Starkey on behalf of the Kokatha People v State of South Australia

Federal Court of Australia (Full Court)|2018-03-16|Before: White JJ

INTRODUCTION 1 Lake Torrens is the second largest salt lake in Australia. It is situated in the mid-north of South Australia. As the primary judge recorded in the Introduction section to his reasons for judgment (Lake Torrens Overlap Proceedings (No 3) [2016] FCA 899) (the Reasons), the surface of the Lake is "unsuited to long-term occupation, save for Andamooka Island", an island "which protrudes into the western side of the lake [and is] accessible by a causeway from the west" (at [2] and [3]). His Honour added that (at [2]): Lake Torrens does not routinely fill but on occasions (e.g. in 1931) it holds substantial fresh water for months which attracts an abundance of wildfowl and other fauna. That is the last recorded occasion when it held substantial water. Otherwise, and for the present hearings, it is a flat white salt lake with some flat areas running into adjacent lands and some rocky outcrops as the entry into the adjacent lands. 2 Despite its harsh environment, Lake Torrens is the central object of these three appeals filed by each of the Kokatha People, the Adnyamathanha People and the Barngarla People. Before the primary judge, each of the three appellants made a competing and entirely overlapping native title claim to the land and waters comprising Lake Torrens or, to describe their claims in the terms of the apposite provisions of the Native Title Act 1993 (Cth) (the NTA), each appellant filed a native title determination application with this Court under ss 13(1) and 61 in which it claimed that, as a community of Aboriginal people, they held native title rights and interests, as defined in s 223, in relation to the land and waters comprising Lake Torrens and sought an approved determination of native title to that effect, as defined in s 225. 3 By the time of the trial of these three competing claims, each of them had a significant jurisprudential background. At that time, each of the claimant groups had already achieved a consent determination of native title in its favour under s 87(1) of the NTA over a separate area of the shores and surrounding land of Lake Torrens. The primary judge summarised the details of those three determinations as follows (at [4]): To the west and contiguous with the western shore of Lake Torrens, the claim area abuts the area recognised as the native title lands of the Kokatha Uwankara People: Starkey v State of South Australia [2014] FCA 924 (Kokatha Part A) [the Kokatha appellants]. To the east and contiguous with the eastern shore of Lake Torrens, the claim area abuts the area recognised as the native title lands of the Adnyamathanha People: Adnyamathanha No 1 Native Title Claim Group v South Australia (No 2) [2009] FCA 359 (Adnyamathanha No 1) [the Adnyamathanha appellants]. Separating the Kokatha Part A determination area from the Adnyamathanha No 1 determination area at the northern tip of Lake Torrens is a narrow strip of land (approximately 200 m wide at the shoreline) determined to be the southern part of the native title lands of the Arabana people: Doss v State of South Australia [2012] FCA 519. Separating the Kokatha Part A determination area from the Adnyamathanha No 1 determination area at the south-eastern part of Lake Torrens is approximately 6 km of shoreline which forms the northern part of lands determined to be the native title lands of the Barngarla People: Croft v State of South Australia (2016) 325 ALR 213; [2015] FCA 9 (Barngarla); Croft on behalf of the Barngarla Native Title Claim Group v State of South Australia (No 2) [2016] FCA 724 (Barngarla No 2) [the Barngarla appellants]. (Emphasis added)

The Kokatha appellants 4 As will be obvious from the existence of these three appeals, all three appellants failed in their claims before the primary judge. The explanations for their respective failures were summarised in the Conclusion section of the Reasons. With respect to the Kokatha appellants, it was because their claimed rights and interests were contemporary in origin, rather than traditional, and they did not therefore meet the requirements of s 223(1)(b) of the NTA. His Honour described that failure in the following terms (at [771]): In the case of the Kokatha claim, I am not satisfied that they occupied or possessed the claim area according to their traditional laws and customs at sovereignty. Clearly, the very strong evidence concerning their mythology in relation to the western part of the claim area, and indeed the Lake itself, is now of great significance to the contemporary claimants, but I am not satisfied that it is other than of relatively recent origin for the reasons given. Consequently, s 223(1)(b) of the NTA is not satisfied.

The Adnyamathanha appellants 5 With respect to the Adnyamathanha appellants (also referred to in the Reasons as the Kuyani People), his Honour concluded that they had not established a continual substantially uninterrupted connection with the claim area under whatever traditional laws and customs they held with respect to that area at sovereignty. He expressed that conclusion in these terms (at [772]): Although the ethno-historical records provide some support for the Adnyamathanha (Kuyani) People being associated at least with part of the claim area, mainly in its northern part at the time of first European contact, it is difficult inferentially to take that back to the time of sovereignty. In addition, I have not found the evidence of continuing and contemporary connection sufficiently persuasive to conclude that (assuming the existence of native title rights and interests at sovereignty in some part of Lake Torrens) there has been a continual connection to the present time substantially in accordance with their traditional laws and customs so as to satisfy that requirement. There is certainly a remnant knowledge of sites, ceremonies and songs but it is broken and not coherent.

The Barngarla appellants 6 Finally, for similar reasons as for the Adnyamathanha appellants, but with greater concerns about the credibility of their evidence, his Honour was not satisfied about the claims of the Barngarla appellants. He said they failed because (at [773]): [W]hilst the ethno-historical record might support a connection to the southern part of the claim area at effective sovereignty, that is at the time of first European contact, I would not draw the inference that that represented the state of affairs at sovereignty. In addition, I do not think that the evidence shows a continuing connection through and substantially in accordance with their traditional laws and customs, which provides that connection at sovereignty and to the present time. I have given reasons for that conclusion. As I have noted, the evidence did not provide a coherent and detailed picture as to where the area covered by Barngarla laws and customs existed, and there was little independent knowledge of laws and customs to serve the proof of a continuing traditional connection to Lake Torrens or to Andamooka Island. 7 The primary judge remarked on the "counter-intuitive" nature of these conclusions in the immediately succeeding paragraph of his Reasons. There, his Honour contrasted the anthropological opinion about the likelihood that Lake Torrens "would have been subject to traditional rights and interests by an Aboriginal society, or societies, at sovereignty and that it is likely that members of country groups closest to Lake Torrens would likely have had stronger rights and interests in the nearby portions of the Lake, its islands and springs, than others" (at [207]) with the "contemporary significant and credible spiritual connection" (at [774]) each of the appellants presently had to parts of Lake Torrens. Nonetheless, he concluded each had failed to meet the requirements of s 223(1) of the NTA as follows (at [774]): The conclusion reached, having regard to the anthropological evidence referred to, is somewhat counter-intuitive of an appropriate starting point. Each of the claim groups now has contemporary significant and credible spiritual connection to parts of Lake Torrens, but it is not possible, in my view, presently to prioritise one set of spiritual beliefs over the other for the purposes of a finding in terms of ss 223 and 225 of the NTA. As Sutton said in the course of his cross-examination, it is the abutment between the Lakes Cultural Bloc and the Western Desert Cultural Bloc that the Court is being requested to determine, and to use his word which, on the evidence I think is appropriate, to "re- imagine" what existed in 1788. The competing or inconsistent spiritual beliefs, which clearly exist, tend to demonstrate also the lack of the continuance of a dominant particular set of spiritual beliefs of one of the three Claim Groups over that of the others for the purposes of s 223(1)(b) of the NTA from sovereignty to contemporary times. (Emphasis added)

The Kokatha appellants 8 The Kokatha appellants have raised eight grounds of appeal against his Honour's judgment, as follows: 1. The trial judge erred: (a) In finding that at sovereignty, the Kokatha People did not possess native title rights and interests in all or any part of the claim area (Reasons at [713], [724], [726], [771]); (b) In failing to find that at sovereignty, the Kokatha People did possess native title rights and interests in all or at least part of the claim area: Particulars (i) The only part of the claim area suitable for long-term occupation was Andamooka Island and it was geographically proximate to acknowledged Kokatha country on the western shore of Lake Torrens; (ii) The universal acceptance by the anthropologists that Lake Torrens would have been subject to traditional rights and interests and that it is likely that members of country groups closest to the Lake would have had stronger rights and interests in the nearby portions of the Lake, its islands and springs, than others (Reasons at [207]); (iii) The Kokatha People are an eastern group of the wider Western Desert society (Reasons at [13]) and the archaeological evidence demonstrates long-term Aboriginal occupation of Andamooka Island and the nearby western shores of Lake Torrens which is at least consistent with occupation by Western Desert people (Reasons at [384]); (iv) The generally agreed anthropological view accepted by the trial judge that at sovereignty, Aboriginal presence and use of Lake Torrens is very likely to have been sourced in a Dreaming story (Reasons at [382]); (v) The very strong evidence which the trial judge accepted that the connection that contemporary Kokatha People have with the claim area is sourced in their Dreaming stories and the claim area is of great significance to them (Reasons at [771] and see too at [605]-[622]); (vi) The constitutional status and elaborate nature of the laws and customs that connect contemporary Kokatha People with the claim area (Reasons at [605]-[622], [771]); (vii) In Western Desert law there are ceremonies, songs and dances for one of those Dreamings (Wati Nyiru) that are performed by Kokatha men (Reasons at [419], [425], [428], [614]); (viii) The Kokatha men engage in gender restricted ceremonies at Crombie Ridge on Andamooka Island (Reasons at [688]) and at other places on the claim area that are not referred to in the Reasons; (ix) The Kokatha possess a number of very significant sacred objects (Reasons at [410], [688]) and contrary to his Honour's finding at [688], two of those objects relate specifically to the claim area (see Ground 2 below); (x) The restricted men's Dreaming stories and ceremonies that connect Kokatha People to the claim area under Western Desert laws and customs are not just known by senior Kokatha men but are also known by senior Western Desert men who are not Kokatha (Reasons at [411], [438]-[442], [717], [722]; although much of the evidence on this topic is not referred to in the Reasons); and (xi) The trial judge accepted that each of the Kokatha witnesses was honest, genuine and truthful (Reasons at [405], [716]), each gave evidence of having learned at an early age that Lake Torrens was within Kokatha country and in the case of the men having learned from more senior men, including non-Kokatha Western Desert men about the Dreaming stories associated with Lake Torrens (Reasons at [717]). 2. The trial judge erred: (a) In finding that none of the "very significant objects" shown to the Court during restricted men's evidence on Andamooka Island were specific to either the Lake generally or to Andamooka Island (Reasons at [410], [608]); (b) The trial judge should have found, consistent with the evidence, that one of those objects was specific to the Lake generally and another was specific to Andamooka Island. 2A The trial judge erred in finding (at Reasons [609]) that the existence of the Angarta, Urumbulla and Wanampi (sic - Wanambi) stories were mentioned only in passing by Michael Starkey and there was no evidence given describing how those stories related to the claim area. 3. As a consequence of the errors in Grounds 2 and 2A, the trial judge did not have regard to and gave no weight to important evidence of connection. 4. The trial judge erred: (a) In finding (Reasons at [724], [341]) that there was an absence of any material that emerges from ethnographic surveys carried out in the 1980s which tends to associate Kokatha interests with Lake Torrens; Particulars of 1980s material which associates Kokatha interests with Lake Torrens (i) In 1981, Dr Vachon recorded an important Kokatha men's Dreaming story travelling from the west to Lake Torrens (not referred to in the Reasons); (ii) In August 1983, Dr Hagen in a gender restricted report recorded a restricted and highly sensitive Kokatha story concerning the creation of Lake Torrens (Reasons at (309]); (iii) In 1984, Gara and Fitzpatrick recorded a highly secret and sensitive male Dreaming story that travels to the claim area; (iv) In 1989, Dr Gara, the State's historian, wrote in a Heritage Survey Report about the reluctance of his Kokatha informants to divulge site names or the identity of mythic beings associated with a site. (b) In failing to consider whether there may have been cultural reasons why particular information about Kokatha interests in Lake Torrens did not emerge from (some of) the ethnographic surveys carried out in the 1980s: Particulars of reasons why particular information did not emerge in the 1980s (i) The claim area is associated with dangerous and restricted Western Desert men's stories and the repercussions for wrongful disclosure are severe (Reasons at [355], [606], [607] and [612]); (ii) The utmost secrecy attaches to those stories and to the ceremonies and ritual practices associated with them (Reasons at [48], [605]-[607], [611]-[615] and [688]); (iii) Western Desert cultural reasons in respect of the confidentiality of placenames is the likely reason why there are no recorded Western Desert Kokatha placenames in and around Lake Torrens (Reasons at [355], [365]); (iv) In a 2007 report, Dr Vachon said that the Kokatha and other Western Desert men with whom he carried out his 2001 survey in respect of the Olympic Dam [Environmental Impact Statement] area, may have withheld information from him about the mythological significance of the area for undisclosed cultural reasons (this evidence is not recorded in the Reasons); (v) Prof Sutton who was called by the First Respondent said that one of the difficulties for an anthropologist working with Western Desert People is that there is a cultural tradition which militates most strongly against a rapid imparting of information relating to the sacred domain and to the interests of persons in places; (vi) In 1989, Dr Gara, the State's historian, wrote in a Heritage Survey Report about the reluctance of his Kokatha informants to divulge site names or the identity of mythic beings associated with a site. 5. The trial judge erred in finding (Reasons at [719]) that in 1981, Dr Vachon did not record any Kokatha or Western Desert stories relating to Lake Torrens (this was contrary to the evidence) and that Dr Hagen, in 1983, did not avert (sic - advert) in his reports to any Kokatha Dreaming stories relating to Lake Torrens (this was contrary to the evidence and to earlier findings made at [309]). 6. The trial judge erred in finding (Reasons at [325]) that if there were any Kokatha sites on Andamooka Island or immediately adjacent to the western edge of Lake Torrens, they would have been disclosed in survey reports written by Dr Gara in 1989. Particulars (i) Dr Gara noted in that report that tribal boundaries were never clearly delineated and were somewhat fluid throughout the Western Desert (Reasons at [319]); (ii) Dr Gara also wrote in the report about the reluctance of his Kokatha informants to divulge information about sites or the identity of mythic beings associated with the site; (iii) The Appellant repeats particulars (i)-(v) under Ground 4(b) above. 7. As a consequence of the errors in each of Grounds 1 to 6, the trial judge erred in holding (Reasons at [771]) that s.223(1)(b) of the Native Title Act 1993 (Cth) was not satisfied.

The Adnyamathanha appellants 9 For their part, the Adnyamathanha appellants have also raised eight grounds of appeal, as follows: 1. The primary Judge erred in failing to determine native title in accordance with sections 223 and 225 of the Native Title Act 1993 which require: 1.1. the identification of an applicant's traditional laws and customs; 1.2. the identification of how, pursuant to an applicant's traditional laws and customs: 1.1.1. native title rights and interests in land and waters arise; and 1.1.2. the applicant connects to land and waters; 1.3. application of those traditional laws and customs to the subject land and waters to ascertain whether: 1.1.1. any native title rights and interests arose; and 1.1.2. the applicant had a connection to land and waters; and 1.4. identification of whether any connection had ceased to exist. 2. In particular, the primary Judge erred: 2.1. in requiring occupation or physical use of the claim area, or exclusive occupation of substantially all of the perimeter of the claim area, at Sovereignty and at all times thereafter in order to prove connection; 2.2. in finding that attenuation of traditional knowledge somehow resulted in an absence of traditional connection; 2.3. in requiring some "maintenance" of the claim area or its spiritual areas or attributes, where there was no evidence that this was required under the traditional laws or customs of the appellant; 2.4. in failing to adopt the prior findings pressed in the appellant's application under Section 86, or alternatively, for failing to give any reasons for not adopting at least those findings which were unopposed by any party; 3. The primary Judge correctly found (at [774]) that the appellant had "contemporary significant and credible spiritual connection to [a part] of [the claim area]", but then erred in attempting to "prioritise" the appellant's spiritual beliefs in relation to those of the overlapping claimants. 4. Although the primary Judge correctly found that any connection of the first applicant (the Kokatha People) to the claim area was not "traditional" in the necessary sense (at [771]), the Judge then erred in not finding that the appellant's contemporary connection was "traditional", particularly in light of: 4.1. the unanimous expert evidence that Adnyamathanha people were present and resident to the east, north, and west of the claim area at first contact and in all likelihood in greater numbers (or higher concentrations as compared with other Aboriginal people) at Sovereignty; 4.2. the unanimous expert evidence that the claim area was highly likely to be part of the country of some Aboriginal group; 4.3. the ethno-historic evidence, which the Judge correctly found (at [772]) provided support for association of at least part of the claim area with the appellant; 4.4. the majority of expert evidence which located the area of abutment between Western Desert and Lakes Society cultural blocs as being to the west of the claim area (and the minority contrary expert evidence was without foundation once the primary Judge found the first applicant's connection to post-date 1980); 4.5. the overwhelming weight of expert evidence which supported continued connection to the claim area (or part thereof) by the appellant; 4.6. the only expert evidence against continued connection was weak and unpersuasive; 4.7. the clear evidence from the Adnyamathanha aboriginal witnesses that, in accordance with their laws and customs, they held native title rights and interests in the claim area and also connected to the claim area in the same way as their ancestors; 4.8. the existence of country of the Adnyamathanha People abutting almost half the claim area; and 4.9. the availability of the usual inferences of traditional connection back to Sovereignty. 5. The primary Judge erred in the effect and weight he gave to, and the use he made of, the prior determinations of native title. 6. In particular, the primary Judge erred in: 6.1. using the prior determinations as a 'filter' or 'yardstick' to assess whether to give any weight to otherwise credible evidence (including expert evidence); 6.2. denying the availability of an inference that proximate native title rights and interests may extend into an un-determined area (but the Judge was correct to hold that this was unavailable in respect of the first applicant); and 6.3. denying, in respect of the Appellant, the usual inference that contemporary connection is proof of historic connection (but the Judge was correct to deny this was available in respect of the first applicant). 7. The primary Judge erred in: 7.1. failing to disregard or discount the expert evidence of Prof Willis once the fundamental bases for that evidence were found to be incorrect; 7.2. not accepting the balance of the expert evidence (including that adduced by the State) insofar as it consistently supported traditional historic and continuing connection by the appellant; and 7.3. making his own evaluation of the ethno-historic record and substituting that for the expert opinion evidence. 8. The primary Judge erred in holding that a determination of native title could not recognise more than one group of persons having native title rights and interests in a claim area.

The Barngarla appellants 10 Finally, the Barngarla appellants have raised seven grounds of appeal. They are as follows: 1. The primary Judge, having correctly recognised that, inter alia, the ethno-historical record "might" support a connection to the south part of the claim area at effective sovereignty and that there was evidence of the Appellant's physical use of Lake Torrens and relating ceremony, ritual practice and sacred objects, erred in: (a) not finding that the Appellant had the requisite connection with Lake Torrens; and/or (b) not sufficiently explaining why such findings were not sufficient for a further finding that the Appellant had the requisite connection with Lake Torrens. 2. The primary Judge erred in his approach to, and use of, the determinations of native title made in Starkey v State of South Australia [2014] FCA 924 (Kokatha Part A determination), Adnyamathanha No 1 Native Title Claim Group v South Australia (No 2) [2009] FCA 359 (Adnyamathanha No 1 determination), Croft v State of South Australia (2016) 325 [ALR] 213, [2015] FCA 9 and Croft on behalf of the Barngarla Native Title Claim Group v State of South Australia (No 2) [2016] FCA 724 (Barngarla No 2 determination), including: (a) The primary Judge erred in concluding that he was unable to go behind the Kokatha Part A determination, the Adnyamathanha No 1 determination and the Barngarla No 2 determination. (b) The primary Judge erred in concluding that the Kokatha Part A determination, the Adnyamathanha No 1 determination and the Barngarla No 2 determination meant that there was no ready inference to be drawn in favour of, among others, the [A]ppellant. (c) The primary Judge erred in his reluctance to place any weight on the evidence as to Barngarla place names in the area to the west of Lake Torrens because he considered it unclear how that evidence could be coherently used consistently with the Kokatha Part A determination that, at the time of sovereignty, the area to the west of Lake Torrens was the land of the Kokatha People. (d) The primary Judge erred in concluding that if the necessary premise for an expert opinion is inconsistent with the Kokatha Part A determination, the opinion itself cannot be given weight. 3. The primary Judge erred in concluding that there was clearly an appropriate body of information which enabled the First Respondent to support the Kokatha Part A determination. 4. The primary Judge erred in failing to have sufficient regard to the evidence of the anthropologists, Dr Fergie and Dr Lucas, and the linguist, Dr Monaghan. 5. The primary Judge erred in not being satisfied that the Appellant has maintained connection they had to Lake Torrens in accordance with their traditional laws and customs since sovereignty. 6. The primary Judge erred in not being satisfied that there is a proper foundation for the contemporary connection with Lake Torrens required for the determination of native title rights and interests in Lake Torrens in favour of the Appellant. 7. The primary Judge erred in not finding that the rights and interests possessed under the traditional laws and customs acknowledged and observed by the Appellant confer possession, occupation, use and enjoyment of the Determination Area on the native title holders to the exclusion of all others.

