC Haddon (ed.) Reports of the Cambridge Anthropological Expedition to Torres Straits. Volume 5. Sociology, Magic and Religion of the Western Islanders, Cambridge University Press pp.248-62.

Haddon, AC and J Hornell 1975 Canoes of Oceania (three volumes), Bishop Museum Press, Honolulu, Hawaii (Special Publications 27, 28 and 29).

Kaus, David 2004 'Material culture collections and research from Tortes Strait', Memoirs of the Queensland Museum, Cultural Heritage Series 3(1):93-104.

Lawrie, Margaret 1970 Myths and Legends of Torres Strait, University of Queensland Press, St Lucia.

McNiven, Ian J 1998 'Enmity and amity: reconsidering stone-headed club (gabagaba) procurement and trade in Torres Strait', Oceania 69(2):94-15.

--2003 'Saltwater people: spiritscapes, maritime rituals and the archaeology of Australian indigenous seascapes', World Archaeology 35(3):329-49.

--and Bruno David 2004 'Torres Strait rock-art and ochre sources: an overview', Memoirs of the Queensland Museum, Cultural Heritage Series 3(1):199-208.

--and Ricky Feldman 2003 'Ritual orchestration of seascapes: hunting magic and dugong bone mounds in Torres Strait, NE Australia', Cambridge Archaeological Journal 13(2):169-94.

--Bruno David and John Brayer 2000 'Digital enhancement of Torres Strait rock art', Antiquity 74:759-60.

--Bruno David and Liam M Brady 2002a 'Torres Strait rock art: an enhanced perspective', Australian Aboriginal Studies 2002/2:69-74.

--Bruno David, Liam M Brady and John Brayer 2004 'Kabadul Kula rock-art site, Dauan Island, Torres Strait', Memoirs of the Queensland Museum, Cultural Heritage Series 3(1):227-56.

--Judith Fitzpatrick, Bruno David and Goemulgau Kod 2002b 'Pulu culture site mapping project (Torres Strait)', unpublished report to Environment Australia, Canberra.

Moore, David R 1984 'The Torres Strait collections of AC Haddon: A descriptive catalogue', British Museum Publications, London.

--1989 Arts and Crafts of Torres Strait, Shire, Aylesbury (Shire Ethnography 10).

Mosby, T, and B Robinson (eds) 1998 Ilan Pasin (this is our way): Torres Strait Art Exhibition Catalogue, Cairns Regional Gallery, Cairns.

Mualgau Minaral Artist Collective 2001 Gelam Nguzu Kazi--Dugong My Son: The first exhibition of limited edition linocuts by the artists of the Mualgau Minaral Artist Collective from Mua Island in the Torres Strait, Australian Art Print Network, Darlinghurst, NSW.

Philp, Jude 2004 'Embryonic science: the 1888 Torres Strait photographic collection of AC Haddon', in R Davis (ed.) Woven Histories, Dancing Lives: Torres Strait Islander identity, culture and history. Aboriginal Studies Press, Canberra, pp.90-106.

Rip, Michael R 1983 'Digital recording and image processing of rock art by computer', South African Archaeological Bulletin 38:77-9.

--1989 'Colour space transformations for the enhancement of rock art images by computer', Rock Art Research 6(1):12-14.

Sharp, Nonie 1992 Footprints Along the Cape York Sandbeaches, Aboriginal Studies Press, Canberra.

--2002 Sahwater People: The waves of memory, Allen & Unwin, Crows Nest, NSW.

Singe, John nd Aspects of rock painting in the Torres Strait with an emphasis on Muralug Island Group (to which is attached copies of drawing of rock paintings from various sites), ms, Thursday Island High School.

Tennant, Kylie 1959 Speak You So Gently, Victor Gallancz, London.

Von Gnielinski, Friedrich E, TJ Denaro, P Wellman and CF Pain 1998 Torres Strait region, JHC Bain and JJ Draper (eds) North Queensland Geology (AGSO Bulletin 240), pp.159-64.

Walker, Donald 1972 'Bridge and barrier', in D Walker (ed.) Bridge and Barrier: The natural and cultural history of Torres Strait, Australian National University, Canberra, pp.399-405.

Willmott, WF, WG Whitaker, WD Palfreyman and DS Trail 1973 Igneous and metamorphic rocks of Cape York Peninsula and Torres Strait, Bureau of Mineral Resources, Geology and Geophysics, Canberra (Bulletin 135).

NOTES

(1.) McNiven and David (2004) have provided a review of Torres Strait rock-art.

(2.) Although the official photographer for the 1898 Cambridge Expedition was Anthony Wilkin, Philp (2004:90) notes that other members of the Expedition including Charles Myers, Charles Seligman, and Haddon also took photographs. Edwards (1998:110) also notes that Expedition records reveal 'that Haddon was directing the photographic work'.

(3.) Due to time restrictions, engravings from the Eastern islands could not be recorded.

(4.) David (et al. 2001) provides details on the logic of computer enhancement, and Brady (2005, 2006) further details on the photographic procedure and computer enhancement steps used in this research.

(5.) The Kirriri rock-art site was also recorded by John Singe (nd) although there is no indication that his recording was systematic.

(6.) Haddon and Hornell (1975) provide details of Torres Strait canoe elements.

Liam M Brady's doctoral research examined rock-art in the Torres Strait region through a series of collaborative, community-based research projects. He received his doctorate from the Programme for Australian Indigenous Archaeology, School of Geography and Environmental Science, Monash University in 2006, and is currently a Postdoctoral Fellow at the Centre for Australian Indigenous Studies, Monash University.

<Liam.Brady@arts.monash.edu.au>

Liam M Brady

Centre for Australian Indigenous Studies, Monash University

Table 1: Total number of pictures determined by computer enhancement
(by island)

 Pictures determined Proportion of pictures
 by computer Total determined by computer
Island enhancement pictures (#) enhancement (%)

Dauan 22 53 41.5
lama 2 8 25.0
Pulu 16 103 15.5
Badu 22 188 11.7
Mua 24 100 24.0
Zurath 4 10 40.0
Kirriri 2 26 7.7
Muralag 6 76 7.9
Ngiangu 3 130 2.3
Somerset 12 289 4.1
Totals 113 983