Webb; Grinding patches on granite bedrock around Cue, WA 
petroglyphs are known, but none with only pictograms. In 
contrast, at half the Kimberley sites, grinding is associated 
solely with pictograms (stencils and/or paintings). Both 
petroglyphs and pictograms were reported at the other 
Kimberley sites. If this dichotomy is real, then two sites 
need to bo explained; Mount Ridley, 65 km north of 
Esperance, and The Granites, 90 km south of Cue. Mount 
Ridley may bo unique. Tlie grinding patch there seems to 
have been used solely fpr grinding ochre (Smith 1997). It is 
located within a decorated shelter, whereas patches 
presumed to have been used for grinding flour are usually 
found in the open air. The Granites is a major 
mythological and ceremonial site complex located a few 
kilometres northeast of Mount Magnet where grinding 
patches, petroglyphs and pictograms have been recorded. 
This is significant, if the distribution pattern in Figure 8 is 
real, since it links The Granites to both the Pilbara and the 
Kimberley, or perhaps to the Western Desert. The file on 
The Granites is closed, so I do not know whether any of 
the grinding patches are associated with either 
petroglyphs or pictograms. 
The DIA database lists three rockshelters located about 
90 km west of Cue as housing pictograms and grinding. 1 
have examined the information on file about these sites. 
Two were recorded by DL McCaskill in 1975. He found a 
number of portable grindstones in one; while at the other 
grooves had been cut into the rockshelter wall. At the 
third site, Gunn & Webb (2003;52) noted that a block of 
saprolite had been abraded on its upper surface; a 
treatment that seemed to us similar to the ritually rubbed 
blocks Mountford (1976) reported from the Western 
Desert. These sites have been omitted from Figure 8 
because they do not house bedrock patches. 
Figure 4 shows that two of the groups of petroglyphs 
found at Camel Soak frame grinding patches 1-6. The 
petroglyphs and patches are patinated to the same 
degree, suggesting that they were made 
penecontemporaneously. At many sites in the Pilbara 
petroglyphs have also been found 'close to or even on 
top of engravings' (Flood 1990). This juxtapositioning 
suggests that the relationship of one to the other may 
have been significant when both were made, raising the 
issue of who made what (Flood 1990, 1997; Mulvaney 
and Kamminga 1999). It is generally accepted that 
grinding flour was 'women's work'. Women probably 
used most of the portable grindstones and made most of 
the grinding patches on bedrock, which are considered 
utilitarian, not to be 'classed as rock art' (Flood 1997). 
Whereas, men are thought to have made most of the rock 
art in Australia, particularly at ceremonial sites like 
Walga Rock (Gunn ef al. 1997). Not all art was sacred or 
not to be viewed by women, however. Women clearly 
made handstencils, for example (Gunn 2006). Whether 
they made petroglyphs is less certain. 
Of course, some grinding patches and petroglyphs 
could have been made sequentially and diachronically. 
Flood (1990) .said that grinding took place 'on top of' 
petroglyphs at many places in the Pilbara. In such cases, 
it is possible that by the time the grinding took place the 
petroglyphs had lost their significance for the people 
doing the grinding. Just as, at Walga Rock (Gunn et al. 
1997), new paintings were made over older paintings, 
possibly because the older motifs were no longer 
significant to the people making the new ones. 
At present, the juxtaposition of petroglyphs and 
grinding patches at Camel Soak is locally unique, making 
the site difficult to interpret. It would appear, however, 
that if men made the petroglyphs, they also either made 
the grinding patches, or had no objections to women 
viewing tlie motifs; they were public images that could 
be viewed by anyone. Or, maybe women made both? 
There is little possibility of verifying any of these 
suggestions, now. 
On the other hand, the occurrence at Camel Soak of 
archaeological evidence more commonly found in the 
Pilbara supports the suggestion made above that the 
people living in the area around Cue were more closely 
linked culturally to the Pilbara than to the Southwest. 
Were grass seeds being ground? 
Tindale (1974) said the Yamaji 'were the 
southwestern-most people to extensively exploit grass 
seeds and wet-grind them for the making of forms of 
bread'. One of his informants told him that the Yamaji 
'had an advantage because they placed great reliance on 
grass seed food, whereas other people lived only on the 
hammered seeds of shrubs, did not wet mill grass seed 
and often went hungry'. Who 'the other people' were is 
not mentioned. Tindale (1974) also said the Yamaji stored 
both grass seeds and 'bulibuli' (Tecticornia arborea) seeds 
for at least six months in kangaroo skin bags or 
containers. T. arborea is a halophytic chenopod 
(samphire) found at claypans in the semi-arid zone 
(Bindon 1996). it is called 'kurumi' in the Eastern 
Goldfields (Dix & Lofgren 1974). If the Yamaji wet-milled 
grass seeds, what sort of grindstones did they use? Gunn 
& Webb (2002, 2003) recorded 42 grindstones during 
their surveys around Cue. Of these, 15 are dished; two 
deeply. One of them is pictured (Figure 9 top). The 
remainder have flat grinding surfaces (Figure 9 bottom), 
like the patches described above. 
Tindale (1959, 1974,1977) and Smith (1985, 1988, 1989) 
argued that wet milling grass seeds led to the 
development of specialised grindstones with deep 
grooves, that contrasted markedly with 'amorphous' 
grindstones with flat grinding surfaces that were used 
for a variety of other tasks. My experiments suggested 
that pounding (of hard Acacia seeds) can be carried out 
more easily on a flat grindstone, whereas rubbing tends 
to create grooves, but whether a specialised grindstone 
was necessar)' to process grass seeds is open to question. 
From their analysis of grinding material, Davidson & 
McCarthy (1957) concluded that it was difficult to classify 
grindstones into discrete types because shape seemed to 
reflect degree of usage. Gorecki et al. (1997) concurred; 
arguing that, rather than being discrete types, 
amorphous and double-groove grindstones probably 
represent opposing ends of a continuous sequence of 
grindstone development. One of the portable grindstones 
Gunn & Webb (2002) recorded seems to support that 
contention. It had a shallow groove ground through a 
previously flattened face (Figure 9 bottom), suggesting 
that the way it was used changed over time. Gorecki et 
al. (1997) argued that a really deeply-grooved grindstone, 
like the one in Figure 9 (top) where the deeper groove is 
40 mm deep, would be difficult to use. They suggested 
that such grindstones were actually discards, not 
specialised artefacts for wet milling grass seeds. It might 
123 