THIS COURT'S REVIEW ROLE 11 Before considering the errors alleged in these three sets of appeal grounds, it is convenient to consider the content of this Court's review role in these appeals. The appellants contended that the principles established by the majority in Warren v Coombes (1979) 142 CLR 531 at 551-553 (Gibbs ACJ, Jacobs and Murphy JJ) applied in these appeals. Specifically, they contended that this Court "is in as good a position as the trial judge to decide on the proper inference to be drawn from facts which are undisputed or which, having been disputed, are established by the findings of the trial judge" (at 551). 12 As well, the Kokatha appellants contended that this Court was required to conduct a full and proper review of the findings of the primary judge and give effect to its own conclusions based thereon. The Kokatha appellants also claimed that the primary judge's conclusion to reject their claim was largely drawn by inference from the facts found by his Honour. In particular, they claimed that his conclusions essentially relied on his interpretation of a handful of ethnographic documents. They further claimed that there were no issues of credit involved and the vast majority of the facts were compelling and undisputed. While the Kokatha appellants accepted that they were required to demonstrate error on the part of the primary judge and that this Court would give respect and weight to the views his Honour expressed in his Reasons, they contended that he "had no position of advantage" over this Court. 13 The State is the main respondent in each of these appeals. It challenged many, but not all, of the appellants' contentions concerning this Court's review role. First, it contended that the appellants' reliance on the judgment in Warren v Coombes belied the complexity of the issues which arose in these appeals. It agreed, however, that some of the facts in these appeals fell into the category of "undisputed or established facts" to which Warren v Coombes applied. As an example, it pointed to the three existing consent determinations of native title described above (at [3]) and agreed that this Court was in as good a position as the primary judge to draw inferences from those determinations. Nonetheless, with respect to most of the Kokatha appellants' grounds of appeal (1, 2, 2A, 3, 4(a), 5 and 6) and some of the grounds of appeal of the Adnyamathanha appellants (1 and 4) and the Barngarla appellants (1, 5, 6 and 7), the State contended that the challenges concerned sought to overturn the facts as found by the primary judge, or to contest his Honour's unwillingness to make positive findings of fact. With respect to those grounds of appeal, it contended that it was necessary for this Court to conclude that those findings of fact, or the primary judge's refusals to make such findings, were erroneous. It further contended that, to succeed in those challenges, the appellant concerned would need to show that the findings in question were of the kind described in Fox v Percy (2003) 214 CLR 118; [2003] HCA 22 at [28]-[29] (per Gleeson CJ, Gummow and Kirby JJ) and Robinson Helicopter Company Inc v McDermott (2016) 331 ALR 550; [2016] HCA 22 (Robinson Helicopter) at [43] (per French CJ, Bell, Keane, Nettle and Gordon JJ). 14 Furthermore, the State took particular issue with the Kokatha appellants on the question of the primary judge's advantage. It contended that, in hearing these three native title claims, the primary judge had a unique advantage over this Court, as explained most recently by the five member Full Court in Banjima People v Western Australia (2015) 231 FCR 456; [2015] FCAFC 84 (Banjima) at [57]-[77]. While it accepted that this unique advantage did not relieve this Court of its obligation to conduct a "real review" of the primary judge's reasons, the State contended that it did present particular difficulties for the appellants in advancing the grounds of appeal identified above. 15 For the reasons that follow, I consider the State has more accurately described the review role of this Court in the particular circumstances of the trial that gave rise to these three appeals. 16 These appeals proceed by way of rehearing (s 27 of the Federal Court of Australia Act 1976 (Cth) and Minister for Immigration and Multicultural Affairs v Jia Legeng (2001) 205 CLR 507; [2001] HCA 17 at [75] per Gleeson CJ and Gummow J and [128] per Kirby J). 17 Ordinarily, in an appeal by way of rehearing, this Court's power to interfere with the decision at first instance is exercisable only if it is satisfied that there is error apparent on the part of the primary judge (see Coal and Allied Operations Pty Limited v Australian Industrial Relations Commission (2000) 203 CLR 194; [2000] HCA 47 at [14] per Gleeson CJ, Gaudron and Hayne JJ). Nonetheless, in an appeal by way of rehearing, this Court is obliged: (a) "… to conduct a real review of the trial and, in cases where the trial was conducted before a judge sitting alone, of that judge's reasons" and "to determine whether the judge has erred in fact or law" (Fox v Percy at [25] and Robinson Helicopter at [43] respectively); (b) to "give the judgment which in its opinion ought to have been given in the first instance" or, to "… make its own findings of fact and to formulate its own reasoning based on those findings" (Fox v Percy at [23] and Robinson Helicopter at [43] respectively); (c) to draw its own inferences "from facts which are undisputed or which, having been disputed, are established by the findings of the trial judge" (Fox v Percy at [25] citing Warren v Coombes at 551). 18 There is, however, a qualification to these principles when the error in contention concerns the primary judge's findings of fact. As the State has correctly pointed out, it is that this Court should not interfere with findings of that kind "unless they are demonstrated to be wrong by 'incontrovertible facts or uncontested testimony', or they are 'glaringly improbable', or 'contrary to compelling inferences'" (Robinson Helicopter at [43] citing Fox v Percy at [28]-[29]). It is worth interposing that the reference, in Robinson Helicopter, to the judge's findings of fact appears to include both findings that, to a substantial degree, are based on the credibility of witnesses; and findings concerning "the weight of lay and expert evidence regarding a range of permissible inferences" (Robinson Helicopter at [43]). The former arose in Fox v Percy where the primary issue was who should be believed out of the plaintiff and defendant in their competing claims to have been on the correct side of the road at the time of an accident (see Fox v Percy at [2]). However, in Robinson Helicopter, there was no issue of credibility but, rather, the question was which of five possibilities was most likely to describe the "absence or presence and condition" of a part of a helicopter that crashed (see Robinson Helicopter at [40]). Whether both situations are included, or only the former, the aforementioned qualification with respect to a primary judge's findings of fact is further affected by another important consideration that arises in an appeal by way of rehearing. It is that such an appeal does not involve a completely fresh hearing by the appellate court of all the evidence before the primary judge. Rather, the appellate court proceeds on the basis of the recorded evidence before the primary judge and any evidence admitted as fresh evidence (Fox v Percy at [22]). With the exception of some evidence directed to the relatively peripheral shared claim issue raised by the Adnyamathanha and Barngarla appellants, there was no fresh evidence adduced on any other issue in these appeals. 19 This reliance on the trial record places some "natural limitations" on this Court sitting as an appellate court. They include "the disadvantage that the appellate court has when compared with the [primary] judge in respect of the evaluation of witnesses' credibility and of the 'feeling' of a case which an appellate court, reading the transcript, cannot always fully share". They also include the fact that "the appellate court does not typically get taken to, or read, all of the evidence taken at the trial. Commonly, the [primary] judge therefore has advantages that derive from the obligation at trial to receive and consider the entirety of the evidence and the opportunity, normally over a longer interval, to reflect upon that evidence and to draw conclusions from it, viewed as a whole" (see Fox v Percy at [23]). 20 It follows that, where an appellate court is called on to consider an error concerning findings of fact, it has to "resolve the dichotomy" between its obligations to conduct a "real review" as mentioned above and the need for it to respect the advantages held by the primary judge as outlined above (Fox v Percy at [25] and [27]-[28]). 21 The resolution of that dichotomy in these appeals is significantly complicated by a number of unique features that are commonly associated with the hearing of a native title claim such as these. As the State noted in its submissions, those features were most recently considered by a five member Full Court in Banjima. The "significant advantage" enjoyed by a primary judge when hearing a native title claim was described in that judgment as follows (at [57]): The primary judge heard substantial evidence on country. He alone saw the witnesses give their evidence and was able to weigh that evidence in the balance having seen the land to which the evidence referred as it was being given. He alone saw the performance of the anthropologists in concurrent session. … The State's submissions fail to come to grips with the obvious significant advantage the primary judge enjoyed over this Court in respect of the overall weighing of the totality of the evidence ... 22 At [58] of Banjima, the Court quoted from five Full Court and High Court decisions dating back to 1999 where similar sentiments had been expressed. Of particular relevance to the issues raised in these appeals are the following parts of three of those decisions: [Commonwealth of Australia v Yarmirr (1999) 101 FCR 171; [1999] FCA 1688 at [639]] In the present case there is the added difficulty that the trial judge's evaluation of the facts is premised upon a plethora of factors which influenced his understanding and impressions of: • the evidence given by the Aboriginal witnesses at various locations; • the extensive documentary material; • the relationship between that evidence and material and the sites to which they relate. [Western Australia v Ward (2000) 99 FCR 316, [2000] FCA 191 (Ward FC) at [222]] In the course of presenting these submissions, the State has sought to challenge many specific findings on matters of detail as to the ancestry and connection of applicants and witnesses to parts of the claim area, and for this purpose the Court has been directed to short passages in the evidence of witnesses which appear to contradict particular findings. These aspects of the State's submissions, in effect, invite the court to re-evaluate the mass of evidence received by the trial judge over the course of a very lengthy trial. Such a task would place an impossible burden on an appeal court … [Members of the Yorta Yorta Aboriginal Community v State of Victoria (2001) 110 FCR 244; [2001] FCA 45 (Yorta Yorta FC) at [202]] His Honour's finding that there was a period of time between 1788 and the date of the appellants' claim during which the relevant community lost its character as a traditional Aboriginal community is not to be lightly disturbed on appeal to this Court. A finding that an indigenous community has lost its character as a traditional indigenous community involves the making of a judgment based on evidence touching on a multitude of factors. The hearing before his Honour was long and complex. As is mentioned in [95] above, evidence was taken from 201 witnesses and his Honour visited, and took evidence on, the claimed land on many occasions … 23 The lengthy hearing, the numerous locations at which the Court sat, the large number of witnesses who gave evidence and the extent and complexity of the oral and written evidence were all features that were present in this native title trial. In the State's written submissions, it summarised those matters as follows: The Court sat for 20 days (sic - 21 days at [51] of the Reasons), four of which were on country at five different locations, and produced over 2,600 pages of transcript. Evidence was taken from 25 lay witnesses and ten experts (one of whom did not complete the trial), with multiple contradicting parties. 24 There are two aspects of this summary that deserve further illumination pertinent to the primary judge's particular advantage in this trial. First, the statement that four of the 21 days, or approximately 20% of the hearing, was held "on country", does not fully convey what that process entailed. In this case, as in most, if not all, of the cases mentioned above, the "on country" hearings included Aboriginal witnesses giving oral evidence about sacred sites, ceremonies and stories relating to, and on, the land in question. In the course of giving this oral evidence, some of those witnesses identified the location of certain sacred sites on the land and pointed out some of the features of the landscape to which their sacred stories related. Furthermore, during the evidence that was given at Andamooka Island, the Aboriginal witnesses present showed the primary judge some traditional objects and explained their cultural and sacred significance. Secondly, the concluding statement above about "multiple contradicting parties" bears further explanation. As I have already alluded to above, in this trial, there were three appellants pursuing entirely overlapping claims to the same area of land and waters. While overlapping claims are, unfortunately, not uncommon in native title trials, the contest in the hearing of these claims did place a particular emphasis on the need for the primary judge to assess the evidence and contentions presented by each applicant in the context of the whole dispute. In both these respects, the primary judge in this matter was therefore at a significant advantage. 25 The following statements about the history to the competing claims of the Kokatha appellants and the Adnyamathanha appellants made by the counsel for the Kokatha appellants during oral submissions serves to highlight the background to some of the tense rivalry that was present at the hearing of these competing native title claims: [Originally] the Kokatha claim included the whole of Lake Torrens, and there was no other current claim over it … there were negotiations with the State for a consent determination, and there was going to be a consent determination in relation to the whole of the Kokatha claim area. It was then that the Adnyamathanha put in their claim over Lake Torrens, including Andamooka Island, and it was then that his Honour split the Kokatha claim into part A and part B. Part A was the western side. Part B was the eastern side, covered by the Adnyamathanha claim. Part A went to a consent determination. Part B ultimately went to a hearing. 26 The Part A consent determination referred to in the penultimate sentence above is the Kokatha consent determination, Kokatha Part A, described at [3] above. The last sentence above obviously refers to the trial to which these appeals relate. It is also worth adding that, apart from the State, there was one other respondent party that took an active role at the trial: a mining company called Kelaray Pty Ltd. 27 The highly contested environment in which the hearing of these three competing claims was conducted led to some obvious difficulties in reconciling the competing evidence adduced by each of the three appellants, particularly as that evidence concerned cultural and sacred matters. The primary judge adverted to these difficulties in his Reasons as follows (at [710]): Moreover, as I also remarked earlier in these reasons, much of the evidence led by each of the [Appellants] was inevitably inconsistent with, rather than complementary to, the claims of competing [Appellants]. By way of example, an obvious illustration derives from the [Kokatha Appellants'] assertion that Andamooka Island is a men's only place where women are not permitted, and the direct contradiction of that on the part of the [Adnyamathanha Appellants] and to some degree the [Barngarla Appellants]. It is an area where the competing claims cannot co-exist. Without going into the detail of the evidence, it is also apparent that each of the [Appellants'] evidence concerning stories about Lake Torrens and its creation or Dreaming relationships are not capable of being accommodated fully consistently with the way in which those Dreamings are recorded from the point of view of the separate [Appellants]. It is not necessary, or appropriate, to highlight those elements in these reasons because, in part, that picture emerged as a result of some of the closed evidence compared to some of the open evidence. 28 The reference to "closed evidence" above raises another peculiar aspect of this trial that affects the capacity of this Court to resolve the dichotomy between its review role and the primary judge's advantage. The "closed evidence" sessions in question resulted from the appellants' request, which the primary judge naturally granted, that the secret aspects of their evidence relating to their laws and customs should be treated as confidential. At [48]-[50] of the Reasons, the primary judge described how he dealt with these requests. Throughout the Reasons, the primary judge was alert to the sensitivities associated with striking a balance between preserving these confidentiality restrictions and openly conveying sufficient information to explain the general effect of the secret evidence to which they related. The following excerpts from the Reasons provide some examples: at [334] "I have not here recorded the details of those two Dreaming tracks, and of a further Dreaming of the Wati Nyiru, as the details may be restricted"; at [461], regarding the evidence of Lee Brady, "[i]t is therefore referred to in these reasons to a limited degree for the reasons given"; at [614], regarding the Kokatha Wati Nyiru, "[d]ue to the restricted nature of this evidence, it is not possible to summarise it in further detail"; at [636], regarding the Bivu Story, "[i]t is male gender restricted and it is not possible to summarise it in detail"; at [682], regarding Lee Brady's visit to Lake Torrens, "[t]hat was described in detail during the hearing, but it is not possible to summarise that here as that evidence was restricted to men only for cultural reasons"; and at [689], regarding Lee Brady's description of sacred objects and sites, "[t]he importance of those objects or sites as described was gender restricted evidence and is not described here in detail". 29 Moreover, his Honour expressly adverted to the effect this balancing exercise had on the content of the Reasons in the following paragraph, addressing the submissions of the Adnyamathanha appellants with respect to the evidence of the Starkey brothers adduced in the case of the Kokatha appellants (at [408]): As it is desirable that the reasons for judgment of the Court should, so far as possible, be publicly available, these reasons touch only generally on the detailed elements of confidential evidence. That should not be taken to indicate that its detailed content has not been carefully and fully considered. There is, however, little point in appending a confidential and restricted summary of that evidence where a redacted version is available, and those who may have access to the confidential transcript may do so in any event. That will explain why, in these reasons, the description of the evidence (for instance) of the three Starkey brothers is relatively brief and does not go into detail of particular stories or sites. (Emphasis added) 30 This treatment of the confidential evidence places an unusual burden on this appellate court when it is reviewing the Reasons to identify whether factual errors are present. It is that, when this Court is asked to interpret particular parts of the Reasons where these confidentiality restrictions applied, it must proceed with even greater caution than it would in the ordinary course. The caution that applies in the ordinary course was highlighted by Lord Hoffmann (with the agreement of all other members of the House of Lords) in Biogen Inc. v Medeva Plc [1997] RPC 1. Speaking of a primary judge's reasons in a patent dispute case concerning the issue of obviousness, his Lordship made the following observations (at 45): The need for appellate caution in reversing the judge's evaluation of the facts is based upon much more solid grounds than professional courtesy. It is because specific findings of fact, even by the most meticulous judge, are inherently an incomplete statement of the impression which was made upon him by the primary evidence. His expressed findings are always surrounded by a penumbra of imprecision as to emphasis, relative weight, minor qualification and nuance (as Renan said, la vérité est dans une nuance), of which time and language do not permit exact expression, but which may play an important part in the judge's overall evaluation. … Where the application of a legal standard such as negligence or obviousness involves no question of principle but is simply a matter of degree, an appellate court should be very cautious in differing from the judge's evaluation. 31 Based upon all these observations, the following are the principles that I consider must be applied when determining these three appeals. While this Court must conduct a "real review" of the Reasons to determine whether the primary judge erred in fact or law, in the peculiar circumstances of this native title trial involving three competing claimants, it is necessary to make even greater allowance for the significant advantage the primary judge possessed. In general, making this allowance in these appeals will require particular caution where the alleged error raised by one of the appellants does not involve a question of "principle but is simply a matter of degree". Similarly, where the alleged error hinges on reasonably open, and yet competing, interpretations of the Reasons. In that situation, this caution requires due regard to the many factors mentioned above that "do not permit exact expression[s]" and, most importantly in this matter, the "nuance[s]" associated with recording in a set of reasons the effect of evidence about cultural and sacred matters, where that evidence was given in secret. As well, where the alleged error relates to the primary judge's factual findings (noting the discussion at [18] above), it requires the appellant concerned to establish that the findings in question are "wrong by 'incontrovertible facts or uncontested testimony'" or are wrong because they are "glaringly improbable" or are contrary to "compelling inferences". Furthermore, where the alleged error involves inferences which the appellant claims were wrongly not drawn, as is the case with all of these appeals, it requires this Court to ensure that, before it draws that inference, it is truly based upon undisputed facts and/or facts that are positively established in the Reasons and it requires this Court to be satisfied that inference is compelling such that the primary judge was wrong not to draw it.

THE PRIMARY JUDGE'S REASONS 32 Next, it is convenient to outline how the primary judge approached his task in the Reasons. This outline will also serve to demonstrate how carefully and coherently his Honour structured the Reasons and set out the reasoning he employed to come to his final dispositive conclusions on the claims of these three appellants. 33 The primary judge began by describing the procedural background to these three competing claims, the Lake Torrens claim area and the details of the three existing consent determinations already mentioned above (at [1]-[16]). His Honour then described the protracted history of these three claims (at [17]-[47]) and the details of the trial (at [48]-[84]). In the next section, The Law ([85]-[129]), his Honour set out the legal principles that he considered were pertinent to the determinations he had to make. I will review that section in the next part of these Reasons. After The Law section, his Honour set out some further details of the background to the three claims in a section headed The (Partly Evolving) Claims (at [130]-[166]). Then he briefly described the issues that arose from the three claims (at [167]-[168]) and the contents of the statement of agreed facts submitted by the parties (at [169]-[174]). There then followed an important section where his Honour considered the competing objections to the anthropological evidence called by the parties: Sutton, for the State; Ellis, for the Adnyamathanha appellants; Fergie, Lucas and Monaghan for the Barngarla appellants; and Willis, for the Kokatha appellants (at [175]-[202]). From that point, his Honour turned to conduct a comprehensive review of the evidence called by the three appellants (at [175]-[704]). 34 That review of the evidence is essentially divided into six sections. In each section, the primary judge reviewed a particular category of evidence, and provided his reasoning and conclusions concerning that category at, or near, the end of most, but not all, of the sections. Occasionally, within a section, his Honour also recorded conclusions with respect to particular pieces of evidence, or particular issues. The six sections proceed in a logical progression with the conclusions drawn in each section generally leading to the subject matter in the succeeding section. 35 The review of the evidence is then succeeded by the final Consideration section ([705]-[769]). In that section, his Honour essentially did two things. First, he drew together the more important conclusions he had reached in the course of his review of the evidence and the submissions made by all the parties with respect to that evidence. Secondly, he expressed his final dispositive reasons for rejecting each of the appellants' claims. It is apparent that that section does not purport to reiterate all of the conclusions his Honour reached in his review of the evidence, much less all the supporting evidence for those conclusions. It is clearly intended to identify the critical factors that affected his Honour's final dispositive reasoning with respect to each claim. I interpose to note that, at, or near, the beginning of my consideration of each appellants' grounds of appeal below, I will summarise the dispositive reasoning in this final Consideration section as it applies to the claim of the particular appellant concerned. 36 Finally, in the final Conclusion section of the Reasons ([770]-[775]), his Honour summarised his final conclusions and made some further general observations. The most important of those conclusions have already been set out above (see at [4]-[7]). There is also a section after this final Conclusion section that deals with the issue of extinguishment (at [777]-[819]), however, that issue does not arise in these appeals. 37 The six sections into which the review of the evidence is divided are as follows: (a) the evidence related to the claim area at sovereignty from [203] to [213] - the three consent determinations mentioned above (at [3]) are considered in this section, the common ground between the anthropologists is recorded at [207] and the countering positions of each of the appellants are set out; (b) the Historical and Early Ethnographic Evidence from [214] to [275] - this review was based on the Gara report of 19 October 2015 tendered by the State (at [71]) and considered the records from first contact - about 1845 on the eastern side of Lake Torrens and 1850 on the western side - including at least 17 named sources from Shurmann in 1844 to Capell in 1963; (c) the More Recent Ethnographic and Historical Evidence from [276] to [345] - the records from Platt in 1967 to Horton in 1994 are reviewed, including 12 other named sources, some of which made multiple contributions, for example, Hagen; (d) the Place Names and Linguistic Identity Evidence from [346] to [366] - this section reviewed the reports prepared by Sutton and Monaghan, who advanced the proposition that there was no place name or linguistic evidence to support the Kokatha appellants' connection to Lake Torrens in the west before the 1850s, but there was evidence to support such connection by the Barngarla appellants in the south-east and the Kuyani in the west (see at [350]); and the countering arguments to those propositions put by the Kokatha appellants (see at [355]-[357]); (e) the Archaeological Evidence from [367] to [386] - this review considered the countering opinions of Professor Draper called by the Kokatha appellants and Mr Carver called by the Adnyamathanha appellants, the areas of agreement between them (at [372]) and the disputed issues (at [373]); and (f) the Lay Evidence from [387] to [704] - this review examined the overlapping ancestral, cultural and geographic relationships between the members of the three applicant claim groups (at [387]-[400]); it summarised the evidence of each of the appellants' lay witnesses (Kokatha at [401]-[456], Adnyamathanha at [457]-[523] and Barngarla at [526]-[601]), it also summarised the Dreaming stories and beliefs recounted by each of the appellants' lay witnesses (at [603]-[675]); and finally it summarised the evidence of other activities with respect to the claim area (at [676]-[697]). 38 The main conclusions his Honour reached during his review of the six categories of evidence above were as follows: (a) that all the anthropologists accepted that the most likely position at sovereignty was: 207 … that Lake Torrens would have been subject to traditional rights and interests by an Aboriginal society, or societies, at sovereignty and that it is likely that members of country groups closest to Lake Torrens would likely have had stronger rights and interests in the nearby portions of the Lake, its islands and springs, than others. (b) that the Historical and Early Ethnographic Evidence did not support any of the three claims: 270 The review of that material to this point does not point in any significant way to supporting any of the three [Appellants'] claims. 271 If one starts with the premises of the three claim groups at settlement occupying the areas as determined (and so virtually all the areas around Lake Torrens), until the work of Mountford and Tindale in the late 1930s and early 1940s, the material does not focus on the "occupation" or the exercise of traditional laws and customs with respect to the area of Lake Torrens, including Andamooka Island, by any particular group of Aboriginal people. 274 There is some ethnographic material from the next decades which places the Barngarla People on the western side of Lake Torrens up to about the Carrapateena Arm, and the Kuyani People on the western side of Lake Torrens north of the Carrapateena Arm, including those adjacent parts of the Lake itself (Pratt 1967, Elkin 1976), although Elkin's later work also places the Kokatha in that vicinity. To the extent that such material would exclude the Kokatha People as the traditional owners of the area to the west of Lake Torrens, both north and south of the Carrapateena Arm, it must be given no weight for the reasons already given. (Emphasis added) (c) that the more recent ethnographic and historical evidence did not support any of the three claims: (i) at [337], with respect to all applications: The material since about 1980 also does not provide any sound basis, absent the interpretive "insights" (or contentions) of the experts about the traditional owners of Lake Torrens itself. (ii) at [341]-[342], with respect to the Kokatha application: 341 On that material, in my view, there is little to support the claims of the Kokatha People to be the sole traditional owners of Lake Torrens either at sovereignty or at any later time, at least up to the late 1980s, in the ethnographic material. … Indeed, given the nature of the informants to certain of those ethnographers and to Gara as an historian it must be said that the material tends to show that, at least until the late 1980s, the Kokatha People themselves did not assert to those sources that their native title rights extended into Lake Torrens itself. 342 If the Kokatha People's claim is to be made out, it must find its proof directly in the lay evidence, or the expert evidence itself (relevantly that of Willis) or from other evidence. It is certainly the case that the picture presented by the Willis-Azziz report in 1996 is a somewhat different one. (iii) at [343]-[345], with respect to the Adnyamathanha and Barngarla applications: 343 The same conclusion is indicated with respect to the Adnyamathanha People's claim and the Barngarla People's claim from that material. There is no cogent ethnographic material up to the late 1980s which attributes to either of them the occupation of Lake Torrens at all, or at least in any sense reflecting occupation or use of Lake Torrens in accordance with their traditional laws and customs. In each of their cases, there is some suggestion of such a relationship with Lake Torrens, partly by direct but minor references which - for the reasons given - I do not find at all persuasive. I do not think that material combined with the fact of their occupation of the eastern side of Lake Torrens, or parts of it, points to Lake Torrens being within their traditional country during the time of the observations recorded. 344 As I have said, I am unable to place any real weight upon the fact of adjacent occupation. The Kokatha Part A determination means that, at and from settlement, the Adnyamathanha and Barngarla Peoples were not the traditional owners of the land immediately to the west of Lake Torrens. That was and is the country of the Kokatha People. The fact that Adnyamathanha and Barngarla People, in the period from settlement may have been and been recorded on the western side of Lake Torrens over that period does not demonstrate or tend to demonstrate that they had at settlement, and have continued to have, any particular traditional relationship with Lake Torrens. In the light of the Kokatha Part A determination, their presence on that country (as clearly occurred from time to time ) does not support the step of saying that their traditional country surrounded Lake Torrens and therefore Lake Torrens is and has been within their traditional country from settlement and since. 345 The ethnographic material then, if assessed on the basis of movement south and west from the northern and eastern boundaries of Lake Torrens for the Adnyamathanha People and if assessed on the basis of movement north and west from the southern boundary of Lake Torrens, does not provide evidence of their occupation and use of Lake Torrens so as to contribute towards a state of satisfaction that they, or either of them, were and are the traditional owners of Lake Torrens or particular parts of it at sovereignty or continuously since sovereignty. (Emphasis added) (a) that the Place Names and Linguistic Identity evidence did not inform the identification of the Aboriginal People holding rights and interests over Lake Torrens (at [364]): Having regard to the decision in Kokatha Part A the pattern cannot show that that presence demonstrated that the Barngarla (or Kuyani) People had at, and therefore since, sovereignty native title rights and interests under their traditional laws and customs in the area west of the western boundary of Lake Torrens. Once that step is taken, it is hard to draw any inference from the linguistic analysis which informs the identification of the Aboriginal People who, by their traditional laws and customs, held native title rights and interests over Lake Torrens or part of it. (Emphasis added) (b) that the Archaeological Evidence did not support the claims of any one of the three appellants (at [386]): That evidence confirms the anthropological evidence that it would be very surprising, given the remarkable features of Lake Torrens, if there were no pre-sovereignty Aboriginal society or societies which had firm cultural connections to Lake Torrens under traditional laws and customs. However, for the reasons given, I do not regard it as tending to add to the proof, more probably than not, that the archaeological materials show that the Kokatha engaged in the activities which they reveal and used the sites where they were carried out, rather than the Adnyamathanha (Kuyani) People or the Barngarla People. (Emphasis added) (c) (i) the Kokatha appellants: that despite accepting them as truthful, the Kokatha appellants' lay evidence did not support their claims: (A) general conclusion at [456]: As that review of that evidence indicates, it shows an awareness of the fact that Lake Torrens is, in Kokatha culture, a men's only place. It does not go further than that to advance the [Kokatha Appellants'] claim, as it is otherwise evidence confirmatory of the facts which underlie the Kokatha Part A determination. (B) particular conclusion at [405] about the reliability of the evidence of the Starkey brothers: The first is prompted by the submission of the [Barngarla Appellants] concerning the reliability of the evidence of the three Starkey Brothers. I did not discern in the extensive course of the evidence any reason to doubt the honesty of any of the lay witnesses, including each of the Starkey brothers. Obviously different witnesses have different personalities. They express themselves more or less forcefully, and more or less positively. That does not mean that the more forceful ones are more or less truthful or persuasive than the less forceful ones. The overall picture I formed was that each of the witnesses was honest and genuine. (C) particular conclusion at [442] regarding Mr Mungkari's evidence: I did not think his evidence really advanced the case of the [Kokatha Appellants] in any real way. (D) particular conclusion at [452] regarding the evidence of the three Kokatha women lay witnesses: None of these witnesses knew of any other story concerning Lake Torrens or, in particular, Andamooka Island. (Emphasis added) It should be noted that, in addition to the above conclusions, his Honour came to a number of important conclusions with respect to the Kokatha appellants' lay evidence in the final Consideration section. Those conclusions are summarised later in these reasons. (ii) the Adnyamathanha appellants: that despite accepting them as truthful, the Adnyamathanha appellants' lay evidence did not support their claims: (A) general conclusion at [524]-[525]: 524 As with the witnesses for the [Kokatha Appellants], I found the witnesses for the [Adnyamathanha Appellants] to be impressive, genuine and truthful. The reliability of what each said, on the whole, was not contentious. There are nevertheless several respects on which their evidence must be assessed in the context of the evidence as a whole. 525 The evidence about the Adnyamathanha's spiritual or cultural or social connections with Lake Torrens under their traditional laws and customs was largely confined to the eastern borders of Lake Torrens. It was in part quite contemporary, and did not have great historical depth. Certain evidence was more geographically extensive, or concerned stories of spiritual significance, which will be further considered later in these reasons. (B) particular conclusion at [471] concerning the truthfulness of Lee Brady's evidence: Despite that, I accept that his evidence is truthful when he says that Lake Torrens (including Andamooka Island) is Kuyani or Adnyamathanha country. (C) particular conclusion at [479] concerning the caution to be applied to the evidence of Michael McKenzie: To an extent, it is necessary to be cautious about giving weight to his evidence where it relates to the area to the west of Lake Torrens and is or may be inconsistent with the findings underlying, or the orders made, in Kokatha Part A. Those informants included Max Thomas. The fact that his evidence about Lake Torrens itself, and including Adnamooka (sic - Andamooka) Island, is in part sourced or learned from those who gave him history or stories not consistent with Kokatha Part A also causes me to pause before placing much weight on his evidence directly concerning the claim area. (D) particular conclusions at [502]-[504] concerning the effect of the cross-examination of Regina McKenzie: 502 The cross-examination elicited that those stories, as related by her, had limits of knowledge or different details from the versions related by others, including the place on the west of Lake Torrens where Kadni had died. 503 During her evidence, she in effect acknowledged that the Barngarla People had a traditional interest in the southern section of Lake Torrens, south of Willochra Creek. Her father had told her Adnyamathanha country was to the north of that area of the Lake. She was always accompanied by Barngarla People, generally Lorraine Dare, when visiting that area so she had implicit consent to do so. 504 The cross-examination explored the reasons why the present Adnyamathanha Application was over Lake Torrens when the application upon which Adnyamathanha No 1 was based extended only to the eastern side of Lake Torrens. It also brought out that the research carried out by Vachon in the early 1980s, referred to above, which involved extensive inquiry of Adnyamathanha People including Regina McKenzie, did not procure the information that Lake Torrens was Adnyamathanha country. That material is, of course, of some significance to what is made of this and other evidence adduced by the [Adnyamathanha Appellants]. It is further considered later in these reasons. (E) particular conclusion at [523] concerning the evidence of Beverly Patterson: She has always believed Lake Torrens is part of her Adnyamathanha country. She learned that from her grandparents and her parents. Her father had once referred to a big fight regarding ownership of the Lake. That topic was only a vague theme through some of the evidence. She did not know much about Dreaming stories concerning the Lake, and what knowledge she has may have come from more recent research she has undertaken. She did, however, have a general understanding of Adnyamathanha laws and customs. (Emphasis added) Again, it should be noted that, in addition to the above conclusions, his Honour came to a number of important conclusions with respect to the Adnyamathanha appellants' lay evidence in the final Consideration section which are summarised later in these reasons. (iii) the Barngarla appellants: that despite accepting them as truthful, the Barngarla appellants' lay evidence did not support their claims: (A) general conclusion with respect to their truthfulness at [529] and [582]: 529 As in the case of those witnesses called by the [Kokatha and Adnyamathanha Appellants], each of the Barngarla witnesses apparently gave evidence to the best of their ability and I am satisfied were witnesses of truth. The significance of their evidence must, of course, be assessed in the context of the whole of the evidence. That is a matter addressed later in these reasons. 582 It is convenient to repeat that the processes or procedures and consultations by which the history of claims over Lake Torrens occurred, as explored with the various witnesses including Linda Dare, did not ultimately cause me to doubt the honesty of any of those witnesses. In respect of each of the witnesses asked about that, it was apparent that advice, pragmatism, and the prospects of securing a consent determination with compromise over claim areas, negotiated at different levels of responsibility, means that those discussions do not provide any foundation for doubting the integrity of any one or other of the witnesses. (B) particular conclusion concerning the evidence of Graham Richards at [547]: He said that he has some memory difficulties which explained the generality of his evidence. Whilst his evidence was generally consistent with some other evidence, I do not think it advanced the case of the [Barngarla Appellants] to any material degree. (C) particular conclusions concerning the evidence of Rosalie Richards at [558] and [560]: 558 I am therefore troubled about accepting her evidence on the issue of the western boundary of Barngarla country. 560 I have confidence in her evidence, so far as it is appropriate to accept it having regard to those two reservations. I accept she is a reliable reporter of what she learned from Leroy Richards. In that, I note the particular purpose of Leroy Richards and his brother arranging a trip in October 2000, to research and record for posterity the stories and sites of significance in the area. (D) particular conclusion concerning the statement of Eileen Wingfield at [594]: It is not clear to me how that statement advances the claim of the Barngarla [Appellants]. (Emphasis added) In addition to these conclusions, his Honour came to other conclusions with respect to the Barngarla appellants' lay evidence in the final Consideration section. Those conclusions are summarised later in these reasons. 39 Finally, I need to outline the contents of an important section of the Reasons which affected the claims of all three appellants. It appears as a sub-section under the heading Preliminary Remarks in the final Consideration section of the Reasons ([705]-[711]). In that sub-section, the primary judge set out the reasons why he did not consider he could draw an inference in favour of any of the three appellants. 40 At the outset (at [706]-[707]), his Honour referred to the observations he had made earlier in The Law section about the common availability of such an inference in native title claims (see at [46(g)], [46(h)] and [49] below). With these observations in mind, his Honour then mentioned a number of factors that may support an inference "to fill the gap in evidence" such as physical and temporal proximity to the claim area (at [706]). With respect to the former, he instanced "a broader cultural bloc of traditional people holding the native title rights and interests in … adjacent country" (at [707]). With the latter, he instanced observations made concerning the claim area and "in relation to an Aboriginal group at the time of the first recorded observations" (at [707]). 41 His Honour also identified a number of features of these three competing claims which stood against such an inference being drawn in favour of any one of the appellants. He included the fact that the archaeological evidence supporting "Aboriginal activity and use on and around the western shore of Lake Torrens [and on Andamooka Island] of considerable antiquity" did not allow an inference to be drawn in favour of "one or other of the three [Appellants]" (at [706]). He also observed that he was cautious about using the linguistic evidence to support such an inference (at [711]) and nor did the evidence of the anthropologists that it was "probable that … there were people who constituted a particular group of Indigenous Australians who 'occupied' in some sense, a part, parts, or the whole, of Lake Torrens including Andamooka Island at sovereignty" (at [708]). Furthermore, his Honour considered that the proximity established by the fact of the existence of the three consent determinations to the area surrounding the Lake Torrens claim area meant that "there is no ready inference to be drawn in favour of any one of the three [Appellants] in respect of their claims". Specifically, he said (at [709]): In this matter, such inferences are not routinely available. That is because there are three Determinations of native title (putting aside the Arabana determination abutting part of the northern section of Lake Torrens) surrounding entirely the present claim area. It is also common ground that the Kokatha Part A determination is a determination in favour of the Kokatha People, as members of the Western Desert Cultural Bloc and the Adnyamathanha No 1 and Barngarla No 2 determinations reflect groups who are members of a separate Lakes Cultural Group or Bloc. Those matters mean that, in relation to these three Applications, there is no ready inference to be drawn in favour of any one of the three [Appellants] in respect of their claims over the whole of Lake Torrens (modified in final submissions particularly by the [Barngarla Appellants]). 42 Lastly, his Honour noted (at [710]) (see at [27] above) that there were significant conflicts in the evidence of the different lay witnesses called by each of the appellants. That included conflicts about the status of Andamooka Island and inconsistencies (the details of which his Honour did not disclose for confidentiality reasons) in the stories about, and Dreaming relationships with, Lake Torrens. On this aspect, his Honour remarked earlier in the Reasons on the mutually destructive effect of this conflict for each claimant as follows (at [45]): At the outset the Court observed that the further the individual cases of the [Appellants] were advanced, the more there would be evidence directed also to showing that another [Appellant], or Applicant group, did not have the native title rights which were asserted. There cannot be three sets of exclusive native title rights held by different groups … 43 Another example of this conflict occurred in the evidence of Mr Lee Brady, one of the main lay witnesses called by the Adnyamathanha appellants. Mr Brady contradicted the evidence of the Starkey brothers in important respects, as his Honour recorded in the Reasons as follows (at [472]-[473]): 472 As a Western Desert lawman, he said the present claim area so far as he knows, is not an important Western Desert place or a place for Western Desert ceremonies. Nor did he know of Western Desert stories (including those to which the Starkey brothers referred in their evidence) concerning Lake Torrens or Andamooka Island. 473 The [Adnyamathanha Appellants] in submissions stressed Lee Brady's evidence that he, as a senior Western Desert man, had never heard the versions of the stories or mythologies told by the Starkeys, except possibly from the Starkeys. This was corroborated by Mick McKenzie. Lee Brady was also definitive that Lake Torrens is not an important men's place under Western Desert law; that it did not belong to the Kokatha; and that he had never heard old people talking about that … 44 Having explained why an inference could not be drawn in favour of any of the three appellants, his Honour said (at [711]) that success in their claims "must necessarily be found in such evidence as there is, either from the lay witnesses, or from the recorded ethnographic and historical documentation, and as the [Kokatha appellants] submitted, as informed by the expert anthropological evidence". He added that he had already explained why the expert archaeological evidence and the linguistic evidence did not "inform my decision in relation to the competing claims". His Honour then turned to consider that evidence with respect to each of the three appellants in turn. As I have already stated above, I will review those parts of the final Consideration section when I come to consider each appellants' grounds of appeal.

THE CONNECTION INQUIRY 45 As is the case in the determination of most native title claims, an inquiry that is commonly referred to as "the connection inquiry" was a central focus at this trial and the primary judge's Reasons above. It is therefore convenient, at this point, to identify the legal principles that affect that inquiry. The two provisions of the NTA that are of central importance to the connection inquiry are ss 223(1) and 225. They provide as follows: 223(1) The expression native title or native title rights and interests means the communal, group or individual rights and interests of Aboriginal peoples or Torres Strait Islanders in relation to land or waters, where: (a) the rights and interests are possessed under the traditional laws acknowledged, and the traditional customs observed, by the Aboriginal peoples or Torres Strait Islanders; and (b) the Aboriginal peoples or Torres Strait Islanders, by those laws and customs, have a connection with the land or waters; and (c) the rights and interests are recognised by the common law of Australia. 225 A determination of native title is a determination whether or not native title exists in relation to a particular area (the determination area) of land or waters and, if it does exist, a determination of: (a) who the persons, or each group of persons, holding the common or group rights comprising the native title are; and (b) the nature and extent of the native title rights and interests in relation to the determination area; and (c) the nature and extent of any other interests in relation to the determination area; and (d) the relationship between the rights and interests in paragraphs (b) and (c) (taking into account the effect of this Act); and (e) to the extent that the land or waters in the determination area are not covered by a non-exclusive agricultural lease or a non-exclusive pastoral lease - whether the native title rights and interests confer possession, occupation, use and enjoyment of that land or waters on the native title holders to the exclusion of all others. 46 The most convenient way to illuminate the legal principles pertinent to the two provisions of the NTA set out above, as they affect the connection inquiry, is to go to that part of The Law section of the Reasons where his Honour extracted those principles from the relevant authorities. The following is a summary of that part: (a) at [85], the relevant starting point for the consideration of an application for a native title determination is the NTA, citing State of Western Australia v Ward (2002) 213 CLR 1; [2002] HCA 28 (Ward HC) at [16] and [25] per Gleeson CJ, Gaudron, Gummow and Hayne JJ; Members of the Yorta Yorta Aboriginal Community v State of Victoria (2002) 214 CLR 422; [2002] HCA 58 (Yorta Yorta HC) at [32] per Gleeson CJ, Gummow and Hayne JJ; and Commonwealth of Australia v Yarmirr (2001) 208 CLR 1; [2001] HCA 56 at [7] per Gleeson CJ, Gaudron, Gummow and Hayne JJ; (b) at [88], for a determination of native title to succeed, all of the elements in s 223(1) of the NTA must be "given effect" to, citing Yorta Yorta HC at [33]; (c) at [89], the native title rights and interests defined in s 223(1) must have three characteristics: (a) they are rights and interests which are "possessed under the traditional laws acknowledged, and the traditional customs observed", by the relevant peoples; (b) by those traditional laws and customs, the peoples "have a connection with" the land or waters in question; and (c) the rights and interests must be "recognised by the common law of Australia", citing and quoting Ward HC at [17]; (d) at [90], consequently, the approach required by the Court in determining whether native title rights and interests exist in an area is to consider the laws and customs of the present day claimant society and to ascertain whether, under those laws and customs, the members of that society have rights and interests in the claim area. If they do, the Court then asks whether those laws and customs can be said to be "traditional laws" or "traditional customs", noting that the concept of "traditional" is one which accommodates adaption of those laws and customs with the evolution of the traditional society concerned, citing Croft (on behalf of the Barngarla Native Title Claim Group) v South Australia (2016) 325 ALR 213; [2015] FCA 9 at [640]; (e) at [91], the only rights or interests in relation to land or waters which are recognised after the assertion of sovereignty are those traditional rights and interests that find their origins in traditional laws and customs established at or before sovereignty, citing Yorta Yorta HC at [43]-[44]; (f) at [92] and [93], issues of extinguishment aside, the native title claimants have both an evidentiary and an ultimate onus of proof. That is unaffected by the difficulty that is often associated with discharging that onus, citing Ward FC; Daniel v State of Western Australia (2004) 138 FCR 254, [2004] FCA 849; Harrington-Smith on behalf of the Wongatha People v Western Australia (No 9) (2007) 238 ALR 1, [2007] FCA 31 at [339]; and Yorta Yorta HC at [80]; (g) at [94], however "gaps in the historical timeline" in establishing the traditional laws and customs at sovereignty may be filled by inviting the Court to draw inferences where there is a proper foundation to do so, citing De Rose v State of South Australia [2002] FCA 1342 (De Rose) at [570]; (h) at [95], "such an inference is more readily drawn where there is a preponderance and weight of available evidence and in the absence of contradictory evidence". Further, "where it is appropriate to do so, weight should be accorded not only to physical use of the claim area, but also to evidence of spiritual connection", citing Yarmirr v Northern Territory (No 2) (1998) 82 FCR 533 (Yarmirr) and Griffiths v Northern Territory (2007) 165 FCR 391; [2007] FCAFC 178 (French, Branson and Sundberg JJ) at [127]; (i) at [97], the Court must be satisfied that the traditional rights and interests relate to the particular area of land and waters under consideration. Where the issue is whether connection has been maintained to a particular area under traditional laws and customs since sovereignty, it is necessary: (i) to examine the traditional laws and customs for s 223(1)(b) purposes as they relate to that area; and (ii) to demonstrate that connection to that area has, in reality, been substantially maintained since the time of sovereignty. That requires a consideration of the historical and contemporary evidence as it relates to that area to ascertain whether "the laws and customs that related to that area had continued to be acknowledged and observed without substantial interruption and that connection likewise has been substantially maintained" since sovereignty, quoting from the Full Court decision in Bodney v Bennell (2008) 167 FCR 84; [2008] FCAFC 63 (Bodney) at [178]-[179] and [187]. 47 Having set out these principles, his Honour foreshadowed, particularly with respect to those principles noted in [46(g)] and [46(h)] above, one of the central difficulties that the three appellants faced in establishing their claims. He did that by observing at [98] that, where there were three competing claimants, it could not be readily inferred "from the existence of adjoining native title rights at sovereignty that such connection, and therefore such rights, extended naturally into Lake Torrens because that inference (without more) would apply equally" to each of the other claimants. As I already noted above, his Honour reiterated this difficulty in his ultimate conclusion on all three claims ([at 774]) of the Reasons (see at [7] above). 48 With particular relevance to the connection inquiry, there is a number of further principles that are worth highlighting at this point. First, with respect to the principle noted at [46(d)] above about the approach required by the Court when considering a native title determination application, since s 223(1)(b) is drawn in the present tense, it is concerned with the current connection with the land and waters concerned (see Yorta Yorta FC at [142]) and Yorta Yorta HC at [85] per Gleeson CJ, Gummow and Hayne JJ). 49 Secondly, concerning the inference that may be drawn to fill the gaps in the history (addressed in the Reasons at [94], see at [46(g)] above), it is to be noted that the question in De Rose to which that inference was directed was the question of connection under s 223(1)(b) (see De Rose at [559]-[570]). By contrast, the following observations of the Full Court in Sampi (on behalf of the Bardi and Jawi People) v Western Australia (2010) 266 ALR 537; [2010] FCAFC 26 (Sampi) about inferences were directed to the question under s 223(1)(a) whether the claimants formed a society at sovereignty with a body of laws and customs that gave rise to rights and interests in land (Sampi at [64]): The question then arises whether the court can infer the existence of that acknowledgement and observance from about the latter part of the 19th century back to sovereignty. Selway J addressed this issue in Gumana v Northern Territory (2005) 141 FCR 457; [2005] FCA 50 and said at [201] by reference to the history of the approach of the common law to the proof of custom: [201] … where there is a clear claim of the continuous existence of a custom or tradition that has existed at least since settlement supported by creditable evidence from persons who have observed that custom or tradition and evidence of a general reputation that the custom or tradition had "always" been observed then, in the absence of evidence to the contrary, there is an inference that the tradition or custom has existed at least since the date of settlement. (Emphasis added) I interpose that, since the connection inquiry in these claims was primarily directed to s 223(1)(b), that provides a likely explanation why De Rose is referred to by his Honour in the Reasons and Sampi is not mentioned. In this respect, it is to be noted that his Honour was a member of the Full Court in Sampi. 50 Thirdly, and further to the above, the character of the evidence necessary to draw such an inference is exemplified by the observations of Olney J (albeit directed to s 223(1)(a)) about the quality of the evidence in Yarmirr (referred to in the Reasons at [95], see at [46(h)] above) at 569 as follows: By way of example, Mary Yarmirr, her brother and their sisters are able to trace their descent through the patriline for four generations. In ordinary circumstances, this would represent about 100 years prior to Mary's birth in 1946. Given the preponderance and weight of the available evidence, and the absence of any contradictory evidence, I draw the inference … 51 Fourthly, and further with respect to the principle noted at [46(h)] above about physical use and spiritual connection being available to establish the presence of Aboriginal people in an area at sovereignty, it is important to note that in Ward HC at [64] the High Court expressly left open the question whether spiritual connection alone would suffice. 52 Further still on this aspect, it is worth recording that connection may still be substantially maintained with land at a spiritual and/or cultural level notwithstanding a community's physical absence from that land (see Bodney at [172]-[174] and the authorities cited). See also De Rose v State of South Australia (2003) 133 FCR 325; [2003] FCAFC 286 at [304]-[307] and De Rose v State of South Australia (No 2) (2005) 145 FCR 290; [2005] FCAFC 110 at [62]-[64]. 53 Fifthly and finally, with respect to the principle noted at [46(i)] above where his Honour quoted from the Full Court decision in Bodney, it is important to mention the five matters recorded earlier in that judgment about the nature and content of the "connection inquiry" under s 223(1). 54 The first was that ss 223(1)(a) and (1)(b) required two distinct connection-related inquiries, the former relating to rights and interests in relation to land and waters and the latter relating to connection with that land or those waters, each sourced in the traditional laws acknowledged and customs observed by the claimants in question (see Bodney at [165]). 55 It is worth interposing that the connection inquiry issue at trial and in these appeals primarily arises under s 223(1)(b). This was identified by the primary judge in the Reasons (at [168]) as: [W]hether all of, or any part of, the rights and interests possessed by any of the three groups, extend to and include all of, or any part of, Lake Torrens, including Andamooka Island, at sovereignty and have since been maintained and exercised by the successors to the relevant group or groups to the present time, albeit appropriately adapted, so as to satisfy s 223(1)(b) requiring the present connection to Lake Torrens (or a part or parts of it) under those traditional laws and customs. 56 The second matter was that both the continued acknowledgement and observance of the traditional laws and customs in question and the connection with the land and waters under those laws and customs must be "substantially uninterrupted" and "substantially maintained" from the time of sovereignty (see Bodney at [168]). 57 The third matter was that, while there is often a significant overlap between the two: "the connection inquiry requires, first, an identification of the content of the traditional laws and customs and, secondly, the characterisation of the effect of those laws as constituting a connection of the people with the land" (Bodney at [169]). 58 The fourth matter was that the connection inquiry requires the claimants to demonstrate the "reality" of the connection to their land by their laws and customs, that is: "the connection itself must have a continuing reality to the claimants and that the evidence of how this is manifest is of no little importance in establishing present connection" (Bodney at [171]): 59 The fifth and final matter was to emphasise that the connection inquiry had a "particular topographic focus" on the claim area in question, particularly where it is inaccessible (Bodney at [175]): … in those cases where the claim area includes within its boundaries portions of land or waters for which there is no evidence of use by the claimants, or which are inaccessible. In such cases the courts have shown a distinct propensity to infer such connection as was practicable with such land or waters from Aboriginal activities in the surrounding areas which were supportive of a connection to the general area within which those apparently unused or inaccessible areas were located. It is worth adding that Bodney at [178]-[179], which his Honour quoted at [97] of the Reasons (see at [46(i)] above), concluded the discussion on this fifth matter. 60 As will appear later in these reasons, the principles outlined above had a particular resonance in the final Consideration section of the Reasons where the primary judge expressed his final dispositive reasoning with respect to each of the three appellants' claims.

The primary judge's dispositive reasoning 61 As I foreshadowed above, I will begin my consideration of the Kokatha appellants' grounds of appeal by summarising the dispositive reasoning contained in the final Consideration section of the Reasons as it applies to the Kokatha appellants. First, at [712], his Honour identified his "starting point" with respect to the Kokatha appellants' claims was the Kokatha Part A determination and the progressive occupation of the area west of Lake Torrens by the Western Desert Cultural Bloc as follows: … to accept, as the Kokatha Part A determination records, that at the time of sovereignty the Kokatha People occupied the western boundary of Lake Torrens according to their traditional laws and customs. It is consistent with the evidence of all the anthropologists that that occupation came about progressively as the Western Desert Cultural Bloc moved from the north and west towards Lake Torrens. I accept that that movement had progressed to the point of the western boundary of Lake Torrens by sovereignty, as determined by the Kokatha Part A determination. 62 Having set out his conclusion at [713] that "it is not possible, on the evidence, to be satisfied that the Kokatha People had extended their country under their traditional laws and customs into the area east of that western boundary of that determination at the time of sovereignty", his Honour proceeded to explain why he had come to that conclusion, beginning with the ethnographic and historical evidence as it applied to the Kokatha appellants' claims, as follows (at [714]-[715]): 714 There is no evidence of Kokatha occupation of any areas to the east of that western boundary of Lake Torrens at the time of first European contact, or indeed in my view in any of the ethnographic or historical material until well into the 20th century, probably about the 1980s … 715 There is, on the other hand, some evidence that the Kokatha occupation of the area immediately to the west of Lake Torrens was, if not relatively recent prior to sovereignty, at least at a point in time when the Kuyani People and to a significant degree also the Barngarla People lived in part on the area to the west of the western border of Lake Torrens … 63 At [716], his Honour confirmed his views about the credibility of the Kokatha lay witnesses, particularly the Starkey brothers (see at [38(f)(i)] above). Then, at [717]-[719], his Honour expressed what were perhaps his most critical views dispositive of the Kokatha appellants' claim. They related to the gap in the evidence between the current genuinely held knowledge and beliefs of the Starkey brothers and the traditional knowledge and beliefs of the Kokatha People at sovereignty with respect to Lake Torrens. Concerning that gap, his Honour began by describing the current knowledge and beliefs of the Starkey brothers and the stated sources of that knowledge (at [717]) as follows: In the 1980s, they were each relatively young men. Each of them relied on Max Thomas as the source of much of their current knowledge. Each of them gave coherent, and persuasive evidence as to their current state of knowledge and their current beliefs, particularly in relation to the Dreaming stories relating to Lake Torrens, and to the claim that the Kokatha People at material times from sovereignty had observed a body of laws and customs under which they possessed rights and interests in, and had a connection with, Lake Torrens. They had also received some knowledge about Lake Torrens from other Western Desert men, and possibly also from Ningel Reid. (Emphasis added) 64 Then, at [718], his Honour expressed difficulty accepting that the knowledge of the Starkey brothers' identified sources was itself sourced in the traditional knowledge and beliefs of the Kokatha People at sovereignty because of what emerged during the investigations in the early to mid 1980s concerning the Olympic Dam mine at Roxby Downs. Specifically, he said: The difficulty I have with accepting that their knowledge represents the knowledge of the Kokatha People at, and subsequent to, settlement or sovereignty in relation to Lake Torrens derives not simply from the absence of any supporting material up to the 1980s, but also from the material which was gathered directly from Ningel Reid and Max Thomas in particular in the period of the 1980s, at the time of the investigations concerning the proposed Olympic Dam mine in the early to mid 1980s. It was during that time that the [Kokatha People's Committee] came to be formed and surveys of the area surrounding Roxby Downs took place in 1980 and late 1981. (Emphasis added) 65 At [719], his Honour proceeded to describe the results of the investigations undertaken by Vachon in 1981 and by Hagen in 1983. His views about the destructive effects of the results of those investigations for the Kokatha appellants' claims are then set out in [719] as follows: An initiated Western Desert anthropologist, Daniel Vachon, was requested to assist in the identification of relevant sacred sites and the recording of the stories relating to the area, including the area to the east of Roxby Downs and what became Olympic Dam. It is fair to say that his record of those investigations, including from his informants, did not expose any Kokatha or Western Desert stories relating to Lake Torrens itself. He remarked that "mythological knowledge is worse than thin, it is non-existent in these informants. The men themselves know that others will need to be involved if the myths are to be re-learned". In November 1981, the second trip, reference is made to Wilyaru, but it was not recorded as relevant to Lake Torrens, as distinct from the area to the west of Lake Torrens. In 1983, Hagen went through the area with Kokatha men, including Max Thomas, but again the Dreaming stories relating to Lake Torrens and its occupation or possession by Kokatha People was not adverted to. (Emphasis added) 66 At [720], his Honour returned to the ethnographic materials and confirmed his view that, in combination, they put the boundary of the Kokatha People's country at the western side of Lake Torrens. 67 At [721], his Honour expressed his views with respect to the three female Kokatha appellants' lay witnesses that: I did not consider that their evidence advanced the Kokatha claim, in the sense of taking it into Lake Torrens area at all, except by reference to relatively recently acquired information. The source of that recently acquired information is, beyond one generation, speculative. Had that state of affairs existed prior to the 1980s, as a state of knowledge and belief, in my view it is highly likely that it would have been recorded at a much earlier point of time. 68 At [722]-[723], his Honour explained why he did not consider Willis' lay and expert evidence bridged the gap in the Kokatha appellants' case. First, it should be noted that his Honour accepted (at [198]) that Willis was "completely honest" in his lay evidence about Western Desert culture and practices but, because of his status as a lay witness, he would be "a little cautious" in weighing his expert evidence. Furthermore, at [202], his Honour stated that, because he had rejected Draper's analysis (see at [383]-[384]), to the extent that Willis had relied on Draper's evidence to express his opinions as an expert, he would place no weight on those aspects of his evidence. Turning, then, to the conclusions his Honour came to about Willis' evidence, first, at [722], he said Willis' lay evidence did not assist to "bridge any gap in knowledge or awareness" that existed in the Kokatha appellants' case essentially because it dated from about the mid 1980s or early 1990s. His explanation for that conclusion is worth quoting verbatim, as follows: It is evident from his own evidence that Willis, as a lay witness, did not himself know or learn of claims of Kokatha People's relationship to Lake Torrens, routinely as part of his exposure to Western Desert society. His exposure to particular Kokatha cultural knowledge runs largely from the mid 1990s, and his opportunity to review and analyse the ethnographic and historical material was (as I have noted above) somewhat restricted on his own evidence. Whilst he became aware of the Western Desert tjukurpa relating broadly to the region near Lake Torrens when working as an anthropologist at the Pitjantjatjara Council in the mid 1980s, his learning as a Western Desert man at that time was restricted. He said it was not until the early 1990s that he learned of the Wati Nyiru and Seven Sisters dreamings having some association with Lake Torrens, information which he described as the "first hint" of that association and which he learned from a Western Desert man but not a Kokatha person. His more precise knowledge came only from surveys conducted in 1996 and 1997. In my view, his knowledge of those matters as a lay witness is shown to be derivative only to informants in a material sense in the 1990s, and does not bridge any gap in knowledge or awareness beyond the period of the mid 1980s to which I have referred. 69 Secondly, at [723], his Honour concluded that Willis' expert evidence did not advance the Kokatha appellants' claim to Lake Torrens. He provided two reasons for that conclusion. First: As an expert witness, the opinions he expressed were tentative and relatively unpersuasive (to the extent that they progressed the Kokatha claim area into Lake Torrens). He pointed out that the ethnographic record was capable of supporting Kokatha occupation of the Territory to the west of Lake Torrens (as determined in Kokatha Part A). He pointed out the inaccuracy or unreliability of early ethnographic recording of Aboriginal group country. That may be correct, but it does not advance the Kokatha People's claim over Lake Torrens or provide a reason why, as an anthropologist, the step should be taken of interpreting the data in a way which attributes Kokatha country as extending over Lake Torrens. And secondly: he pointed out the fact of major Dreaming stories or tjukurpa traversing the Western Desert area both running north and south and to the south-west and north-east. Again, that did not in my view support any anthropological opinion attributing Lake Torrens itself to the Kokatha People at sovereignty, or indeed for a considerable time thereafter. He referred to the archaeological investigations, about which I have made separate findings. (Emphasis added) 70 At [724], his Honour summed up the overall effect of the views he had expressed between his starting point at [712] and his dismissal of Willis' lay and expert evidence at [722]-[723] as follows: In my view, the absence of any material at all which tends to associate Kokatha interests with Lake Torrens until the relatively recent past, particularly in circumstances where it may have been expected that such Kokatha Dreaming stories as relate to Lake Torrens (according to the contemporary evidence) would have emerged in the course of those investigations particularly in the 1980s, and having regard to the fact that the maps drawn by others, including Tindale, in the earlier decades of the 20th century draw the Kokatha boundary on the western shore of Lake Torrens lead to the conclusion that, at sovereignty, the Kokatha People by their traditional laws and customs did not possess or occupy any part of Lake Torrens east of the western shore of Lake Torrens (as determined in the Kokatha Part A determination). 71 Before coming to his ultimate dispositive conclusion on the Kokatha appellants' claims at [726], his Honour expressed the view, at least implicitly, that the Kokatha appellants' lay evidence about Kokatha stories relating to Lake Torrens was contemporary in origin and not traditional, as follows (at [725]): I have referred above to the Kokatha stories relating in some way to Lake Torrens, including Andamooka Island. In essence, that evidence came from the three Starkey brothers, other than some evidence of the Seven Sisters tjukurpa which the three female witnesses spoke of. 72 Finally, at [726], his Honour's final dispositive conclusion on the Kokatha appellants' claims was set out as follows: On the whole of the evidence, I accept the submission of the State that those tjukurpas relied upon by the Kokatha claimant must be post-sovereignty in origin. As I have said, I have no doubt they are sincerely believed by the witnesses concerned, but their exposure to information came from those involved in the 1980s when the proposed Olympic Dam mine was being considered and those sources of information did not at that time present that material when it might otherwise have been expected to have been presented. 73 From this review of the dispositive reasoning in the final Consideration section of the Reasons, it can be seen that the Kokatha appellants failed to establish that their current knowledge and beliefs, genuinely held as they were, were sourced in the traditional laws and customs of the Kokatha People such that they could claim to have native title rights and interests in Lake Torrens. As the State pointed out in its written submissions, correctly in my view, the deficiency in the Kokatha appellants' case on connection was that their lay evidence did not "take Kokatha occupation or connection to [Lake Torrens] anywhere near sovereignty". This was so despite the Kokatha appellants' lay witnesses, particularly the Starkey brothers, being accepted by the primary judge as honest witnesses as to their underlying current state of knowledge about the Kokatha connections to the claim area. Furthermore, inference could not be used to fill the gap that existed between their current state of knowledge and the traditional laws and customs of the Kokatha People at sovereignty in circumstances where there were multiple competing claims to the same claim area. 74 Having set out the primary judge's dispositive reasoning with respect to the Kokatha appellants' claim, I now turn to consider their grounds of appeal.

Ground of appeal 1 - not drawing an inference 75 Ground of appeal 1 of the Kokatha appellants' appeal (see at [8] above) expresses the same high level error in the positive and in the negative. That is, it asserts that the primary judge erred because he came to the conclusion that the Kokatha appellants did not possess native title rights and interests in any part of the claim area and because he did not come to the conclusion that they did possess such rights and interests in that area. Four paragraphs of the Reasons are said to evidence this error: [713], [724], [726] and [771]. 76 With the exception of [771], each of these paragraphs appears in the final Consideration section, the contents of which have been reviewed immediately above (see at [61]-[72]). As can be seen from that review, those three paragraphs essentially sum up his Honour's reasoning for rejecting the Kokatha appellants' claims. Paragraph [771] (see at [4] above) appears in the final Conclusion section of the Reasons. It is similar in that there, his Honour provided a further one paragraph summing up of the reasons why he was not persuaded that a determination of native title should be made in favour of the Kokatha appellants. 77 Eleven particulars were provided in support of this ground of appeal. Those particulars variously refer to and rely upon approximately 40 paragraphs of the Reasons predominantly selected from the sections where his Honour reviewed the evidence of the Kokatha appellants' lay witnesses. Those paragraphs are said in the submissions to be the "findings and other facts" which are relied upon under this ground of appeal. 78 The only error specifically identified in those particulars is that set out in Particular (ix) (see at [8] above). As that Particular records, that error is the subject of ground of appeal 2. For the reasons I have given below with respect to that ground, I do not consider any such error exists. I should add that, while there is no mention of them in this ground of appeal, or its supporting particulars, in their written submissions the Kokatha appellants have also pointed to the errors alleged to exist in the factual findings raised by grounds of appeal 2, 2(a), 4, 5 and 6, as supporting this ground. For the reasons I have also given below with respect to those grounds of appeal, I do not consider any of them raises an appealable error. 79 The main contention advanced by the Kokatha appellants in their submissions (although, as the State pointed out in its submissions, this ground of appeal itself does not mention it) is that, in the circumstances where the primary judge had accepted the honesty of their lay witnesses, particularly the Starkey brothers, about their spiritual beliefs and practices, his Honour was wrong not to have drawn an inference in their favour as described in Sampi at [64]-[65]. In support of this contention, they pointed to a number of matters which they claimed were undisputed facts or facts established by his Honour's findings and contended that this Court was in as good a position as the primary judge to decide on the proper inference to be drawn from those facts. Their attempt to dismiss the primary judge's advantage is already recorded above (at [12]). 80 It is apt to begin the consideration of these contentions by identifying the circumstances in which an inference can, or cannot, properly be drawn. In the first place, an inference cannot properly be drawn unless there are "positive proved facts from which [it] can be made" (see Caswell v Powell Duffryn Associated Collieries, Limited [1940] AC 152 at 169-170; Seltsam Pty Ltd v McGuiness (2000) 49 NSWLR 262; [2000] NSWCA 29 at [80]-[91] per Spigelman CJ and Rawson Finances Pty Ltd v Federal Commissioner of Taxation (2013) 93 ATR 775; [2013] FCAFC 26 at [88]). This is undoubtedly what the primary judge had in mind when he said at [94] of the Reasons (see at [46(g)] above) that an inference could be drawn where there was a "proper foundation" to do so. Further, where the evidence gives rise to conflicting inferences, an inference cannot be properly drawn unless it "might reasonably be considered to have some greater degree of likelihood". Or conversely, it cannot be properly drawn if those conflicting inferences are "of equal degree of probability" (see Holloway v McFeeters (1956) 94 CLR 470 at 480-481). 81 For the reasons that follow, I consider the Kokatha appellants fail on both of these criteria. First, on the need for a proper factual foundation, they cannot point to a sufficient body of "positive proved facts" from which an inference can be drawn that they hold native title rights and interests over Lake Torrens. Secondly, in the circumstances of these three competing claims, with each claimant advancing a similar conflicting inference of approximately equal probability, no inference can be drawn in favour of any one of them. 82 There are essentially three groups of facts which the Kokatha appellants have sought to advance as "positive proved facts". In no particular order, they are the facts the subject of the alleged errors in the factual findings made by the primary judge, certain undisputed facts and certain facts which they claim his Honour found in the Reasons. As to the first group, for the reasons given at [78] above, I do not consider any of those errors exist. The Kokatha appellants cannot therefore resort to any of those facts as a proper factual foundation for the drawing of an inference. 83 The second and third groups can be considered together. Those groups are variously set out in the particulars of this ground, or in the supporting submissions. In my view, none of those facts could truly be said to be positively established facts. Particular (i) is an assertion or submission. On its face, it is not a fact, undisputed or otherwise. Insofar as it is an assertion or submission, I will consider the proximity factor to which it refers when considering particular (ii) below. The remainder of the particulars rely upon specific statements contained in the Reasons without having regard to countering statements made elsewhere within the same paragraph of the Reasons, or in the paragraphs appearing before and after that paragraph, or without regard to the import of the statement relied upon when it is read in the context of the Reasons as a whole. Specifically, Particular (ii), which relies on the agreed position of the anthropologists recorded in [207] of the Reasons, fails to acknowledge the circumstance I have mentioned above that, because of the effect of the consent determinations, each of the other two appellants was in the same position as the Kokatha appellants in having traditional rights and interests in country in close proximity to Lake Torrens and each of them made similar competing claims before his Honour to an inference relying upon that physical proximity. Unsurprisingly, his Honour concluded that he was unable to draw such an inference in favour of any one of the three appellants (at [709]: set out at [41] above). 84 Particulars (iii) and (iv) rely upon statements appearing in his Honour's consideration of the archaeological evidence at [382] and [384] of the Reasons, including Draper's evidence, who, as noted above, was called by the Kokatha appellants. However, those particulars fail to have regard to the countering evidence of Carver called by the Adnyamathanha appellants nor, most importantly, to his Honour's conclusion that Carver's evidence should be accepted ahead of Draper's (see the Reasons at [383]-[385]). The remainder of the particulars from (v)-(xi) (excluding Particular (ix): see above) rely upon his Honour's acceptance of the honesty and truthfulness of the lay evidence called by the Kokatha appellants, particularly that of the Starkey brothers, about the state of their "current" knowledge without having any regard to the reasoning in the final Consideration section where his Honour explained in some detail why he rejected their claims, notwithstanding those findings (Reasons at [712]-[726], reviewed at [61]-[73] above). On this aspect, it is important to note that each of the appellants could rely upon similar findings in their favour, again preventing the probabilities necessary for an inference being tipped in favour of any one of them. The Kokatha appellants cannot therefore point to a body of undisputed facts, or facts which have been positively established by findings made by the primary judge that are sufficient to support the inference they would have this Court draw. 85 Even if such a body of undisputed or established facts existed, the Kokatha appellants have not explained why that inference is so compelling that the primary judge was wrong in not drawing it. Earlier in these reasons I set out, by reference to The Law section of the Reasons, the principles that apply to drawing inferences in native title claims such as the present. They include the need for the claimant seeking such an inference to provide credible evidence in support of the inference and the need for there to be an absence of evidence to the contrary (see at [46(g)], [46(h)], [49] and [50] above). As well, in my outline of the primary judge's Reasons, I summarised the important sub-section appearing at the beginning of the final Consideration section where his Honour explained why he was not willing to draw an inference in favour of any of the present appellants. It can be seen from that outline that his Honour did not consider the matters of physical or temporal proximity advanced by all of the appellants supported such an inference (see at [40] above), particularly having regard to the exigencies of the three competing claims (see at [41] above) and the significant conflicts in the evidence of the lay witnesses of the three appellants (see at [42]- [43] above). To all these impediments to the drawing of an inference in favour of any one of the appellants may be added the three other matters that were also highlighted in The Law section of the Reasons: that each appellant bore the ultimate onus to prove their claim (see [46(f)] above); that they each had to establish "the reality" of their connection under s 223(1) (see at [58] above); and that the connection inquiry in these three claims was necessarily topographically focused on Lake Torrens (see [59] above). 86 Expressed in the terms of the judgments in De Rose and Sampi (see at [49] above), all these matters meant that each of the appellants had considerable difficulties assembling credible evidence in support of their claims to such an inference, or avoiding "evidence to the contrary". All three conflicting inferences were therefore of an equal degree of probability and none of them could reasonably be considered to have some greater degree of likelihood. 87 Nothing has been advanced by the Kokatha appellants to show why his Honour was wrong in any of this reasoning. That is, no attempt has been made to show that, notwithstanding this reasoning, the inference they sought to have drawn was so well supported by positive proved facts, and so compelling as to its probabilities that his Honour should have drawn it. 88 For all these reasons, I do not consider there is any merit in the Kokatha appellants' ground of appeal 1.

Grounds of appeal 2 and 3 - the very significant objects 89 The Kokatha appellants' ground of appeal 2 concerns "the very significant objects" shown to the primary judge during the restricted men's evidence session on Andamooka Island. They claim that his Honour erred in finding that none of those objects was "specific to either the Lake generally or to Andamooka Island" and that his Honour should have made findings to the contrary. Ground of appeal 3 also partly relies upon this alleged error to assert that "the trial judge did not have regard to and gave no weight to important evidence of connection". Accordingly, I will deal with these two grounds of appeal together. The two paragraphs of the Reasons where these errors are said to be present are [410] and [608] (sic - [688]). The former appears in that part of the Reasons where his Honour is recording the evidence given by Andrew Starkey as follows: On Andamooka Island, he explained the significance of Andamooka Island generally, including particular physical features and during a view he showed certain stone tools and certain other very significant objects (not permanently there). 90 The latter appears in that part of the Reasons where his Honour is describing the Kokatha appellants' evidence concerning ceremony, ritual practice and sacred objects. Related to those topics, his Honour described the evidence of Andrew Starkey in the following terms: Andrew Starkey gave evidence of the importance of Andamooka Island and in particular the area marked as Crombie Ridge where he provided an explanation as to why the place is only visited by senior men and is significant. As noted, the Court was taken on a walk and shown some objects, the significance of which was explained in the restricted evidence of Andrew Starkey. They were not specific to either the Lake generally or to Andamooka Island. He also provided detail of ceremonies at Crombie Ridge in his restricted evidence. He also discussed the importance of secrecy and privacy in relation to undertaking ceremonies. The objects viewed included objects which were the law that gives the Kokatha the authority to talk for Kokatha country generally. He provided an example of the use of a certain object in ceremony and ritual practice. There was some detail of ceremonies (in his restricted evidence) at or on, or in the vicinity of, Lake Torrens itself. 91 The object of the Kokatha appellants' concerns is the third sentence of [688] and the words "specific to either the Lake generally or to Andamooka Island". In my view, no error is apparent in [410]. Without disclosing what Andrew Starkey said in the restricted evidence session on this topic, it appears from the restricted transcript of that evidence that Mr Starkey did refer to at least one of those sacred objects as having significance for Lake Torrens itself. There was also restricted evidence on the same topic given by Michael Starkey and Robert Starkey. However, in my view, that evidence is broadly consistent with the observations his Honour has made in the final two sentences of [688] above. As well, there are observations elsewhere in the Reasons where his Honour accepts as honest and truthful the evidence of the Starkey brothers and states that they gave evidence about their "current" knowledge and belief which is specific to Lake Torrens (see, for example, [405], [716] and [717] set out at [38(f)(i)] and [63] above, [724] set out at [70] above and [725] set out at [71] above). 92 While it is difficult to determine what his Honour meant by the words "specific to either the Lake generally or to Andamooka Island", given all the unambiguous statements I have referred to above and having regard to the advantage he possessed, I am not willing to conclude the former statement is reflective of appealable error on the part of the primary judge.

Grounds of appeal 2A and 3 - the Kokatha Tjukurpa and Crombie Ridge 93 Ground of appeal 2A concerns the "Angarta, Urumbulla and Wanampi (sic - Wanambi) stories" with respect to which the primary judge is alleged to have erred at [609] of the Reasons. According to the Kokatha appellants, the primary judge stated in that paragraph that the stories in question "were mentioned only in passing by Michael Starkey and there was no evidence given describing how those stories related to the claim area". For the same reasons as I have given above, I will deal with this ground of appeal and ground of appeal 3 (to the extent it relies upon this ground of appeal) together. 94 In their submissions, the Kokatha appellants referred to the restricted evidence that Andrew Starkey and Robert Starkey gave at Crombie Ridge on Andamooka Island about these Tjukurpa and their connection to the claim area, and to "secret men's rituals conducted on the claim area". In response to the State's submissions that this evidence was brief, they contended that that was so because it related to particularly "secret and dangerous men's law". They claimed that the primary judge did not consider any of this evidence. They also contended that the primary judge erred by placing Crombie Ridge off the claim area and to the west of Lake Torrens. They claimed that this appeared from [454], [534] and [593] of the Reasons. 95 Paragraph [609] appears in a section of the Reasons which is headed Summary of Dreaming Stories and Beliefs and Activities. That section follows immediately after his Honour's review of the lay evidence of the three appellants ([401]-[602]). In it, the primary judge "briefly but thematically" summarised the main Dreaming stories and activities exposed by the lay evidence of the three appellants. The paragraph in contention ([609]) appears at the end of the first part of the section where his Honour described the Kokatha Tjukurpa, or Dreaming stories and beliefs. His Honour began that part (at [605]-[607]) by describing the evidence given by Andrew Starkey with respect to the Kokatha appellants' Wati Nyiru and Willaroo (Wilyaru) stories and the fact that those stories were "highly sacred and dangerous men's stories". He concluded that description (at [607]) by stating "Consequently, much of the [Kokatha appellants'] evidence on this topic was provided to the Court as restricted evidence by the Starkey brothers." 96 At [608], his Honour mentioned some other stories, including the three stories in contention in this ground of appeal in the following terms: Other stories to a degree associated with Lake Torrens were the Seven Sisters Tjukurpa, told by Kokatha women, and references which were made to the Kalta, Angkarta, Urumbulla and Wanambi. 97 Finally, at [609]-[610], he concluded as follows: 609 The existence of the Angkarta, Urumbulla and Wanambi stories were mentioned only in passing by Michael Starkey and there was no evidence given describing how those stories related to the claim area. 610 Much of the evidence given about the Tjukurpa was given in closed session, so brief "open" descriptions only are given. 98 Thereafter, his Honour proceeded to summarise the evidence concerning the Wati Nyiru story (at [611]-[614]), the Wati Willaroo (Wilyaru) story (at [615]) and the Seven Sisters story (at [616]-[619]). 99 The Kokatha appellants' lay evidence which his Honour summarised in these paragraphs is described earlier in the Reasons at [409]-[456]. In the introductory paragraphs to that description (at [401]-[408]), his Honour made two important observations for present purposes, both of which are already mentioned above. First, at [405], his Honour said that he accepted the Starkey brothers as honest witnesses. He then reiterated this conclusion at [716] in terms that made it clear that he was referring to their current knowledge and beliefs about Dreaming stories and related matters. 100 Secondly, at [408], his Honour emphasised the balance he needed to maintain between preserving the secrecy and confidentiality of the restricted evidence given at the trial and conveying sufficient information about that evidence to explain his reasons for reaching the conclusions he did. This paragraph has already been set out above, but it is so important for this ground of appeal it bears repeating: As it is desirable that the reasons for judgment of the Court should, so far as possible, be publicly available, these reasons touch only generally on the detailed elements of confidential evidence. That should not be taken to indicate that its detailed content has not been carefully and fully considered. There is, however, little point in appending a confidential and restricted summary of that evidence where a redacted version is available, and those who may have access to the confidential transcript may do so in any event. That will explain why, in these reasons, the description of the evidence (for instance) of the three Starkey brothers is relatively brief and does not go into detail of particular stories or sites. (Emphasis added) The part of this paragraph emphasised above makes it clear that, notwithstanding his failure to mention the details of any restricted evidence, his Honour had "carefully and fully considered" it. Furthermore, in the concluding section of this paragraph, his Honour made a specific reference to the evidence of the three Starkey brothers. 101 In the paragraphs of the Reasons immediately following these observations, his Honour proceeded to record the evidence of the three Starkey brothers: Andrew Starkey, Robert Starkey and Michael Starkey, in that order. With respect to the evidence of Andrew Starkey, there are references in the Reasons to stories and to the Kokatha Tjukurpa, including Wati Nyiru and Wati Willaroo (Wilyaru) at [411]-[413]). At [412], it is noted that "the details [of that evidence] were given in a closed session". Similarly with respect to the evidence of Robert Starkey, there are such references at [419] and [421]-[425]. As well, it should be noted that [421] contains the same notation as at [412] described above. Finally, the evidence of Michael Starkey contains similar references at [429]-[430] and [432], including a similar notation to that at [412] at [430]. In particular, at [430], his Honour noted that Michael Starkey's evidence concerning the three stories mentioned in [609] were "stories associated with Lake Torrens". 102 When [609] is read in the context of all the paragraphs of the Reasons set out above, I do not consider his Honour should be interpreted as saying that Michael Starkey's evidence on the three stories mentioned in [609] was the "only" evidence on that topic, or that there was no evidence describing how those stories related to the claim area. The transcript of the restricted evidence shows that the topic was addressed by the Starkey brothers, particularly Andrew Starkey and Robert Starkey, with differing levels of detail, albeit that evidence was predominantly directed to the Wati Nyiru and Wati Willaroo (Wilyaru) Tjukurpa. It is clear from a fair reading of the paragraphs of the Reasons referred to above that his Honour was well aware of that evidence. Any failure to mention it, or particular aspects of it, in more detail is, in my view, most likely explained by the confidentiality restrictions his Honour was being particularly careful to maintain. Furthermore, and relatedly, having regard to the specific assurance his Honour gave at [408] (see at [100] above), which was pointedly directed to the evidence of the Starkey brothers, I reject the Kokatha appellants' contention that he did not consider all of the evidence they gave on that topic. I do not therefore consider [609] is reflective of any relevant error on the part of the primary judge. 103 Finally, I reject the Kokatha appellants' claim that his Honour thought that Crombie Ridge was off the claim area to the west of Lake Torrens. The apocryphal nature of this proposition underscores how cautious this Court needs to be when considering arguments advanced by highly selective quotations from a complex and lengthy set of reasons, unassisted by the great advantage the primary judge possessed. This is so because when all the references to Crombie Ridge in the Reasons are duly considered, it becomes apparent that, insofar as the Kokatha lay evidence was concerned, his Honour well understood that the Crombie Ridge being referred to by them was the senior men's site which was located on Andamooka Island. It is also apparent, unsurprisingly, that his Honour well understood that the Crombie Ridge to which the two Starkey brothers (Andrew Starkey and Robert Starkey) were referring when they gave restricted evidence on Andamooka Island was the Crombie Ridge on that Island on which that evidence was being given. I interpose to note that the third Starkey brother, Michael Starkey, gave his evidence at Port Augusta (see at [427] of the Reasons). 104 The Crombie Ridge on Andamooka Island is first mentioned at [372(8)] of the Reasons where his Honour listed the sites where the archaeologists had agreed there was evidence of Aboriginal people's presence on the claim area. Paragraph [372(8)] states: It was agreed as between the archaeologists that there was archaeological evidence of Aboriginal people's presence at the following sites: … (8) Crombie Ridge (Andamooka Island); and … 105 Andamooka Island is then identified as the location at which Andrew Starkey and Robert Starkey gave their restricted evidence: at [409]-[410] (Andrew Starkey) and [417]-[418] (Robert Starkey). Finally, the restricted evidence of Andrew Starkey at Crombie Ridge on Andamooka Island is described in greater detail (to the extent that it could be recorded openly) at [688] (see at [90] above). 106 From these references to Crombie Ridge in the paragraphs describing the lay evidence of the Kokatha appellants, particularly the evidence of the Starkey brothers, it is clear, in my view, that his Honour well understood that the restricted evidence that two of the Starkey brothers gave at Crombie Ridge on Andamooka Island concerned the senior men's site on Andamooka Island known as Crombie Ridge. If it were necessary to further confirm that, such appears at [615] where, in describing the Kokatha Wati Willaroo (Wilyaru) Tjukurpa, his Honour said: This is a deeper story, with a relationship to the Wati Nyiru story, although they do not touch. It involves a description of what Lake Torrens represents and of a particular ceremony performed at Crombie Ridge. 107 Conversely, when one reads the paragraphs to which the Kokatha appellants have selectively referred, together with all the paragraphs containing references to Crombie Ridge, including those mentioned above, it becomes quite apparent that the references to Crombie Ridge or Crombie Sandhills at [454], [534] and [593] concern a women's site described by some of the Barngarla appellants' lay witnesses. It is that site which the two Starkey brothers said in their evidence was located somewhere to the west of Lake Torrens. In context, it is clear that they were not saying that the Crombie Ridge to which they referred in their evidence was located in that direction. 108 The women's site in question is first mentioned at [454] where his Honour refers to a statement of Eileen Wingfield tendered by the Barngarla appellants which "included reference to a special place for women called Crombie Ridge (which is marked on Exhibit S1 as being on Andamooka Island, but was said by the Starkey brothers as being to the west of Lake Torrens)." The reason why this statement is referred to in the section of the Reasons where his Honour is describing the Kokatha appellants' lay evidence appears in the concluding sentence to [454] where his Honour records that none of the Kokatha appellants' women lay witnesses "… had been told of that place or the significance of it. They did not recognise its name." 109 His Honour returned to Eileen Wingfield's statement, and the "apparent dispute about the location of Crombie Ridge" enlivened by it, at [591]-[593] (the latter of which is relied upon by the Kokatha appellants in this ground of appeal). Those paragraphs are in the section of the Reasons where his Honour is describing the lay evidence adduced by the Barngarla appellants. At [593], he made the following observations about Crombie Ridge: She agrees women are not allowed on Andamooka Island at all, and only wati can do so. She refers to the Seven Sisters tjukurpa or dreaming, including that the women were chased to Lake Torrens, where they sat and talked on the shore. That site is, by inference, identified as Crombie Ridge which is a restricted women's site and with restricted women's stories. As noted, there is an apparent dispute about the location of Crombie Ridge. She says the whole of Lake Torrens is significant to the Seven Sisters story, including a "big site" at a place around the Lake. 110 It is apparent, therefore, that the dispute referred to above was that first introduced in the concluding sentences of [454] earlier developed in the description of the evidence given by the lay witnesses called by the Barngarla appellants. The next reference to the issue occurred at [534] (the second of the three paragraphs relied upon by the Kokatha appellants in this ground of appeal) where his Honour described the evidence of Eric Paige, a lay witness called by the Barngarla appellants. Eric Paige gave his evidence in two stages: first, on Andamooka Island and then later at Port Augusta. The evidence he gave on this topic at Andamooka Island is recorded at [533]-[534]. In relation to Crombie Ridge, (at [534]): He asserted cultural knowledge of two particular sites on Andamooka Island: Crombie Ridge (as marked on Exhibit S1) related to a Kalta-Widjeru story, and Crombie Sandhills (not marked on that map). I note that the Kokatha evidence referred to above also identified a site called Crombie Ridge, but placed it off Andamooka Island to the west. The former was told to him by Bill Lennon on one of his visits. The latter was told to him by an old lady Mrs Crombie; she said it was a women's place only so men should not go there. He also said this had come from senior APY women. But for its content, having regard to the evidence of the Starkey brothers, that could support the Kokatha claim, but it is not consistent with their evidence if the site is on the Island, as the Kokatha evidence is that Andamooka Island is a men only area. (Emphasis added) 111 Then, at [540] (a paragraph not relied upon by the Kokatha appellants in this ground of appeal), he is recorded as giving the following evidence at Port Augusta about the two visits he made to Andamooka Island - one in 2005 and one in 2007: He explained that his first visit in 2005 was a site clearance visit. He agreed that the women present (including both some Western Desert women and Lorraine Dare and Eileen Crombie) would not fly over Andamooka Island by helicopter for fear of seeing certain men only sites. Eileen Crombie had gone on to the Island by vehicle and had spoken about the Crombie Sandhills. He agreed that on the second visit, specified in the questions as November 2007, the women present (including Linda, Lorraine and Patricia Dare) had simply refused to go on to the Island at all. The reason for their change of attitude is not explored (and of course is probably not a matter Eric Paige could give evidence about). 112 Finally, at [589] (another paragraph not relied upon by the Kokatha appellants in this ground of appeal), his Honour records the evidence of Harry Dare, another lay witness called by the Barngarla appellants as follows: He has been to Andamooka Island three or four times, including on one occasion with some women, but they were required to stay behind at the homestead. His description of that occasion identifies it as the second visit by Eric Paige (and the second visit by Rosalie Richards). What he said about that suggests that the women did not go on to the Island because the purpose of the visit included some men only areas, rather than that no women could go on to the Island at all. In fact, he said expressly that women were not prohibited from going on to Andamooka Island. He referred to another occasion (which by the description of those present included Eileen Crombie) was the first visit by Eric Paige when women did go on to Andamooka Island. 113 This review of all of the references to Crombie Ridge in the Reasons reveals that two sites bearing the name Crombie Ridge were referred to in the evidence at the trial. The Kokatha appellants' lay witnesses, particularly the two Starkey brothers, gave evidence at, and described the significance of, a very important senior men's site called Crombie Ridge on Andamooka Island. As well, some of the Kokatha appellants' lay witnesses and various Barngarla appellants' witnesses described a women's site called Crombie Ridge or Crombie Sandhills. It was the location of that women's site to which the dispute about the location of "Crombie Ridge" was directed. No dispute was raised with respect to the location of the men's site called "Crombie Ridge" on Andamooka Island. 114 For these reasons, I do not consider ground of appeal 2A and, to the extent that it relies upon this ground, ground of appeal 3 have any merit.

Ground of appeal 4 - the 1980s ethnographic surveys

An overview of the alleged errors 115 In ground of appeal 4, the Kokatha appellants seek to attack the primary judge's conclusions at [341] (see at [38(c)] above) and [724] (see at [70] above) about the absence of any support for their claim with respect to the claim area in the ethnographic surveys "until the relatively recent past". They also claim that the primary judge erred by failing to consider whether there may have been cultural reasons why particular information about Kokatha interests in Lake Torrens did not emerge from some of the ethnographic surveys carried out in the 1980s. In support of this ground of appeal, they provided two sets of particulars (see at [8] above). The first set lists four items which they claim "associates Kokatha interests with Lake Torrens". The second set lists six reasons why "particular information did not emerge in the 1980s". 116 In their submissions, the Kokatha appellants pointed to the statement at [724] about the absence of "any material at all" associating Kokatha interests with Lake Torrens and contrasted that with the ethnographic surveys listed in the four items in the particulars which, they claim, recorded such an association. Further, they claimed that those surveys were either not considered at all by the primary judge, or were disregarded when his Honour made this statement at [724]. These disparities, they submitted, indicated that the primary judge had not taken account of all of the important ethnographic evidence when he came to the conclusion at [724]. They also pointed to the lay evidence of the three Starkey brothers where they said they were told about the Wati Nyiru connection to Lake Torrens by their uncle, Max Thomas. In this respect, they also referred to the evidence of Peter Mungkari, a senior Tjilpi, who had also said he gained his knowledge from Max Thomas. Further, to explain the sparsity of reference to the Kokatha association with Lake Torrens in the ethnographic surveys, they also contended that none of those surveys focused on the claim area itself and they sought to highlight the "great secrecy and danger" that attach to the men's Tjukurpa as appeared from the evidence. They submitted that evidence included the Hagen reports of August and September 1983. Finally, as with ground of appeal 1 above, they sought to rely upon inferences to fill the gap between the evidence given by their lay witnesses and the traditional knowledge of earlier generations of Kokatha and Western Desert people. 117 The State responded in its submissions by listing 12 examples where the ethnographic materials dating from the 1960s and early 1970s had failed to mention any connection between Kokatha interests and Lake Torrens. It claimed that all of these materials had been adverted to by the primary judge in his Reasons and his conclusions, based upon that assessment, did not evidence any error. The State also provided two examples dating from 1983 and 1984 where there had been some mention of Kokatha myths connected to Lake Torrens. They were Hagen's August 1983 report, which referred to a Lake Torrens creation myth that was adverted to by the primary judge; and the Fitzpatrick and Gara October 1984 report, which referred to a Kokatha myth travelling north to Lake Torrens that, it accepted, was not recorded in the Reasons. Nonetheless, it submitted that these examples did not provide any basis for upsetting the primary judge's conclusions based on his assessment of the ethnographic evidence. Finally on the secrecy explanation advanced by the Kokatha appellants, the State submitted that the primary judge had, in fact, heard evidence in closed session about many of the highly secret stories and cultural practices to which the Kokatha referred. Further, it submitted that secrecy did not, in any event, explain the contradictions between the Kokatha lay evidence and the Kokatha People's Committee's report and the Vachon report, both produced in 1981. Lastly, it submitted that, even if the absence of evidence were explained, that did not provide positive evidence in support of the Kokatha appellants' case, as the primary judge himself had observed with respect to the linguistic evidence at [355]. 118 Since the Adnyamathanha appellants tendered the 1989 Gara report, it is appropriate to record their submissions concerning the 1984 Fitzpatrick and Gara report. First, they submitted that report did not record any Kokatha mythology relating to the claim area itself, but instead it referred generally to the story travelling from Port Augusta north "to" Lake Torrens. Secondly, they submitted that the 1989 Gara report, which, they claim, was "far more significant" did not mention this mythology or, for that matter, the 1984 Fitzpatrick and Gara report itself. Furthermore, they submitted the Kokatha appellants had not sought to raise this discrepancy with Gara in cross-examination notwithstanding that they were given an opportunity to do so. Finally, they submitted that secrecy did not explain the sparsity of references to a Kokatha connection to Lake Torrens when one had regard to the fact that the existence of the relevant Dreamings have all been disclosed in ethnographic materials dating from the early 1940s, albeit without the secret details. 119 It is convenient to begin the consideration of these submissions with the four items that the Kokatha appellants have particularised as materials that were either not considered at all by the primary judge, or disregarded by him in his conclusions at [341] and [724].

The Vachon 1981 report 120 Whilst it does not expressly mention it, Particular 4(a)(i) (see at [8] above) refers to the report Dr Vachon prepared at the request of the Kokatha People's Committee (KPC) in late 1981. Part 1.0 of that report relates to the "activities of Roxby Mining Services on its mining lease on Roxby Downs Station". It appears that the KPC were particularly concerned to determine whether those activities would result in the destruction of sites of cultural significance to them. As a result of those concerns, Dr Vachon was appointed to assist them to record the sacred sites on Roxby Downs. To that end, Dr Vachon conducted a field trip to the area between 8 and 10 October 1981. Following that trip, he prepared a field report and a report to the KPC. 121 Curiously, the report to the KPC is relied upon by both the Kokatha appellants and the State in their submissions to advance two diametrically opposed positions. Obviously both cannot be correct. On the one hand, as I have already mentioned above, the Kokatha appellants have included certain contents of that report (without expressly identifying the report as such) in its list of four items which they claim "associate[s] Kokatha interests with Lake Torrens". On the other hand, the State has listed the report among the 12 examples of ethnographic materials that do not reveal Kokatha interests extending to the claim area, that is, farther east than the western side of Lake Torrens. To compound this unusual contest about the effect of this report, both parties have referred to the identical part of its contents to advance their stated position. That part is Part 2(b), which states: (b) Wilyaru: It is still unclear what this term refers to, but the men said during the trip that it is a name. This track apparently began at Miringil ("Iron Knob"), north along the west side of Lake Torrens to Andamooka. No other information. It is known in the North-West, however, that a track referred to as Wilyaru goes north from Coober Pedy to a placed called malu piti near Mabel Creek and then on to yulpala, south-east of Granite Downs Homestead and then east to Lake Torrens. It is unknown whether two separate myths are involved. 122 Further still, both parties have also relied upon Dr Vachon's 1981 field report, the Kokatha appellants generally and the State with respect to the following two parts: (a) at p 11, where the Wilyaru track referred to above is described as extending east to the "western shore of Lake Torrens"; and (b) at p 25, where the same track is extending east "down the western side of Lake Torrens". 123 Before I address this contest about these two reports, it is worth setting out what the primary judge had to say about them in the Reasons. At [289]-[291], his Honour said: 289 In October 1981, Dr D Vachon produced a report for the KPC in response to the KPCs concerns that the activities of Roxby Mining Services on its mining lease on Roxby Downs Station would result in the destruction of sites of cultural significance to the Kokatha people. Dr Vachon conducted a field survey with information from Max Thomas, Jeffery Brown and Richard Reid. 290 He recorded from them the extent of Kokatha country as being to the western shore of Lake Torrens. 291 He recorded several archaeological sites and 14 ethnographic sites in the Olympic Dam Area as having possible significance, none which were within the present claim area. He noted that none of these sites had been visited for cultural purposes for at least 20-25 years, and in summary under the heading "Assessment" he said: The men who made up the field party were selected by the men themselves as being most knowledgable (sic - knowledgeable) of the study area. In my opinion, unless other people can accompany us on a second trip, we will be forced to indicate to RMS archaeological only - places the KPC would want no mining on the basis of protecting a forgotten heritage. Mythological knowledge is worse than thin, it is non-existent among these informants. The men themselves know that others will need to be involved if the myths are to be re-learned. Unlike these people who still retain their religious traditions, these men… (1) Sang no songs during the trip - a constant practice in the North-West, for instance. They admit they do not know them. (2) Know no Aboriginal place names. (3) … (4) identify wilyaru as the only "dreaming". The term seems to be used in a generic fashion for any site. Since the songs and place names are not known, there are no "ethnographic sites" to protect. Without the place names, it becomes almost impossible (but not entirely) to elicit information on the area from others. That is not to say that no-one knows the country. But unless a process of re-learning begins immediately, whatever knowledge of the area remains will be lost within the next few years. These men know that the situation is desperate, but have little confidence that the younger people share their enthusiasm. In truth, only Richard Ried (sic, Reid) and, maybe, Max Thomas are serious in re-learning their heritage. Both are willing to consider having white people who know anything teach them the country. 124 I turn, then, to the question: whether or not Part 2(b) of the report to the KPC provides evidence that associates Kokatha interests with the claim area? I do not consider it does. At its highest, I consider that paragraph reinforces (if that were necessary) the Kokatha interests in the area immediately to the west of the claim area. I do not consider any of the words "along the west side of Lake Torrens", or "down the western side", or "east of Lake Torrens", or "to the western shore of Lake Torrens", permit a conclusion that that association extended east beyond the western boundary of Lake Torrens. 125 I also reject the Kokatha appellants' contention that the primary judge did not refer to this report or, more specifically, to the contents of it they have identified, in coming to his conclusions at [341] and [724]. In the first place, at [291] (see [123] above), I consider his Honour gives an accurate, albeit truncated, summation of the effect of the information in Part 2(b) about the location and extent of Kokatha interests. Furthermore, the significance of the Wilyaru story itself, to the extent that it is pertinent to this issue, is mentioned in that part of the field report which his Honour has quoted extensively at [291]. It follows that I consider his Honour's statement at [341] about there being little in the ethnographic materials "to support the claims of the Kokatha People to be the sole traditional owners of Lake Torrens either at sovereignty or at any later time, at least up to the late 1980s" can be taken to include the ethnographic materials described at [290]-[291] (see [123] above). This conclusion is reinforced by the words "that material" at the very beginning of [341] (see at [38(c)] above), which plainly refer to the whole of the more recent ethnographic and historical material examined by his Honour commencing at [276] and concluding at [340]. 126 As to the conclusion at [724], first, I do not consider the statement "any material at all" should be interpreted as indicating that his Honour disregarded what materials there were that associated Kokatha interests with Lake Torrens. It is clear, in my view, from the whole of his Honour's Reasons that he took those materials into account, along with the countering materials, when he reached that conclusion. At worst, therefore, his Honour used that statement in [724] to overemphasise the paucity of those materials. More likely, his Honour was referring to the position up to "about the 1980s" consistent with the conclusion he earlier recorded at [714] as follows: There is no evidence of Kokatha occupation of any areas to the east of that western boundary of Lake Torrens at the time of first European contact, or indeed in my view in any of the ethnographic or historical material until well into the 20th century, probably about the 1980s. I do not regard any of the earlier ethnographic or historical material, including any roughly drawn maps, as progressing that physical occupation either actually or conceptually into Lake Torrens itself. I have not overlooked the Curr map, but having regard to the written observations and Willis' appropriate observation to the effect that early mapping was not intended to be used for forensic purposes, I do not place any real weight on it. (Emphasis added) 127 I do not therefore consider his Honour's treatment of the Vachon 1981 report, or the contents of that report particularised by the Kokatha appellants at ground of appeal 4(a)(i), provides any support for any of the errors raised under this ground of appeal.

The Hagen report of August 1983 128 The second item particularised by the Kokatha appellants is the report prepared by Dr Hagen in August 1983, which became a restricted exhibit (Hagen August 1983 report). As Particular 4(a)(ii) indicates, that report is mentioned in the Reasons at [309] as follows: This is a restricted exhibit, so reference to it is oblique. It is entitled "Mythological Tracks in the Vicinity of the Olympic Dam Mining Venture". It reports on Dreaming travels through a triangle "bounded loosely" by Lake Torrens in the east, and then to the west. I considered each of the myths referred to. There is no clear indication, with one exception, that the tracks concern an area to the east of the western boundaries of Lake Torrens. One myth concerns the creation of Lake Torrens, including "where the Kokatha and Kuyani meet". It therefore records a mythological track which has Kokatha associations and concerns Lake Torrens itself. 129 In submissions, in addition to the Kokatha story mentioned in that particular, the Kokatha appellants' claimed that this report recorded "two other Tjukurpa stories that travelled to or across Lake Torrens". The Kokatha appellants also referred to the preceding paragraphs of the Reasons which recorded the details of a joint report prepared by Hagen and Martin in 1983 entitled "Report on Kokatha interests in sites and land in the vicinity of the Olympic Dam Mining Project" (the 1983 Hagen and Martin report). That report contains the transcript of an interview conducted on 14 June 1983 between Max Thomas (M) and the interviewer, Jean Aitken (J), as follows (at [306]-[307]): 306 The interview appears to have occurred at Phillips Ridge/Pine Ridge, which is south-west of Lake Torrens. During the interview, in response to the question "where does the Kokatha country finish - on the other side of Andamooka?", Max Thomas replies "On the lake…Lake Torrens". 307 Relevantly, J - And where does the Kokatha country finish - on the other side of Andamooka? M - On the lake J - Lake Torrens? M - Lake Torrens J - Around the boundary edge of it? M - That's the boundary of the Kokatha people 130 The Kokatha appellants placed particular reliance on Max Thomas' statement above that Kokatha country finished "On the lake". They did not specifically challenge his Honour's observations in the immediately succeeding paragraph (at [308]) about the meaning of the words "on" in that statement, or the conflict between that evidence and the Tregenza and Hagen report of 1983, as follows: The use of the word "on" does not designate a clear boundary. The [Kokatha Appellants] did not maintain that the boundary between Kokatha country and Adnyamathanha or Barngarla country bisected Lake Torrens. Indeed, such an interpretation would not be consistent with evidence given by the Kokatha witnesses. Nor do I understand that the "boundary edge" is referring to the eastern boundary. That is so different from what Hagen recorded in the Tregenza and Hagan report of the same year to the KPC that, if that were the suggestion, it would have prompted some follow up. 131 The background to the preparation of the 1983 Hagen and Martin report is described earlier in the Reasons at [305] as follows: The [Kokatha Appellants] also relied on the 1983 report by Hagen & Martin (1983), "Report on Kokatha interests in sites and land in the vicinity of the Olympic Dam Mining Project". The cover page says the document is confidential, but it was not marked as confidential when received. It contains extracts from a transcript of an interview, apparently taken as part of a women's survey. The date is recorded as 14 June 1983 and the names of 7 participants, including Max Thomas, whom I have taken to have been referred to as "M" in the transcript. He is interviewed by "J" whom I have taken to be Jean Aitken, although nothing turns on that. 132 There are two other reports mentioned in the Reasons of which Hagen was at least one of the authors. First, at [310]-[312], his Honour described a report of September 1983 prepared by Hagen as follows: 310 This is also a restricted exhibit (although it is not clear why), being three pages of a Report by Hagen headed "Kokatha, Kuyani and Arabana Interests in the Olympic Dam Area: Results of the Current Field Survey". 311 As with other material, it records that "invariably" the direct response to questions about the interests in land was that the western shore of Lake Torrens belonged to the Kokatha and the eastern shore was Kuyani. The focus of questions, in the context of the inquiry, was the Olympic Dam area. The range of informants included Adnyamathanha men, senior Yankuntjatjara men, and Aranda men. Reference is made in responses to the activities of Jack Davis, Dicky Thomas and Max Thomas. The informants included Richard Reid, John Stewart and Max Thomas. 312 There is nothing in that document to suggest that the Kokatha country extended eastwards into Lake Torrens from its western shores. 133 Secondly, at [292]-[295], the Tregenza and Hagen report referred to above (at [130]) was described by his Honour as follows: 292 In or around February 1983, Tregenza and Hagan produced a report for the KPC entitled "Sites of Importance to Aboriginal People within the Olympic Dam [Environmental Impact Statement] Area" based on fieldwork they had conducted in January 1983, and informants who identified as Kokatha, Pitjantjatjara, Yankuntjatjara and Antakirinya. 293 That report concerns the land on the north-western side of Lake Torrens but does not directly concern the claim area. 294 Hagen recorded a number of sites associated with the mythology of the [Environmental Impact Statement] area, which were said to be based on two major groups of mythic beings passing through the area. Firstly, the Wati Kalta (sleepy lizard) and Wati Angata (frill necked lizard) (Wilyaru), and secondly the Kunkarankara or "Seven Sisters". 295 Under the heading "Kokatha or Kuyani" concerning the affiliation of Aboriginal groups with the Olympic Dam Area, Hagen writes: The limited time available to me has made it impossible to arrive at an unchallengeable conclusion on the matter. It is however worth nothing that all Aboriginal people approached on the question agreed that Kokatha territory encompassed the entire [Environmental Impact Statement] area. The eastern boundary of Kokatha territory was invariably given as the entire western shore of Lake Torrens. 134 I have set out all of the references in the Reasons to the four reports of which Hagen was either the sole, or a joint, author in order to demonstrate that there is no merit to the Kokatha appellants' contention that his Honour did not consider the contents of those reports that described the association of Kokatha interests with Lake Torrens. His Honour specifically stated at [309] that he had considered each of the myths referred to in the 1983 Hagen report. He then proceeded to described in "oblique" terms the particular Kokatha myth concerning the creation of Lake Torrens. While he did not specifically mention the other two myths that are described in that report as travelling to, or across, Lake Torrens, there is nothing to gainsay his statement that he had considered those myths along with all the other myths referred to in that report. As for the other report mentioned in the Kokatha appellants' submissions, the 1983 Hagen and Martin report, because that report was not a restricted exhibit, his Honour was able to quote verbatim the particular part of it that recorded Max Thomas' apparent claims that Kokatha country extended on to Lake Torrens. Having regard to that quote as set out at [307] of the Reasons (see at [129] above), there is no merit in the Kokatha appellants' claims that his Honour thereafter ignored that evidence. Indeed, far from ignoring it, as is apparent from [308] (at [130] above), he immediately proceeded to explain why he did not accept that it placed the boundary of Kokatha country either in the middle of Lake Torrens, or on its eastern shore. 135 For these reasons, I do not consider his Honour's treatment of the Hagen report of August 1983 or the Hagen and Martin report of the same year, particularised by the Kokatha appellants at 4(a)(ii), or as developed in submissions, provide any support for any of the errors raised under this ground of appeal.

The Fitzpatrick and Gara 1984 report and the Gara 1989 report 136 The final two items which the Kokatha appellants alleged were either not considered at all by the primary judge or, if they were considered, were subsequently disregarded by his Honour in his conclusion at [724] are particularised at 4(a)(iii) and (iv) (see at [8] above). Those particulars describe two reports prepared by Dr Gara: first, a report prepared in conjunction with Fitzpatrick in 1984 (the 1984 report); and, secondly, a Heritage Survey Report solely prepared by Dr Gara in 1989 (the 1989 report). I should note that, in their submissions, the Kokatha appellants do not mention the 1989 report, but instead, under this ground of appeal, they refer to the thesis prepared by Ms Jane Jacobs in 1983. Nonetheless, I propose to deal with this ground as it is particularised in the notice of appeal. With respect to the 1984 report, the Kokatha appellants claimed that it was not referred to in the Reasons and they complained that that indicated that his Honour did not consider it at all. Further, they claimed that, if his Honour had considered that report, he could not have come to the conclusion that he did at [341] and [724] about the absence of any support for their claims with respect to the Kokatha appellants in the ethnographic surveys "until the relatively recent past". 137 While it appears to be common ground between the Kokatha appellants and the State that the primary judge did not refer to the 1984 report in the Reasons, that fails to acknowledge the express references to that report in the Reasons and the important statement that follows those references. The express references to the 1984 report appear at [73(13)] and [73(14)] as follows: (13) Fitzpatrick and Gara 1984, "Anthropological Survey of Extractive Mineral Leases on Birthday, Four Mile and Nantilla Creeks Near Port Augusta - A Report for the Kokatha People's Committee" p 5 of the report (p 7 of the pdf document) (14) Fitzpatrick & Gara, "Anthropological Survey of extractive mineral leases on Birthday, Four Mile and Nantilla Creeks near Port Augusta - a report for the Kokatha People's Committee" - October 1984 The first item appears to be a single page extracted from the latter. While it is not entirely clear from the transcript, the latter appears to have been tendered as Exhibit K15 (see at 1925 of the transcript). 138 The important statement to which I have referred above appears in the section of the Reasons recording the tender of these reports, together with all the other evidence tendered at the trial. That section commences at [48], under the heading The Hearing, and concludes at [84] with the following statement: All that material has been considered in the course of, and for the purposes of, this judgment although there may be no specific references to it where the issue raised has been addressed in much the same manner by one of the active parties during the hearing. (Emphasis added) By this ground of appeal, the Kokatha appellants would appear to be challenging this statement. 139 That said, it is difficult to discern whether his Honour referred to Exhibit K15 in that part of the Reasons containing his review of the More Recent Ethnographic and Historical Material (at [276]-[345]). That is so partly because of the number of reports of Dr Gara that were tendered in evidence and also because the references to those reports in the Reasons do not readily correspond to the exhibit descriptions. 140 As to the first difficulty, each of the parties tendered one or more reports prepared by Dr Gara as follows: (a) the Kokatha appellants: the two reports mentioned above (at [137]); (b) the State: at [71]: a 2015 report described in parenthesis as the "Gara Report"; and at [78]: a 1988 report; (c) the Adnyamathanha appellants: at [74(4)]: the 1989 Gara report; and (d) the Barngarla appellants: at [77]: research based on the Gara Report. Thus, there appears to have been at least four reports in evidence of which Gara was at least one of the authors (allowing for the duplication mentioned above and excluding the separate research tendered by the Barngarla appellants which does not appear to have been prepared by Dr Gara). It is also worth mentioning that the 1989 report described in Particular 4(a)(iv) above was actually tendered by the Adnyamathanha appellants and not the Kokatha appellants. 141 The confusion mentioned in the second difficulty above emerges from that section of the Reasons where his Honour considered the 1989 report ([317]-[326]). Because of confidentiality restrictions, his Honour conducted that review by reference to another document in evidence: a 2003 report prepared by the National Native Title Tribunal (see at [317] of the Reasons). After reviewing the Gara material by reference to that report (at [318]-[323]), his Honour referred, at [324], to having "separately considered the Gara survey". It is not clear what his Honour meant by the word "separately", or what "the Gara survey" referred to. It is possible, although unlikely, that it was a reference to Exhibit K15. That is possible because, as is stated at [324], part of the field work for that survey was conducted in 1984. It is unlikely because that field work continued until 1988, well beyond the date of the 1984 report. It is also unclear what his Honour meant at [338], where he referred to Gara presenting the same picture between 1984 and 1988 as follows: It is significant, in my view, that the KPC in 1981 depicted the Kokatha People's traditional area as extending to the western side of Lake Torrens but not east into Lake Torrens itself, and the Kokatha informants to Vachon in 1981, and to Tregenza and Hagen in 1983, and to Gara between 1984 and 1988 presented the same picture. 142 On balance, however, it is probably correct to say that his Honour did not refer to the 1984 report in his consideration of the More Recent Ethnographic and Historical Material at [276]-[345] inclusive. 143 The question that then arises is: if Exhibit K15 was not referred to in that review, should it be inferred that his Honour did not consider it? It is appropriate to preface my answer to this question by noting that the primary judge was not required to refer to every piece of evidence tendered during the trial of this matter, all the more so where the trial was as lengthy and complex as this one and where he gave at least two other assurances in the Reasons (apart from that mentioned at [138] above) that he had considered all of the evidence (see at [408] and [601]). Turning then to the question posed above, I have just provided one compelling reason why I do not consider this inference should be drawn: his Honour's statement at [84] (set out at [138] above). 144 There are, as well, three other reasons why I do not consider that inference should be drawn. First, there is the lack of significance attached to Exhibit K15 by the Kokatha appellants themselves at the trial and, secondly, there is a similar lack of significance apparently attached to that exhibit by Dr Gara himself. Thirdly, and perhaps most significantly, when it is analysed, I do not consider the 1984 report is probative of the question raised by the Kokatha appellants in this ground of appeal: whether there is any support in that report to associate Kokatha interests with the claim area in the 1980s. 145 On the first reason, despite the importance that the Kokatha appellants have attached to the 1984 report in these appeals, I have been able to find only one passing reference to it in all of their extensive written and oral closing submissions at the trial. That appears in a table which is included in their submissions in reply to the submissions of the Adnyamathanha appellants. The table is said to summarise "both the early ethnography and more recent research in relation to the location of Kokatha and Kuyani (Adnyamathanha) country". In that table, the 1984 report is described among 24 other reports in the following terms: Date Ethnographer Data Exhibit Identification of the Claim Area - whose country Kokatha Adnyamathanha 1984 Fitzpatrick & Gara Notes the Seven Sisters (Kunkaralinya) mythology traveling (sic) to Lake Torrens K15 p.5 Claim area again linked to Kokatha mythology

146 This item does not place any great significance on the 1984 report. If the Kokatha appellants failed to tell the primary judge about that significance at the trial, it is hardly surprising that his Honour did not specifically mention it in the Reasons. This conclusion is supported by the second reason: Dr Gara himself does not appear to have discussed his 1984 report in his 1989 report, despite the fact that the latter report surveys, among other things, the ethnographic records in the early 1980s (see at [321]-[323] of the Reasons) and specifically mentions his fieldwork with Kokatha elders and other Kokatha people between 1984 and 1988 (see at [324]). On this aspect, it is also worth noting, as the Adnyamathanha appellants have pointed out, that the Kokatha appellants had the opportunity to explore this omission in cross-examination, but they chose not to do so. 147 Finally, there is the third, and perhaps most important, reason: consistent with the other two reasons above, the 1984 report does not, in fact, have the great significance the Kokatha appellants now wish to attach to it. To explain why that is so, it is necessary to review its contents. In doing so, I will need to be discreet because it was tendered as a "RESTRICTED - MEN ONLY" exhibit at the trial. 148 The 1984 report was prepared on behalf of the KPC and the Aboriginal Heritage Section at the request of the South Australian Department of Mines and Energy. It details the traditional significance of a number of areas in Birthday, Four Mile and Nantilla Creeks, near Port Augusta, which were at the time held under Extractive Mineral Leases. It was based on a survey undertaken in late 1984 by Dr Gara, on behalf of the KPC and two other persons on behalf of the Aboriginal Heritage Section. The survey was undertaken in conjunction with two Yunkuntjajara elders, the traditional Kokatha custodian for the area, and a person representing the KPC. The area of the survey was the three Creeks mentioned above. That area is located north of Port Augusta and some distance south of Lake Torrens. A large number of Aboriginal informants are recorded as providing information for the survey. While it was undertaken more than 30 years before the trial of this matter, none of those informants' names appears to correspond with the names of any of the lay witnesses who gave evidence at the trial, particularly the primary sources for the Kokatha appellants' lay witnesses, namely Max Thomas and Ningel Reid. 149 The two paragraphs of the report that are critical for present purposes cannot be set out here because they contain information that is secret and restricted. Nonetheless, the description in the table at [145] above provides a sufficient summary of the contents of those two paragraphs to demonstrate why the report does not have the great significance the Kokatha appellants contend for. The critical words are: "traveling (sic) to Lake Torrens". As the Adnyamathanha appellants have correctly pointed out, the word "to" does no more than connect the Kokatha appellants to the western side of Lake Torrens. It does not evidence a connection with Lake Torrens itself. I therefore agree with the State and the Adnyamathanha appellants that the 1984 report did not provide any support for Kokatha connection to the claim area. Put differently, even if, contrary to the conclusions I have reached above, his Honour did not consider the 1984 report in preparing the Reasons, I do not consider it is of such probative value that it would have led his Honour to a different conclusion about the absence of any support for the Kokatha appellants' claims with respect to the claim area in the ethnographic surveys produced in the 1980s. 150 Having regard to all these matters, I am therefore not willing to infer that his Honour did not consider the 1984 report in his consideration of the More Recent Ethnographic and Historical Material at [276]-[345] of the Reasons. Alternatively, if he did not do so, I do not consider that would have made any difference to the outcome to the Kokatha appellants' claim such that it constitutes appealable error. For these reasons, I do not consider his Honour's treatment of the 1984 report, as particularised by the Kokatha appellants at 4(a)(iii), provides support for any of the errors raised under this ground of appeal. 151 The 1989 report can be dealt with briefly. Even if the primary judge failed to take into account the reluctance of Dr Gara's informants in 1989 to divulge site names or the identity of mythic beings associated with sites, that does not provide positive evidence of a fact that would assist the Kokatha appellants to establish their connection to Lake Torrens, all the more so where the deficiency in their evidence related to the period before about the mid 1980s.

The four items were properly considered and not subsequently disregarded 152 For these reasons, I do not consider that the four items, as particularised by the Kokatha appellants at Particular 4(a), identify materials that were either not considered by the primary judge in his examination of the ethnographic surveys carried out in the 1980s which tended to associate Kokatha interests with Lake Torrens or, in the case of the 1984 Gara report, if it was not considered, it would have made any difference on that issue if it had been. 153 While this conclusion disposes of the Kokatha appellants' concerns with respect to [341] of the Reasons, it is necessary to address separately their contention that, even if the primary judge did consider those materials, he disregarded them when he made the statement at [724] about the absence of "any material at all" associating Kokatha interests with Lake Torrens. 154 I have already dealt with the words "any materials at all" at [724] when considering the Vachon 1981 report (see at [126] above). As to the broader import of those words, in my view, this contention places a particularly slanted interpretation on that particular phrase when it is considered in its full and proper context. To begin with, it ignores the words "until the relatively recent past" appearing in the next line. It also ignores the references to the contemporary evidence of Kokatha stories relating to Lake Torrens later in [724] and in [725]. Furthermore, it ignores the statement at [714] about the ethnographic or historical material containing records "probably about the 1980s" and the discussion of that evidence and all the other evidence called by the Kokatha appellants at [714]-[723]. Hence, when that phrase in [724] is read in the context of all these related statements, I do not consider it provides any support for the Kokatha appellants' contention that his Honour disregarded the materials associating Kokatha interests with Lake Torrens when he came to express his reasoning at [724].

The secrecy explanation 155 I turn to the second issue raised by this ground of appeal, the Kokatha appellants' claim that the primary judge failed to consider whether there may have been cultural reasons why particular information about Kokatha interests in Lake Torrens did not emerge in the ethnographic surveys carried out in the 1980s. As mentioned above, in their Particulars of this ground of appeal (at 4(b) set out at [8] above), the Kokatha appellants have listed six reasons why that information did not emerge in the 1980s. 156 There is a short answer to all of these reasons and, indeed, this aspect of this ground of appeal. It is set out in the Reasons at [355]. There, his Honour accepted that cultural reasons, namely the secrecy and danger associated with some Kokatha sites, may explain the absence of any record of the place names of those sites. However, as his Honour then went on to observe, "[w]hile that may explain the absence of recorded Kokatha place names in certain areas, it is not capable of positively supporting a Kokatha presence on the claim area". In my view, this observation is correct and applies equally to the absence of any records about the Tjukurpa the subject of this ground of appeal. If secrecy and danger explain why there is a dearth of mention of such records in the 1980s, that does not provide the Kokatha appellants with the positive evidence they need to establish the necessary connection to the Lake Torrens claim area. 157 Before I leave this aspect of this ground of appeal, I should deal separately with the reason advanced in 4(b)(iv) concerning Dr Vachon. There is a short answer to that reason as well. It is that the 2007 report to which that particular refers was not tendered in evidence at the trial. Cross-examining Sutton with respect to his reference to that report in one of his reports plainly does not constitute a tender of the former report, or even that part of it to which reference was made by Sutton. If the Kokatha appellants wished to rely upon Dr Vachon's 2007 report, they should have tendered it.

Conclusion on ground of appeal 4 158 For the reasons set out above, I do not consider ground of appeal 4 has any merit.

Ground of appeal 5 - Drs Vachon and Hagen and the Kokatha stories relating to Lake Torrens 159 This ground of appeal focuses on [719] of the Reasons (see at [65] above). First, the Kokatha appellants contended that the statement at the end of the second sentence, "did not expose any Kokatha or Western Desert stories relating to Lake Torrens itself", is contrary to the evidence of Dr Vachon. For the reasons I have already given when considering Dr Vachon's 1981 report above (at [120]-[127]), I reject this contention. Secondly, the Kokatha appellants contended that the last sentence of [719] is contrary to the evidence of Hagen and to the findings at [309]. That sentence reads: "In 1983, Hagen went through the area with Kokatha men, including Max Thomas, but again the Dreaming stories relating to Lake Torrens and its occupation or possession by Kokatha People was not adverted to." In its submissions, the State accepted that the last part of this sentence was not correct, but the Adnyamathanha appellants contended in their submissions that it was. However, whether or not it is correct, having regard to my review at [128]-[135] above of the reports of which Hagen was the sole, or a joint, author and the full discussion of that and other related evidence at [718]-[724] of the Reasons, I consider the words in that sentence, at worst, amount to a slip, but are more likely explained by the meaning that is attached to the pivotal words near the middle of the sentence: "relating to". Whichever it is, having regard to the observations I have made at [31] above about how this Court should approach the primary judge's advantage in these appeals, I am not willing to conclude that the contents of that sentence demonstrate appealable error. For these reasons, I do not consider this ground of appeal has any merit.

Ground of appeal 6 - Dr Gara and the Kokatha sites on Andamooka Island 160 Ground of appeal 6 alleges error in [325] of the Reasons based upon the contents of the "survey reports written by Dr Gara in 1989". The particular part of [325] that appears to be in issue is the statement in the last sentence that "if there were significant sites on or immediately adjacent to Lake Torrens, there is no reason why they would not have been identified". 161 While this ground of appeal refers to the "survey reports", the particulars to it identify two aspects of that "report". The two aspects identified are the discussion in the "report" of tribal boundaries and the recorded reluctance of the Kokatha informants to divulge information about sites or the identity of mythic beings associated with those sites. The particulars also rely upon the particulars in ground of appeal 4(b) above. 162 With one addition, the Kokatha appellants' submissions on this ground are to the same effect as set out above. The addition is to place reliance on his Honour's alleged failure to consider "an earlier report written in 1984 by Gara and Fitzpatrick [where] they recorded a highly secret and dangerous Kokatha Tjukurpa that travels to Lake Torrens". I have already rejected as unmeritorious the contention with respect to the Fitzpatrick and Gara 1984 report above (see at [136]-[150]). 163 As to the balance of this ground of appeal, or the particulars, or submissions, supporting it, I consider it, too, is unmeritorious for the same reasons as I have given above with respect to particular (b) in ground of appeal 4 (see at [155]-[157]). In essence, even if the boundaries between the Kokatha, Kuyani and other Peoples were flexible and even if Kokatha informants were unwilling to divulge secret information, neither of these factors assists to provide positive evidence in support of the Kokatha appellants' case to possess native title rights and interests in the Lake Torrens claim area. For these reasons, I do not consider this ground of appeal has any merit.

Ground of appeal 7 - a catchall ground 164 This ground of appeal is a catchall ground which does not require separate consideration. It falls with all of the other unmeritorious grounds above.

Other matters 165 Before I conclude on the Kokatha appellants' grounds of appeal, there are two final matters I should mention. First, in submissions, the Kokatha appellants asserted: "that [the primary judge] … erred in failing to consider the Kokatha claim in respect of only a part of the claim area." This appears to refer to Andamooka Island. If that is so, in their written submissions, the Kokatha appellants conceded that they had not made an alternative claim to part of the claim area before the primary judge. Self-evidently, his Honour could not therefore have erred in not considering such a claim. In any event, since this alleged error is so significant and since it is not mentioned anywhere in the Kokatha appellants' notice of appeal, it is not appropriate to entertain it in these appeals. 166 Secondly, in oral submissions, the Kokatha appellants raised another error that does not appear in their notice of appeal. It was that, at [724], the primary judge sought to "bolster" his conclusion about the absence of materials tending to associate Kokatha interests with Lake Torrens by relying upon Tindale's 1974 maps. That was said to be an error because all of the anthropological experts had disagreed with the views expressed by Tindale in those maps. While this error is also not raised by their notice of appeal, it is a relatively minor matter so I will address it. I do not consider his Honour made any such error. It is relatively clear, in my view, that the maps to which his Honour referred at [724] were not Tindale's 1974 maps, but rather those produced by him in 1940, as described earlier in the Reasons at [262]-[267].

Conclusion on the Kokatha appellants' appeal 167 For the reasons set out above, none of the Kokatha appellants' grounds of appeal has any merit and their notice of appeal must therefore be dismissed.

An overview of the alleged errors 168 Based upon the content of their written and oral submissions and having regard to the terms of their notice of appeal (see at [9]), the Adnyamathanha appellants alleged that the primary judge made the following errors: (a) various errors with respect to connection - grounds of appeal 1, 2.1 to 2.3 and 3 to 4; (b) misuses of the prior consent determinations - grounds of appeal 5 and 6; (c) errors on particular evidentiary matters - ground of appeal 7; and (d) the failure to make a ruling under s 86 - ground of appeal 2.4; (e) the rejection of the shared claim with the Barngarla appellants - ground of appeal 8. 169 Greater details of some of these alleged errors (and the paragraphs of the Reasons concerned) are provided in the Adnyamathanha appellants' written submissions where they claimed that the primary judge had made three "core errors" as follows: (a) "to require proof of occupation or possession of the claim area [[d]espite [Reasons] at [677], [604], [733] (part in parentheses), see [342]-[345], [708], [734], [742], [754], [235], [271]. See also [192], [206], [212], [234], [707]-[709], [724], [746]-[747]] when the appellant's laws and customs required neither"; (b) "to mis-use inferences arising from the prior consent determinations as a 'filter' or 'yardstick' against which to assess, weigh and reject evidence [[Reasons] at [171], [709], [52], [178]-[195], [204], [334]-[345], [364]-[365], [479], [738], [743]-[744], [762]-[763]. See also [170], [174], [224], [226], [250], [304], [315], [331], [557], [587]-[588], [598], [712], [715], [742], [758]]"; and (c) "to refuse to draw inferences favourable to the appellant which were properly open [[Reasons] at [707]-[711], [772]]" (cases omitted). 170 Paragraph [169(b)] above clearly concerns the issue raised by grounds of appeal 5 and 6 above (see [168(b)]). Paragraphs [169(a)] and [169(c)] above provide details of the core errors relating to the connection issue raised by grounds of appeal 1 to 4 (excluding 2.4) above (see at [168(a)]). I will therefore consider those grounds of appeal in light of the further details provided in these three core errors. 171 There are two other aspects of the list of alleged errors at [168] above that can be disposed of at this point. The first concerns the s 86 issue, which is raised by ground of appeal 2.4 (set out at [9] above). During oral submissions, whilst not abandoning that issue, counsel for the Adnyamathanha appellants conceded that it did not have a practical consequence in these appeals. In its submissions, the State explained why that was so. It submitted (correctly in my view) that, while the primary judge did not make a ruling under s 86, such a ruling was implicit from the Reasons because his Honour gave the "fullest weight to the determinations in favour of each of the appellants in rejecting any evidence that challenged the fundamental bases for native title existing in each of those parties in the respective areas that those determinations covered". This explanation accords with, among other things, the description of the Adnyamathanha No 1 determination at [4] and [9] of the Reasons and his Honour's observations at [11] and [12] of the Reasons as follows: 11 Obviously, at sovereignty, the native title rights and interests so recognised would not have been extinguished either in whole or in part by the acts which have occurred subsequent to sovereignty. 12 While the named applicants who comprise each of the claimant groups are no longer identical to the applicants who were authorised by the claimant groups and in whose favour the determinations were made in Kokatha Part A, Adnyamathanha No 1 and Barngarla No 2, they identify as being part of the same societies as those recognised as the holders of native title rights and interests under the NTA in Kokatha Part A, Adnyamathanha No 1 and Barngarla, respectively. Nothing presently turns on those differences. However, for reasons which are addressed below, the precise composition of the three competing claim groups provokes certain observations about the extent of certain common ancestry. 172 Having regard to these matters and except to the extent that the effect of the three consent determinations will be considered under grounds of appeal 5 and 6 above, I do not consider it is necessary for this Court to consider the s 86 issue as a separate issue in these appeals. 173 The second matter concerns the shared claim issue which is raised by ground of appeal 8 above. That ground relies upon the Adnyamathanha appellants and/or the Barngarla appellants succeeding in their appeals and, because of my conclusion that all of these appeals must fail, it becomes unnecessary to consider that issue. Accordingly, in considering the grounds raised by the Adnyamathanha appellants' notice of appeal, I will deal with the three alleged core errors set out at [169] above in that order and then turn to consider the various evidentiary-related errors that are mentioned in [168(c)] above. In the latter, I will also include the other evidentiary-related errors mentioned in the Adnyamathanha appellants' submissions, but not specifically covered by their grounds of appeal.

The primary judge's dispositive reasoning 174 Before turning to consider their alleged errors, I will first, as I have foreshadowed above, summarise the dispositive reasoning contained in the final Consideration section of the Reasons as it applies to the Adnyamathanha appellants' claims. 175 It is convenient to record first how that section is structured. In the introductory paragraph (at [727]), the primary judge set out the various issues the State and the Kokatha appellants raised with respect to the Adnyamathanha appellants' claims. First, the Kokatha appellants, but not the State, raised the issue whether at sovereignty the Adnyamathanha People "acknowledged and observed a body of laws and customs under which its members possessed traditional rights and interests in, and had a relevant connection with, Lake Torrens including Andamooka Island" (at [728]). Secondly, both the Kokatha appellants and the State raised the issue whether there was "any continuity of any relevant connection with, any part of Lake Torrens either at sovereignty or subsequently" (at [727]). Further on that issue, they both claimed that the only evidence of any present day use of Andamooka Island by the Adnyamathanha People was non-traditional in nature (at [727]). 176 At [728], his Honour expressed doubts about the Adnyamathanha appellants' position on the first issue, saying "[i]t is not clear, in my view" and he then went on to state why he held those doubts at [729]-[731]. 177 However, at [732], he said that he did not have to resolve that issue and then explained at [733] that was so because, in any event, their claim failed on the second issue. 178 Thereafter, from [734]-[753], his Honour explained in some detail why he had reached the view he had reached on the second issue that the Adnyamathanha appellants had failed on the connection inquiry as follows (at [754]): The end result is that, upon the whole of the evidence, as I have indicated, I accept the State's submission that the Adnyamathanha claimants have not established that any occupation of Lake Torrens, which they had according to traditional laws and customs at the time of sovereignty or at the time of first European contact has been maintained substantially in accordance with their traditional laws and customs over Lake Torrens to the present time. 179 Having outlined the structure of the final Consideration section as it applies to the Adnyamathanha appellants' claim, it is next convenient to summarise the explanation his Honour gave for the conclusions he reached on the two issues set out above. 180 As to the first issue, after noting the evidence confirming the presence of some Kuyani People and some Barngarla People on the western side of Lake Torrens around the 1850s, and "possibly considerably earlier", and assuming the failure of the Kokatha appellants' claim to Lake Torrens, his Honour postulated the possibility (at [729]) of an inference that the Kuyani and/or Barngarla had traditional rights and interests in Lake Torrens consistent with the archaeological evidence and the agreed anthropological evidence. However, his Honour then observed at [730] that, in the ethnographic and historical material, "there is virtually no specific reference (other than some generic and imprecise mapping as discussed in the analysis of the ethnographic material) that the Adnyamathanha People in fact enjoyed such rights and interests over Lake Torrens itself at the time of first European contact, or working backwards at sovereignty". On this aspect, his Honour expressed the view that the evidence that some Adnyamathanha People traversed Lake Torrens in the early decades of the 20th century was "not very strong evidence". 181 Furthermore, his Honour noted (at [731]) that, "even if that state of affairs were shown to have existed at the time of the first European contact", the fact that the Kokatha People held native title rights and interests in the area to the west of Lake Torrens at sovereignty and that the Kuyani People, whilst still in part of the western side of Lake Torrens, were moving eastwards in that period, would "operate to discourage any inference that, for all of the 60 years or so proceeding (sic - preceding) European contact, the Kuyani (Adnyamathanha) People held native title rights and interests over the entirey (sic - entirety) of Lake Torrens to its western boundaries". Nonetheless, as noted above, his Honour then concluded he did not "need formally to resolve that question" and explained why that was so at [733] (see at [177] above). 182 Thereafter, as to the second issue, his Honour expressed his views on numerous aspects of the evidence called by the Adnyamathanha appellants and explained why he concluded at [754] that on "the whole of the evidence" the Adnyamathanha appellants had failed to discharge their onus on the connection inquiry under s 223(1). That was so because they had "not established that any occupation of Lake Torrens, which they had according to traditional laws and customs at the time of sovereignty or at the time of first European contact has been maintained substantially in accordance with their traditional laws and customs over Lake Torrens to the present time" (at [754]). In highly summarised form, in that review of the whole of the evidence, his Honour highlighted: (a) the lack of "appropriate topographic focus" on Lake Torrens itself in the Adnyamathanha appellants' lay evidence, notwithstanding that some Adnyamathanha People "continued to assert responsibility to look after Adnyamathanha country and its spirits and Dreaming stories, including in relation to Lake Torrens" (at [734]); (b) the unavailability of an inference to fill the approximately 60 years temporal gap between first European contact and sovereignty to establish continuity of connection over either the western or eastern side of Lake Torrens (at [735]); (c) the fact there was "virtually no ethnographic or historic material identifying any particular activities or ceremonies of the Adnyamathanha People in relation to the Lake itself" (at [736]); (d) the evidence relating to the Adnyamathanha mura (Dreaming stories and beliefs) was not "persuasive evidence of the continued connection required by s 223 and 225 of the NTA as prescribed in Ward HC" (at [738]); (e) the conflicts and/or deficiencies in the male and female Adnyamathanha appellants' lay witnesses' evidence relating to Lake Torrens (at [739]-[741], [750]-[752] and [754]); (f) the lack of evidence of ceremonial practices on, or immediately adjacent to, Lake Torrens, save for the storage of some sacred objects in an unidentified location and the sparsity of the evidence that the Adnyamathanha appellants "maintained an ongoing spiritual connection with and over Lake Torrens, according to their traditional laws and customs, since sovereignty to the present time" (at [747]-[748]); and (g) the fact that the Adnyamathanha appellants' "lay evidence exposed that there is no clear Adnyamathanha/Kuyani word for Lake Torrens or of other physical features in Lake Torrens and Andamooka Island" (at [753]). 183 Having outlined how his Honour disposed of the Adnyamathanha appellants' claims in the final Consideration section, I now turn to consider the first of their alleged errors set out above.

Error (a): no requirement for occupation or physical possession 184 In this first core error, the Adnyamathanha appellants alleged that the primary judge focused on the "occupation or use of Lake Torrens", in the sense of physical presence on, or use, and thereby applied the wrong test for connection under s 223(1) of the NTA. In support of this contention, they provided a list of approximately 20 usages in the Reasons of the expression "occupation", some of which they conceded were not reflective of error. However, they claimed that when the whole of the usage of that expression is considered, the identified error emerges from the Reasons. 185 I do not accept this contention. In my view, it is abundantly clear from the Reasons that his Honour was well aware that occupation or use for the purposes of s 223(1) of the NTA did not involve western legal concepts of the physical presence, or the use or possession of land, but rather occupation, use, or possession according to the traditional laws and customs of the Aboriginal or Torres Strait Islander Peoples concerned. His Honour clearly demonstrated that he was alert to this distinction by his use throughout the Reasons of words to the latter effect, or that expressed that concept of occupation. For example, "[w]here it is appropriate to do so, weight should be accorded not only to physical use of the claim area but also to evidence of spiritual connection" (at [95] in The Law section); "occupation or use of Lake Torrens in accordance with their traditional laws and customs (at [343]); "the [Appellants] relied heavily on evidence of Dreaming stories and beliefs to demonstrate connection … in part due to the nature of the claim area being unsuited to long term occupation" (at [604]); "ongoing physical occupation of the Lake is not required to be shown to demonstrate physical presence" (at [677]); and "despite the uninviting environment (other than on Andamooka Island) - there were people … who 'occupied' in some sense, a part, parts, or the whole, of Lake Torrens including Andamooka Island at sovereignty according to their traditional laws and customs" (at [708]). 186 The two occasions in the proffered usages where the word "occupation" did not appear in conjunction with the words "by their traditional laws and customs", or words that conveyed that concept, both occurred in the section of the Reasons where his Honour was recording the details of the Historical and Early Ethnographic Evidence (at [234] and [235]). I do not consider those usages reflect any relevant error on the part of the primary judge when they are considered in the context which his Honour described at [204] where he drew a distinction between the two concepts with respect to that evidence: "In any event, the early ethnography provides only indirect evidence from which the use of Lake Torrens may be inferred, as opposed to direct evidence of rights and interests being held and/or exercised in respect of the claim area itself." 187 For these Reasons, I do not consider that there is any merit in this first core error.

An overview of the alleged error 188 In this second core error, the Adnyamathanha appellants claimed that the primary judge had misused "inferences arising from the prior consent determinations as a 'filter' or 'yardstick' against which to assess, weigh and reject evidence". As with the previous core error, they provided a list of numerous paragraphs of the Reasons where they claimed this error was manifest (see at [169(b)] above). 189 The Barngarla appellants raised a similar issue to this core error in their second and third grounds of appeal (see at [10] above). Accordingly, I will deal with that issue in this part of these reasons as well.

The contentions 190 In their submissions, the Adnyamathanha appellants submitted that the three consent determinations described at [4] and [8]-[10] of the Reasons were conclusive as to which groups now hold certain native title rights and interests in the determined areas, but they are not conclusive as to why those groups hold those rights and interests. Nonetheless, they sought to emphasise that they did not seek any findings at the trial "which were necessarily inconsistent with" those three determinations. Further, they submitted that a judgment in rem "is only a basis for drawing inferences, or finding further facts, which flow from the determination itself (rather than facts or findings assumed to underpin the determination)". Noting that the primary judge had ruled at [54], [57], [171], [364], [479] and [743] that no weight would be given to any lay or expert evidence that was inconsistent with those three determinations, they contended that his Honour had erred in making and applying those rulings. In particular, they contended that the primary judge was "wrong to rule (at [192]) that it was 'a determined fact' that the Kokatha People occupied the area immediately west of the claim area at Sovereignty". They also contended that the primary judge was wrong at [194] and [203]-[340] "to prefer his own analysis of the ethno-historic record … to the opinions of Mr Ellis, Prof Sutton and Dr Fergie et al". They further contended that the primary judge was wrong at [205], [235], [250], [274] and [315] to "filter" those materials by reference to the prior determinations and to "disregard or de-weight" those parts of them that did not support Kokatha Part A. Finally, with respect to the Kokatha Part A determination, they expressly acknowledged that an inference necessarily arose that "at some point prior to that determination, [they] ceased to hold [native title rights and interests] in any part of the Kokatha determination area". They added "one may infer this occurred in 2014". 191 The Barngarla appellants made submissions to similar effect. They contended that the primary judge had erred in failing to attempt to reconcile the three consent determinations, including the Kokatha Part A determination, with the evidence before him, particularly the expert and historical evidence.

Consideration 192 Before proceeding to consider these contentions, it is worth identifying the position the Adnyamathanha and Barngarla appellants adopted at the trial with respect to the three consent determinations and particularly the Kokatha Part A determination. At [52], the primary judge recorded that: "It was common ground as between the parties, that no expert proposed to give evidence directly to attack any finding of fact which is fundamental to or underlies any existing determination and that the focus of the experts' reports was on who, if anyone, occupies or is entitled to native title rights over Lake Torrens." Accordingly, his Honour recorded that, during the course of the trial, "the parties accepted the primary findings of fact necessary to support the three [consent] determinations on the areas of land surrounding Lake Torrens". 193 Despite that common ground, his Honour noted at [54] that, in respect of the expert evidence, there "were in fact [attacks] on findings in [the] existing determinations". The form of those attacks is recorded at various points throughout the Reasons. For instance, at [178] (Sutton), [180] (Ellis) and [181] (Fergie et al), three of the expert anthropologists were recorded as advancing the position that the area of the Kokatha Part A determination was Kuyani and/or Barngarla traditional country and not the traditional country of the Kokatha People at sovereignty. The same position was advanced by the Adnyamathanha appellants and the Barngarla appellants at [211]-[212] with respect to the ethno-historical evidence and examples have already been given earlier in these reasons of a similar position being adopted by the lay witnesses called by both of those appellants. This contrasts with the more nuanced position the Adnyamathanha appellants have adopted in these appeals, as can be seen from the outline of their contentions above. They now accept that, at some point prior to the Kokatha Part A determination in 2014, they ceased to hold native title rights and interests in any part of the Kokatha Part A determination area. 194 It is these attacks on the findings in the three consent determinations, particularly the Kokatha Part A determination, that explain why his Honour found it necessary to make the rulings he did at [54]-[57] with respect to the evidence generally, at [170]-[174] with respect to the agreed facts and at [175]-[202] with respect to the anthropological evidence. Specifically, at [54], his Honour ruled that the expert evidence was admitted "to inform the extent to which there are native title rights in any one of the [Appellants] over Lake Torrens, and for any other legitimate purpose … [but] would not be used to undermine an existing determination and that, in the event that it had no other relevant use, there would be no weight accorded to it". This ruling was then applied to the lay evidence at [57] where his Honour ruled that "no weight would be given to evidence that was inconsistent with [the consent] determinations". 195 Then, later in the Reasons, when considering the statement of agreed facts, these rulings were reiterated with a particular focus on the Kokatha Part A determination in the following terms at [171]: "To the extent that evidence is sought to be led which is directly inconsistent with the necessary findings in Kokatha Part A, I have not accorded it any weight." On this theme, at [189], his Honour made the following observations, which were specifically directed to the evidence of Sutton, Ellis and Fergie et al mentioned above: However, I accept the contention of the [Kokatha appellants] that certain aspects of the expert anthropological evidence referred to does not merely assert that there were shared rights as between the Kokatha and the Kuyani at sovereignty, but rather it asserts that native title rights and interests in the area immediately to the west of Lake Torrens at the time of sovereignty were held by only the Kuyani or the Barngarla, and it also says that at sovereignty most, if not all, Kokatha People resided or were widely thought to be associated in some way with country further to the west, and not with areas immediately to the west of Lake Torrens. From that premise, it is said that it is unlikely that the eastern border of Kokatha country extended west into Lake Torrens at the time of sovereignty, because the Kokatha People were not at sovereignty in the western side of Lake Torrens. In my view, such a premise is not consistent with the decision in Kokatha Part A. (Emphasis added) 196 The "shared rights" his Honour mentioned in this paragraph refer back to the following observations he made at [186]: [H]aving regard to the approach of the Court in Banjima FC, a finding that at sovereignty more than one group, and particularly more than one of the Applicant groups, might have held some rights and interests in the land which has now been determined to be the land of the Kokatha, Adnyamathanha or Barngarla People respectively would not necessarily be inconsistent with the [consent] determinations … (Emphasis added) His Honour made some similar observations about this possibility in The Law section of the Reasons at [116]. However, it was unnecessary for him to explore this shared rights issue because, as is recorded at [189] (see at [195] above), neither the Adnyamathanha appellants, nor the Barngarla appellants, advanced such a claim. 197 His Honour further reiterated these rulings above at [190], [192] and [193], where he said: "[c]onsequently, to the extent that the expert anthropological views are premised upon the Kokatha People not having native title rights and interests in the area immediately to the west of Lake Torrens at sovereignty, I do not place weight on it" ([190]); and "I have ruled that it is a determined fact that the Kokatha People occupied, and held native title rights in, the area immediately to the west of Lake Torrens at sovereignty" ([192]); and finally, "to the extent that the dates adopted by an expert or experts indicate that the Kokatha People annexed the land to the west of Lake Torrens only post-sovereignty, or are otherwise inconsistent with the findings in Kokatha Part A, no weight has been placed on that evidence". He added that he would also approach "with particular caution" expert witnesses who relied "in part on those dates" (at [193]). Despite these rulings, his Honour did not consider it necessary to strike out any parts of the expert reports (see at [195]). 198 I turn then to consider the claims the Adnyamathanha and Barngarla appellants have made that the primary judge erred in making and applying the rulings described above. It is convenient to begin that consideration by observing that the fundamental matters of which a determination of native title dispose, once and for all, are determined according to the provisions of the NTA, particularly ss 223(1) and 225 (see Ward HC at [16] and [25], Yorta Yorta HC at [32], Munn (for and on behalf of the Gunggari People) v State of Queensland [2002] FCA 486 at [8] per Emmett J, Gumana v Northern Territory of Australia (2005) 141 FCR 457; [2005] FCA 50 at [127] per Selway J and Akiba v Queensland (No 3) (2010) 204 FCR 1; [2010] FCA 643 (Akiba) at [157] per Finn J). Consequently, it is the provisions of the NTA, not common law principles, that have led to a native title determination being commonly described as a judgment in rem that is binding on the whole world (Kokatha People v State of South Australia [2007] FCA 1057 (Kokatha) at [33] per Finn J; referring to Wik Peoples v State of Queensland (1994) 49 FCR 1, Ward FC at [368]-[369]. See also Dale v Western Australia (2011) 191 FCR 521; [2011] FCAFC 46 at [92]). 199 Under ss 13(1), 61, 81 and 87, this Court has the jurisdiction and power to make an approved determination of native title as defined in s 13(3), including a consent determination. Section 94A requires any such determination to set out "details of the matters mentioned in section 225". Section 225 is already set out above (see at [45]). Putting aside s 225(e), which is not relevant for present purposes, that section requires a determination of native title in relation to a particular area of land to determine who holds the rights comprising the native title concerned (s 225(a)); the nature and extent of those native title rights and interests (s 225(b)); the nature and extent of any other interests in relation to that area (s 225(c)); and the relationship between these two sets of rights and interests taking into account the provisions of the NTA (s 225(d)). Accordingly, it was those details of the matters in the three consent determinations that the primary judge briefly summarised in the Reasons at [8]-[10]. 200 The native title and the native title rights and interests referred to in s 225 are defined in s 223(1) (see at [45] above). It is that native title and those rights and interests that are recognised and protected by the provisions of s 11(1) of the NTA (see Ward HC at [16]). Fundamental to that definition are the traditional laws and customs from which those native title rights and interests derive (see Ward HC at [20]). Specifically, those rights and interests which "owed their origin to the traditional laws acknowledged and the traditional customs observed by the Indigenous peoples concerned" (see Yorta Yorta HC at [37]). To gain that recognition and protection the rights and interests concerned must therefore be possessed under traditional laws and customs (see Yorta Yorta HC at [40]-[42]). It was only those rights or interests which survived the change in sovereignty such that they could be recognised after the assertion of the new sovereignty. Any rights and interests in land that came into existence post sovereignty were not given effect by the new sovereign order (see Yorta Yorta HC at [43]-[44]). 201 It necessarily follows that one of the most important fundamental matters disposed of, once and for all, by a determination of native title made under the NTA is that the native title rights and interests described in that determination are of that traditional nature and having those pre-sovereignty origins. It therefore follows that one of the main reasons why the Kokatha Part A determination was made was because the Kokatha People had established that their native title rights and interests in the area covered by that determination area were of that traditional nature and having those pre-sovereignty origins. The Adnyamathanha appellants are therefore wrong in their contention that the three consent determinations and, in particular, the Kokatha Part A determination, were not conclusive as to why the Kokatha People hold the native title rights and interests they do in the Kokatha Part A determination area. 202 This conclusion has adverse consequences for two other contentions the Adnyamathanha appellants made in these appeals. First, as I have explained when setting out the principles concerning the connection inquiry above ([46] and [51]-[59], in order to obtain the Kokatha Part A determination, the Kokatha People also had to establish that the acknowledgement and observance of their traditional laws and customs which gave rise to their rights and interests in that determination area had continued, substantially uninterrupted, and their connection with that area under those laws and customs had been substantially maintained, since sovereignty. This continuity of observance and connection was therefore fundamental to their rights and interests in that area. Accordingly, I consider it follows that the Kokatha Part A determination also determined as a fundamental matter, once and for all, that the Kokatha People's native title rights and interests in that area held that intrinsic continuity element. That is to say, those rights and interests did not come into existence as a consequence of, and from, the Kokatha Part A determination. Rather, that determination recognised as a fundamental matter that those rights and interests had existed continually in that form at all times in the past to at least sovereignty. The Adnyamathanha appellants are therefore wrong in their contention that the Kokatha Part A determination was only conclusive as to the native title rights and interests that the Kokatha People now hold in that area, or at least they have held since it was made in 2014. 203 Secondly, and perhaps most importantly for present purposes, this analysis explains why the Adnyamathanha appellants are also wrong in their contention that the native title rights and interests they purportedly held in the Kokatha Part A determination area inferentially ceased in 2014 at about the time the Kokatha Part A determination was made. In this respect, it is important to recall that at trial, and in these appeals, the Adnyamathanha appellants have maintained that they had traditional rights and interests in the Kokatha Part A determination area and that they did not share those rights and interests with the Kokatha People. So much is clear, for example, from the position recorded by the primary judge at [189] (set out at [195] above), which led directly to the rulings he made with respect to the expert evidence. This is important because, as I have explained above, traditional native title rights and interests cease to attain their traditional nature and cease to hold their intrinsic continuity element if there is a substantial interruption of the acknowledgement and observance of the laws and customs under which they arise, or if the connection with the area of land and waters concerned under those laws and customs has not been substantially maintained. Once either, or both, of these incidents occurs, those rights and interests cease to be traditional rights and interests. They are no longer rooted in their traditional pre-sovereignty traditional laws and customs (Yorta Yorta HC at [53]). They, therefore, lose the intrinsic continuity element I have mentioned above and they thereby cease to exist as traditional rights and interests for all time. Expressed in the terms used by Brennan J in Mabo v Queensland (No 2) (1992) 175 CLR 1, "the foundation of native title has disappeared"; and "it cannot be revived for contemporary recognition" (at 60). And, critically for the purposes of the NTA, they no longer retain the three characteristics necessary to meet the definition of traditional rights and interests in s 223(1) of the NTA (see Ward HC at [17]). They therefore cannot be recognised under the NTA. While it may be accurate to say that this state of affairs occurs at a particular point in time, for example in 2014, it is not accurate to say that it occurs from that point in time. Specifically, I do not consider it is accurate to contend, as the Adnyamathanha appellants did, that, prior to that point in time, those rights and interests still existed as traditional rights and interests in that area of land under the NTA such that they could be used to support their claim to hold traditional rights and interests in the adjoining claim area. That is to say, once they disappeared as traditional rights and interests, their past existence became immaterial for the purposes of the NTA. 204 For these reasons, I do not consider the primary judge committed any error in any of the rulings above concerning the effect of the consent determinations and, in particular, with respect to the Kokatha Part A determination. That is, his Honour was correct in ruling that those determinations determined that the three Peoples concerned held (and hold) the traditional rights and interests described in them with respect to the area to which they related at sovereignty and at all times since then. Furthermore, I consider his Honour committed no error when he ruled that he would accord no weight to any evidence that was directly inconsistent with those determined facts. Expressed with particular reference to Kokatha Part A, I do not consider his Honour committed any error in ruling that he would place no weight on any evidence that was premised upon the Kokatha People not holding traditional native title rights and interests in the area immediately to the west of Lake Torrens at sovereignty, and at all times since then. As his Honour correctly explained at [190], that premise was fundamental to the Kokatha Part A determination. The Adnyamathanha and Barngarla appellants' contentions to the contrary must therefore be rejected. Furthermore, for the reasons that follow, I also reject their related contentions that his Honour incorrectly applied those rulings in the Reasons when he came to weigh the lay and expert evidence before him. In this analysis, I have had particular regard to the paragraphs of the Reasons identified in the Adnyamathanha appellants' submissions (see at [169(b)]). I have also based it upon the positions the Adnyamathanha appellants and Barngarla appellants adopted in relation to this evidence at trial. 205 The paragraphs to the end of the section of the Reasons where his Honour dealt with the objections to the expert anthropological evidence have already been considered above. That includes [52], [170]-[171], [174] and [178]-[195] about which complaint is made by the Adnyamathanha appellants. Those paragraphs comprise the rulings and the reasoning supporting them rather than any application of those rulings. Insofar as the reasoning is concerned, I have explained above why I consider none of those paragraphs discloses any errors concerning the effect of the consent determinations. Thereafter, his Honour turned to review the evidence adduced by the three appellants commencing with the Historical and Early Ethnographic Evidence from [203]. Two complaints are made by the Adnyamathanha appellants about his Honour's review of that evidence. First, they claimed he was wrong to conduct his own analysis of that evidence. That contention can be rejected at once. It was a matter for his Honour to decide how he proceeded to conduct that review. Choosing to do his own analysis of the evidence was a course that was properly open to him and it certainly does not demonstrate error. 206 Secondly, the Adnyamathanha appellants complained about the remarks his Honour made: [204], [224], [226] and [250]. At [204], his Honour said he would view the Historical and Early Ethnographic Evidence with caution because "it has been established in Kokatha Part A, Adnyamathanha No 1 and Barngarla that in the areas surrounding Lake Torrens the Kokatha People, the Adnyamathanha People and the Barngarla People, and in respect of a small stretch of land the Arabana People, held native title rights and interests at sovereignty and continue to do so". As I have explained above, these remarks are correct. They are consistent with the rulings above and, indeed, they appear to be entirely consistent with the contents of the statement of agreed facts (see at [169] of the Reasons). 207 At [224], in the course of attempting to interpret a statement made by Shurmann in 1844, his Honour said that whatever that statement meant, it was "not inconsistent with the decision in Kokatha Part A that the area to the west of Lake Torrens is, and was at sovereignty, Kokatha country or the agreement by the expert anthropologists that the inhabitants to the west of Lake Torrens at sovereignty may have included Adnyamathanha People". The agreement referred to in this statement is recorded at [191]. These statements are correct and consistent with his Honour's rulings. They certainly do not manifest error on his part. 208 The same observations apply to the statements at [226], with respect to a report by Oakden and Hulkes in 1857, that it "does not place the Kokatha People any further east than the western side of Lake Torrens, already determined to be Kokatha country in Kokatha Part A". Finally, it was clearly open to his Honour at [250] to conclude with respect to the debate (see at [249]) about the contents of the map and paper produced by Mathews in 1900 that: I do not regard the map and paper as probative in any material way of any one of the applicant groups holding native title rights in the area immediately to the western side of Lake Torrens, or on Lake Torrens itself. Given the decision in Kokatha Part A, I am unable to ascribe any real significance to the reference to "the Hillary Kakkarurra, Yallingara … tribes" to the west of Lake Torrens if it excludes the Kokatha People. Apart from this conclusion being open to his Honour, these statements are consistent with his rulings. 209 There is, therefore, no validity in any of the Adnyamathanha appellants' complaints about his Honour's review of the Historical and Early Ethnographic Evidence. I should add that I do not consider that his Honour engaged in any "filtering" of that evidence in any way indicative of error on his part. 210 In the next section of the Reasons, his Honour turned to review the More Recent Ethnographic and Historical Material (at [276]-[345]). The Adnyamathanha appellants complained about the use his Honour made of the Kokatha Part A determination at four points in that review: at [304], [315], [331] and at [334]-[345]. 211 At [304], his Honour made some concluding remarks following his review of a series of maps appearing in the 1983 publication of Jacobs' entitled "Aboriginal Land Rights in Port Augusta". In that context he said: All of the lines on the maps are in broad lines rather than apparently responsive to particular geographical features. I do not think they can be used other than for giving a general picture. In my view, that general picture so far as presently relevant is consistent with the determinations in Kokatha Part A and Adnyamathanha No 1, but does not advance the critical issues. I do not consider that any of these statements are inconsistent with his Honour's rulings or demonstrate any error on his part. 212 At [315], his Honour made the following comments which the Adnyamathanha appellants contend involved a "filtering" of the evidence because, so they claim, his Honour wrongly considered it was inconsistent with the Kokatha Part A determination: Whilst that material cannot be useful to demonstrate a state of affairs inconsistent with the Kokatha Part A determination, it provides no support for the Kokatha claim extending over Lake Torrens itself. 213 "That material" to which his Honour refers appears at [314] where he reviewed two maps produced by Berndt from 1964 to 1988 and observed: A dividing line is shown on the map to the west of Lake Torrens running through Kingoonya, Bon Bon and Mount Eba. To the west, it shows people with generation moieties, including the 'Gogoda' people, and to the east including all of Lake Torrens are matrilineal moiety societies, described by Sutton as Lakes societies. The objectionable aspects of Sutton's evidence (mentioned in the last line of [314] above) were recorded at [178] of the Reasons (summarised at [193] and [195] above). The proposition in [314] above appears to be in the same category. That statement is therefore consistent with the rulings. 214 It follows that, when [315] is viewed in the context of the debate about the admissibility of the expert evidence described earlier in these reasons, it is apparent that, in that paragraph, his Honour has accurately reiterated the effect of his rulings. Otherwise, his Honour's comments are adverse to the Kokatha appellants' claims, not those of the Adnyamathanha appellants. I do not therefore consider that the contents of [315] are indicative of any error on his Honour's part. The observation at [331] regarding the map in the Encyclopaedia of Aboriginal Australia edited by Horton in 1994 is in the same category. It accurately applies the rulings and is otherwise adverse to the Kokatha appellants' claims. 215 Paragraphs [334]-[345] appear at the end of the More Recent Ethnographic and Historical Material section. Those paragraphs essentially fall into three groups. The first, at [334]-[336], concerns the 1996 report of Willis and Abdel-Azziz. There does not appear to be any mention of any of the consent determinations in those paragraphs. The second group, at [337]-[340], contains some comments on the ethnographic materials since about 1980. Again, there does not appear to be any mention of any of the consent determinations in those paragraphs. However, the last group (at [341]-[345]), under the heading Further Comment, is in a different category. One of those paragraphs does expressly refer to the Kokatha Part A determination (at [344]) as follows: As I have said, I am unable to place any real weight upon the fact of adjacent occupation. The Kokatha Part A determination means that, at and from settlement, the Adnyamathanha and Barngarla Peoples were not the traditional owners of the land immediately to the west of Lake Torrens. That was and is the country of the Kokatha People. The fact that Adnyamathanha and Barngarla People, in the period from settlement may have been and been recorded on the western side of Lake Torrens over that period does not demonstrate or tend to demonstrate that they had at settlement, and have continued to have, any particular traditional relationship with Lake Torrens. In the light of the Kokatha Part A determination, their presence on that country (as clearly occurred from time to time) does not support the step of saying that their traditional country surrounded Lake Torrens and therefore Lake Torrens is and has been within their traditional country from settlement and since. 216 As I observed at the outset of this analysis, the rulings his Honour made in his review of all this evidence must be considered in the light of the Adnyamathanha appellants' and Barngarla appellants' positions before him at trial, namely that the Kokatha appellants did not have traditional rights and interests to the land immediately to the west of Lake Torrens and they did. On that assumption, I consider all of these comments are entirely consistent with his Honour's rulings. Even if this assumption were not adopted and these comments are considered in light of the Adnyamathanha appellants' position in these appeals, that they held traditional rights and interests in the land immediately to the west of Lake Torrens prior to 2014, such that those traditional rights and interests should be taken into account in determining their claim to Lake Torrens, for the reasons I have expressed above (at [203]), I do not consider that contention is valid or correct. On either approach, I do not therefore consider that the comments his Honour made in this paragraph evidence any error. 217 Next, the Adnyamathanha appellants complained about the use his Honour made of the Kokatha Part A determination at [364]-[365]. Those paragraphs appear among the concluding paragraphs of the Place Names and Linguistic Identity section of the Reasons. They contain his Honour's conclusions with respect to that evidence as discussed earlier in these reasons. In the course of expressing those conclusions, his Honour made the following references to both the three consent determinations and the Kokatha Part A determination in particular: [at 364] The difficulty, in my view, with taking the linguistic evidence too far is the extent to which the pattern it is said to disclose is related more or less consistently to the western side of Lake Torrens (as well as to the areas to the south and north) in terms which point to a conclusion inconsistent with Kokatha Part A. [at 364] The pattern can, and does, show significant Barngarla presence and probably significant Kuyani presence in the areas around Lake Torrens. That that was the state of affairs accepted by Willis in any event, and was the general effect of the evidence. Having regard to the decision in Kokatha Part A the pattern cannot show that that presence demonstrated that the Barngarla (or Kuyani) People had at, and therefore since, sovereignty native title rights and interests under their traditional laws and customs in the area west of the western boundary of Lake Torrens. [at 365] Indeed, given that evidentiary pattern of word use and word derivation in the area to the west of Lake Torrens, and the Kokatha Part A determination in respect of that area, the suggested explanation for the absence of Kokatha word usage and word derivation in place names in the area, namely that in Western Desert societies such word usage or word derivation is kept confidential, appears the more likely. 218 In their submissions, the Adnyamathanha appellants made the following challenge to these conclusions: "The Judge was wrong to reject the linguistic placename evidence for 'pointing to a conclusion inconsistent with Kokatha Part A' (when it was, in fact, consistent)." They contended elsewhere that the words "inconsistent with Kokatha Part A" were ambiguous. This contention can be rejected briefly. Having regard to the context in which they were made as outlined above (at [192])-[197]), I consider his Honour's rulings were unambiguous in their terms. As to the Adnyamathanha appellants' main complaint, it is necessary to understand the context in which his Honour expressed these conclusions. I have already described the broader context above (at [213])-[214]) when considering his Honour's observations at [314]-[315]. With respect to the place names and linguistic evidence, that context is specifically described from [348]-[351] of the Reasons where his Honour summarised the opinions of Sutton and Fergie et al. In brief summary form, they were to the same effect as I have outlined above, namely that the Kokatha had no connection with the western side of Lake Torrens at "practical sovereignty" and the Adnyamathanha/Kuyani and the Barngarla most likely did have traditional rights and interests in that area (see the Reasons at [350]). 219 In this context, it can be seen that his Honour's comments in the paragraphs above were consistent with the fundamental premise of the Kokatha Part A determination which his Honour identified when making his rulings. The position may have been different if the effect of this evidence was not to mount a direct challenge to the Kokatha Part A determination. The difference may be subtle, but his Honour was in the best, perhaps the only, position to assess the effect of this evidence and whether the Adnyamathanha appellants and the Barngarla appellants were attempting to use it for a proper, or an improper, purpose. Viewed in this context, I do not consider his Honour erred in applying his rulings in any of these paragraphs. 220 In his Honour's lengthy review of the lay evidence (at [387]-[704]), the Adnyamathanha appellants have identified two parts where, they claim, his Honour misapplied his rulings. One relates to the evidence of their lay witness, Mr Michael McKenzie at [479] and the other relates to three paragraphs in the evidence of some of the Barngarla appellants' lay witnesses. 221 Mr McKenzie gave his evidence at Andamooka Island and at Port Augusta, partly in closed session (see at [475]). In his evidence, he said his father is an Adnyamathanha man and his mother is of Kokatha country (at [476]). This mixed heritage led his Honour to observe that (at [476]): "Like Lee Brady therefore, he has extensive genealogical tracks back to both the Adnyamathanha People and the Kokatha People. As I have remarked elsewhere, such shared genealogies make it somewhat hard to understand the vehemence of the competing Claim Groups." The reference to having "remarked elsewhere" would appear to include the comments his Honour made in the Interrelated Applicant Groups section of the Reasons (at [387]-[400]). In that section, his Honour recorded the peculiar circumstance that some of the lay witnesses called by the three appellants before him identified with, or traced their ancestry back to, more than one applicant group (see at [387]-[391]). He remarked that this unusual feature had the following consequences (at [394]): That was a particular feature of the evidence of traditional stories, where some witnesses were said to be have been told their stories by elders who knew those stories but who identified with another group, either in addition to their own group, or in some cases senior men and/or women who only identified with another group. 222 With respect to that peculiar feature, his Honour made the following remarks about the evidence of Mr Brady and Mr McKenzie that (at [395]): In addition some witnesses gave evidence that they had been initiated into the law in more than one group. Mick McKenzie and Lee Brady, who both gave evidence on behalf of the Adnyamathanha People, also described going through Western Desert law. Mick McKenzie is also a Western Desert-initiated wati who gave evidence of going through Western Desert law ceremonies. Lee Brady is also recognised as being a very senior Adnyamathanha person and a very senior Western Desert tjilpi with extensive knowledge of both Adnyamathanha and Western Desert law, including dealings with senior Kokatha men Robert Starkey and Mick Starkey. 223 While his Honour did not consider these matters "in and of themselves" were fatal to the cases of any one of the three appellants, he foreshadowed the care with which he intended to weigh the lay evidence concerned (at [399]) and concluded (at [400]) that: However, that degree of inter-relationship and clearly shared knowledge to a significant degree because of the shared sources of it means that the Court should be cautious above (sic - about) using evidence of such knowledge as supporting the claim of one of the three competing claim groups over the other. Secondly, it attracts the same comments as made earlier: the positive evidence given by one claim group may well have a negative significance to the interests of another claim group. That situation is confirmed by the closing written submissions of the other [Appellants]. It is not just the State or Kelaray which positively asserted in submissions that a particular application should fail; it is also each of the competing [Appellants]. 224 Clearly his Honour had this background in mind when he said of Mr McKenzie's evidence that (at [479]): To an extent, it is necessary to be cautious about giving weight to his evidence where it relates to the area to the west of Lake Torrens and is or may be inconsistent with the findings underlying, or the orders made, in Kokatha Part A. Those informants included Max Thomas. The fact that his evidence about Lake Torrens itself, and including Adnamooka (sic - Andamooka) Island, is in part sourced or learned from those who gave him history or stories not consistent with Kokatha Part A also causes me to pause before placing much weight on his evidence directly concerning the claim area. As has been noted earlier in these reasons, Max Thomas was also the primary source of the Starkey brothers' sacred and cultural knowledge about Lake Torrens (see, for example, the Reasons at [416], [420] and [429]). 225 The Adnyamathanha appellants have made two complaints about the statements in [479]. The first was that the reference to "inconsistent with … the Kokatha Part A" is ambiguous. While, in that reference to Kokatha Part A, his Honour did express the effect of his rulings slightly differently - "inconsistent with the findings underlying, or the orders made" - that statement unambiguously expresses the effect of his rulings about the weight he intended to apply to this kind of lay evidence (see at [194] above). The second complaint was that "[t]he Judge wrongly extended or applied [his] … rulings to matters not inconsistent with prior determinations as set out below". Thereafter, the Adnyamathanha appellants identified a number of paragraphs of the Reasons where this arose, but they did not specifically mention [479]. 226 It is true that it is difficult to discern what particular inconsistency his Honour was referring to in [479]. For confidentiality reasons, he was obviously deliberately guarded when summarising the effect of Mr McKenzie's evidence and the evidence of the many other witnesses (including the Starkey brothers) that gave rise to the inconsistency in question. Again, it is necessary to take account of the context in which his Honour made these comments. While they concern lay evidence, rather than expert evidence, as I have already noted above, the Adnyamathanha appellants and the Barngarla appellants also sought to mount the same direct challenge to the Kokatha Part A determination in their lay evidence and that necessitated his Honour making a ruling directed to that evidence (see at [194] above). It is therefore reasonable to assume that the aspect of Mr McKenzie's evidence to which his Honour referred in this paragraph fell into that category. 227 The primary judge was best placed to assess this evidence and to determine where it fell. He alone heard all the evidence, confidential and open, from all the various witnesses concerned. He heard the submissions from all the competing parties and was therefore fully informed about the dimensions of the dispute between them as it related to this evidence. He was well aware from his rulings where he had drawn the line. It is also apparent from elsewhere in his Reasons that he fully appreciated the distinction between the permissible and impermissible uses that may be made of the evidence, lay and expert. Having regard to all these aspects of the primary judge's advantage, I do not consider error can be discerned in the comments his Honour made in this paragraph, nor, for that matter, in any of the paragraphs about which the Adnyamathanha appellants have complained on this particular aspect. In all the circumstances I have outlined, I consider this Court should defer to the considerable advantage the primary judge possessed and accept that his Honour properly used it. I therefore reject the Adnyamathanha appellants' contention that the contents of this paragraph show that his Honour erred by misapplying his rulings. 228 The other concerns the Adnyamathanha appellants have about his Honour's review of the lay evidence relate to four paragraphs in the evidence of some of the Barngarla appellants' lay witnesses: Rosalie Richards (at [557]), Harry Dare (at [587]-[588]) and the statement of Lorraine Dare at [598]. Without setting out the contents of these paragraphs, it suffices to say this evidence addresses the same subject matter as the evidence of Mr McKenzie (above) which his Honour thought gave rise to an inconsistency with the Kokatha Part A determination. Obviously, his Honour assessed this evidence as being affected by that same inconsistency. For the reasons I have expressed above, I do not consider this Court is in a position to gainsay the considerable advantage the primary judge had in making that assessment. I do not therefore consider the contents of any of those paragraphs show that his Honour erred by misapplying his rulings. 229 In their final group of challenges under this alleged core error, the Adnyamathanha appellants' pointed to the use the primary judge made of the Kokatha Part A determination in various paragraphs of the final Consideration section of the Reasons. They were, in order: [709], [712], [715], [738], [742]-[744], [758] and [762]-[763]. Since [709] appears in that part of the final Consideration section where his Honour explained why he was not willing to draw an inference in favour of any of the appellants and that matter is the subject of the third core error, that paragraph will be included in the consideration of that error below. 230 Paragraphs [712] and [715] appear in that part of the final Consideration section where his Honour disposed of the Kokatha appellants' claims. Both of those paragraphs, in my view, merely restate the effect of the Kokatha Part A determination in terms of his Honour's rulings. As well as complaining about [738] as involving a misuse of the Kokatha Part A determination, the Adnyamathanha appellants' alleged that: "The Judge was wrong to find (at [738]) that only the Irti Vardnapa mura and Artunyi & Arkaru mura specifically related to the claim area. So too do the Bivu/Vidni Murunha, Wilka & Warratyi, Pungka Pudunha, Idhi, and Alda (Kalta) and frill-neck lizard mura." This statement does not, in my view, accurately state what his Honour said about the first two mura. His Honour did not say that those were the "only" mura, but rather that they were the only two "detailed" ones. This is made clear by his reference to the other more generic stories in the penultimate sentence. That aside, the Kokatha Part A determination is not expressly or implicitly mentioned in this paragraph. Otherwise the observations in [738] are entirely consistent with the principles his Honour outlined in The Law section of the Reasons, particularly the need for "continued connection required by s 223 and 225 of the NTA as prescribed in Ward HC." They do not therefore disclose any relevant error. 231 Finally, for the reasons I have already explained above, I consider all of the references to the Kokatha Part A determination (and all three consent determinations in the case of [763]) in [742]-[744], [758] and [762]-[763] accurately described the effect of a consent determination and/or the Kokatha Part A determination in particular. 232 Accordingly, none of the paragraphs in the final Consideration section of the Reasons about which the Adnyamathanha appellants complained shows any error by his Honour misapplying his rulings, or in misusing the effect of the Kokatha Part A determination or the three consent determinations generally. 233 For these reasons, I do not consider there is any merit in the Adnyamathanha appellants' second core error.

Error (c): failing to draw an inference 234 The third and final core error advanced by the Adnyamathanha appellants is that his Honour wrongly refused to draw inferences in their favour that were properly open to him. Six paragraphs of the Reasons were relied upon: [707]-[711] and [772]. 235 I have already dealt with this issue in considering the similar ground of appeal raised by the Kokatha appellants with respect to these paragraphs of the Reasons (see at [85]-[87] above). For the reasons there expressed, I do not consider his Honour erred in refusing to draw an inference in favour of the Adnyamathanha appellants, or indeed any of the other appellants. His conclusions on that issue are entirely consistent with the established principles I there outlined. For these reasons, the Adnyamathanha appellants' third core error is unmeritorious and must be rejected.

Other errors 236 The Adnyamathanha appellants' general response to the dispositive reasoning in the final Consideration section with respect to their claims was to challenge as "wrong" almost all of the conclusions his Honour reached. The difficulty with this approach is that almost all of those conclusions concerned in whole, or in part, findings of fact and the Adnyamathanha appellants have not explained why these findings involved any relevant error (see the discussion at [31] above). In particular, more than one third of the paragraphs containing these challenges took the form of bald assertions of error, or claimed that the primary judge had failed to give sufficient weight to evidence without any attempt being made to explain why those findings were wrong. For example: "The Judge was wrong to find (at [738]) that"; "The Judge was wrong to find (at [740], and contrary to [516]) that"; "The Judge at [747] gave insufficient weight to the appellant's evidence of customary use of the claim area"; "The Judge was wrong at [752] that"; "The Judge was wrong (at [751] and [753]) to criticise the appellant's lay evidence as fragmented or contradictory"; and so on. 237 On other occasions, the allegation of error was itself erroneous. For example, the Adnyamathanha appellants submitted that, at [774] (in the final Conclusion section), the primary judge attempted "to prioritise spiritual beliefs as between claim groups". On a fair reading of that paragraph of the Reasons, I do not consider his Honour did anything of the sort. Even on the few occasions where the Adnyamathanha appellants went further than the bald assertion of error, they did not come close to providing a satisfactory explanation as to how the error arose. An example of this deficiency is their submission that the primary judge was wrong, at [733], to find "that any at-Sovereignty connection had not been maintained" because, so they claimed, "the totality of the evidence was overwhelmingly persuasive that connection had been maintained". 238 For these reasons, I do not consider any of the Adnyamathanha appellants' challenges to the factual findings in the Reasons have any merit. That being so, they have not, on this aspect of their appeal, demonstrated any other error on the part of the primary judge.

Conclusion on the Adnyamathanha appellants' appeal 239 For the reasons set out above, none of the Adnyamathanha appellants' grounds of appeal, or core errors, has any merit and their notice of appeal must therefore be dismissed.

The primary judge's dispositive reasoning 240 As I have foreshadowed above, I will begin my consideration of the Barngarla appellants' grounds of appeal by summarising the dispositive reasoning with respect to their claims as contained in the final Consideration section of the Reasons. 241 The primary judge reached his final dispositive conclusion with respect to the Barngarla appellants' claims at the end of that Consideration section stating (at [769]) that: I am not satisfied that there is a proper foundation for the contemporary connection with that section of the Lake required for the determination of native title rights and interests in that section of the Lake in favour of the Barngarla People. 242 The reasoning used to reach that conclusion is essentially divided into two parts. The first concerns the primary judge's rejection of the attempt made by the Barngarla appellants in final submissions to recast their claims so that they related to the area south of the "Fergie Line" also described as the "fall back" line (at [755]-[759]). The second concerns his Honour's rejection of the Barngarla appellants' claims insofar as they related to the "south gap" area, which he thought was "more realistically" contiguous to the corresponding parts of the area covered by the Barngarla appellants' consent determination (at [760]-[769]). 243 With respect to the Fergie, or fall back lines, although noting the archaeological and agreed anthropological evidence as to "the historical use and likelihood of traditional rights and interests being held in Lake Torrens by an Aboriginal society, or societies", his Honour concluded at [756]: As with the Adnyamathanha claim, there is little evidence to support the assertion that, at the time of sovereignty, the Barngarla held rights and interests in that part of Lake Torrens (or indeed to the lower section of Lake Torrens where the "fall back" line position was drawn) in accordance with their traditional laws and customs ... 244 At [757]-[758], his Honour explained the origins of the fall back line and expressed his view that there was no support for it in the ethnographic evidence, or in the expert anthropological evidence, or in the contemporary Aboriginal evidence. 245 At [759], his Honour referred to the evidence he had earlier summarised under the heading Borders of Country (at [698]-[704]). Relevantly, as a part of that summary, his Honour summarised the difficulties he observed the Barngarla appellants' lay witnesses had in describing the borders of their country in the following terms (at [701]): Barngarla witnesses had difficulty in consistently effectively describing the borders of country. Eric Paige said that the line was either Carrapateena or Yeltacowie on the west, going directly east. He identified South Gap and Andamooka Island as the two sites of Barngarla significance that he knew about, but said he wouldn't draw the line from South Gap eastward because "that's not givin' us much then". Graham Richards said the line was around Beltana, and Parachilna on the east, bending southwards to "below Arcoona Station" on the west. Linda Dare claimed the area south of a line from Beltana on the east, but her evidence was not specific about the western side of the Lake. Eventually she said that her mother told her, "the Lake was ours.. It was Barngarla country". Rosalie Richards identified the Nilpena railway siding on the east to "probably a bit north of Andamooka" although she was not sure. And Amanda Richards said "around Beltana" on the east. She was not sure on the west, but said she thought it was north of Andamooka. Patricia Dare and Harry Dare said that the whole Lake was Barngarla country. 246 These observations led his Honour to conclude (at [759]) that that evidence "was not persuasive, as it appeared to be inconsistent, speculative and somewhat indecisive". 247 As to the "south gap" area, his Honour concluded (at [760]) that the evidence to support the Barngarla appellants' claims to that area was "relatively slight" and, in any event, he concluded (at [761]) that: I am not satisfied that the Barngarla People have maintained any connection they had to that part of the Lake in accordance with their traditional laws and customs in a continuous way substantially in accordance with their traditional laws and customs since sovereignty. 248 Then, at [762]-[769], his Honour summarised his reasons for reaching this conclusion. First, at [762], he identified two reasons why he did not consider he was able to draw an inference that would apply from the time of first European contact "for the preceding 60 years or so to sovereignty". Those reasons were the existence of Kokatha Part A, reinforced by the evidence of the eastward migration which he described earlier in the evidence and summarised as follows: … reinforced by the evidence about the eastwards tending movement of the Western Desert Cultural Bloc and the accepted western movement of the Barngarla (and Kuyani) People from the western side of Lake Torrens over that period and presumably subsequent decades. The observations I have made on this topic in relation to the Adnyamathanha claim apply equally to the issue of continuity concerning the Barngarla claim. 249 Next, at [763]-[764], his Honour noted that the expert anthropological evidence did not support the Barngarla appellants' claims. In expressing those views, his Honour noted Fergie's conclusion after a detailed and lengthy study of the ethnographic evidence on that question that: "Nonetheless, Fergie agreed that it is uncontroversial that the Barngarla evidence with respect to Lake Torrens is 'impoverished'." 250 Finally, at [765]-[769], his Honour referred to and explained why there was "little persuasive" lay evidence to support the necessary "continuous ongoing and contemporary connection with Lake Torrens, according to Barngarla traditional laws and customs". His Honour's summary of that evidence included: there was no Barngarla name offered for Lake Torrens (at [765]); such evidence as there was about Andamooka Island was inconsistent as to whether it was a men's only place, a women's area or both (at [765]); that there were "variations in the mythology known and described by different Barngarla witnesses … indicative of fractured and localised knowledge" (at [768]); and the absence in the ethnographic evidence until relatively recently of "any significant reference to Barngarla occupation" of the Lake in that area (at [769]). 251 From this summary of the dispositive reasoning in the final Consideration section of the Reasons, it can be seen that the Barngarla appellants failed on the evidence to establish a factual foundation for their claim. That was so because, as his Honour found, their claim was bereft of credible supporting evidence. He variously described their evidence as "impoverished", "not persuasive", "inconsistent", "speculative" and "somewhat indecisive" (see [759] and [763]). While his Honour found the witnesses who gave this evidence were truthful, that does not detract from the findings he made about its credibility. It is trite to observe that even truthful witnesses can give evidence that is unreliable or lacking in credibility.

The Barngarla appellants' grounds of appeal 252 The Barngarla appellants' grounds of appeal are set out above (at [10]). I have already disposed of grounds of appeal 2 and 3 relating to the misuse of the consent determinations when dealing with the similar issue raised by the Adnyamathanha appellants (see at [188]-[233] above). Similarly, I have disposed of the inference issue raised by ground of appeal 2(b) when considering the similar issue raised by the Kokatha appellants (see at [85]-[87] above). Putting aside the two matters which I have dealt with separately below, as to the balance of their grounds of appeal, in essence they are confined to challenging findings his Honour made which were, in my view, to a substantial degree based upon the credibility of their witnesses. In this respect, it is important to record that the decision in Warren v Coombes, upon which the Barngarla appellants placed reliance, does not stand for the proposition that this Court can simply ignore those credibility findings and set about drawing inferences from facts which they carefully selected from the Reasons. 253 In Fox v Percy, while reiterating the need for an appeal court to conduct a "real review", the High Court made it clear that the trilogy of cases that followed Warren v Coombes concerning the particular advantage a trial judge possesses in assessing the credibility of witnesses remained the instruction of the High Court to Australian appellate courts (see at [26]-[27]). In the terms of the last case in that trilogy, Devries v Australian National Railways Commission (1993) 177 CLR 472 at 479, that instruction was that: "If the trial judge's finding depends to any substantial degree on the credibility of the witness, the finding must stand unless it can be shown that the trial judge 'has failed to use or has palpably misused his advantage' or has acted on evidence which was 'inconsistent with facts incontrovertibly established by the evidence' or which was 'glaringly improbable'" (footnotes omitted). The Barngarla appellants have not come close to meeting any of these tests with respect to the factual findings challenged in these grounds of appeal. 254 There are two other matters raised by the Barngarla appellants' grounds of appeal, or their supporting submissions, that should be addressed. In their ground of appeal 1(b), there is a suggestion of an allegation that the primary judge failed to give reasons for his rejection of their claims. They have not identified a particular part of the Reasons where that occurred. Nonetheless, to the extent that it appears to relate to the issue identified in the introductory words of ground of appeal 1, I consider there is no validity in that claim. In the final Consideration section of the Reasons where his Honour explained why he rejected the Barngarla appellants' claims (outlined at [240]-[251] above), he has, in my view, more than adequately explained why he came to that conclusion. 255 The second matter appears in the Barngarla appellants' submissions in support of ground of appeal 2. On that issue, they sought to draw a distinction between a native title determination made following a contested hearing and a consent determination made under s 87 of the NTA. I have already touched on this issue above, but it warrants an express rejection here. A consent determination made under s 87, and complying with the apposite provisions of the NTA, is as much an approved determination of native title under the NTA as a determination made following a contested hearing. That is, it disposes once and for all of the fundamental matters decided by it (see the discussion above at [198]-[203]). 256 For these reasons, I do not consider any of the Barngarla appellants' grounds of appeal has any merit. Their notice of appeal must therefore be dismissed.

CONCLUSIONS AND ORDERS 257 For these reasons, I consider that each of these appeals fails and must, therefore, be dismissed. Accordingly, I would make the following orders:

The Kokatha amended notice of appeal filed 15 February 2017 is dismissed.
The Adnyamathanha notice of appeal filed 30 August 2016 is dismissed.
The Barngarla amended notice of appeal filed 27 September 2016 is dismissed. I certify that the preceding two hundred and fifty-seven (257) numbered paragraphs are a true copy of the Reasons for Judgment herein of the Honourable Justice Reeves.

The appeals 258 These are three appeals from orders dismissing three overlapping native title claimant applications in respect of Lake Torrens in South Australia. By their native title claimant applications the appellants each claimed native title rights and interests over the whole of Lake Torrens to the exclusion of the native title rights and interests claimed by the other appellants. The overlap in claimed rights and interests was thus complete. So too was the failure of the parties. 259 The appellants all contend that the primary judge's reasons for judgment in Lake Torrens Overlap Proceedings (No 3) [2016] FCA 899 disclose error in the consideration and resolution of their claim. 260 In the appeals the appellants sought orders not only that their appeals be allowed, but also that this Court make a native title determination under the Native Title Act 1993 (Cth) (the NTA) recognising their native title rights and interests in relation to Lake Torrens. Subsequently, they agreed that if one or more of their appeals succeeded on certain grounds, this Court would not be able to make a determination but would need to remit the matter for consideration by a single judge (the primary judge having retired shortly after the judgment was published). 261 I have concluded that the appeals should be allowed and the matters remitted to a single judge. Before explaining why this is so, it is appropriate to record that the task which confronted the primary judge was by no means straightforward. It is not apparent, moreover, that the task was made any easier by the parties' conflicting approaches to the many and varied issues which the primary judge had to resolve, including the admissibility of and weight that could be given to many aspects of the evidence. 262 The spelling of Aboriginal names and words in the evidence varied. I have used the spelling as it appears in the particular document or evidence or transcript.

The circumstances confronting the primary judge 263 The appellants are applicants for three competing native title claims in relation to Lake Torrens. One claim to the whole of Lake Torrens was filed on behalf of the Kokatha People. Another claim to the whole of Lake Torrens was filed on behalf of the Adnyamathanha People. A third claim to the whole of Lake Torrens was filed on behalf of the Barngarla People. 264 While overlapping claims under the NTA are not uncommon and, indeed, are expressly contemplated by s 67 of the NTA (which requires overlapping claims, to the extent of the overlap, to be dealt with in the one proceeding), in the present case each of the claimant groups had the benefit of a native title determination recognising their native title rights and interests in land immediately adjoining Lake Torrens. The Kokatha have native title rights and interests recognised in a determination which encompassed all of the land to the immediate west of Lake Torrens. The Adnyamathanha have native title rights and interests recognised in a determination which encompassed all of the land to the immediate east of Lake Torrens excluding a small area abutting the south-eastern corner of the lake. The Barngarla have native title rights and interests recognised in a determination which encompassed land to the south of Lake Torrens including to the boundary of the lake in the south-eastern corner not included in the determinations in favour of the Kokatha or Adnyamathanha Peoples. 265 As the primary judge noted at [3], apart from Andamooka Island, an habitable island which is located to the western side of Lake Torrens and separated from the land the subject of the Kokatha determination by a narrow causeway, Lake Torrens is a salt lake which occasionally fills with fresh water. When filled the lake attracts abundant wildfowl and other fauna (at [2]). When empty, which it is most of the time, it is unsuitable for long-term occupation, being a large flat expanse of salt (at [3]). 266 It was common ground that the Kokatha People, holding native title rights and interests to the immediate west of Lake Torrens, are an eastern group of the wider Western Desert society. Professor Peter Sutton, anthropologist, gave evidence in the hearing describing the wider Western Desert society as one in which people subscribe to a religious and mythological system that permeates relationships with landscape and to a body of law often referred to as Tjukurrpa, with a particular type of kinship system. It was also common ground that the Adnyamathanha and Barngarla Peoples, holding native title rights and interests to the immediate east and south of Lake Torrens, are part of the wider Lakes Group society which shares cultural laws and norms distinct from those of the Western Desert society. 267 At [98] the primary judge said this: The necessary connection must be shown in relation to Lake Torrens, or parts of it, notwithstanding its harsh physical features. And, moreover, it will not readily be inferred on any of these three Applications from the existence of adjoining native title rights at sovereignty that such connection, and therefore such rights, extended naturally into Lake Torrens because that inference (without more) would apply equally to the Kokatha People from the west and to the Adnyamathanha (or Kuyani) People from the east, although perhaps not so strongly to the Barngarla People from the south, except to a limited extent into the southern part of Lake Torrens. 268 While this was subject to challenge by the appellants, [98] represents nothing more than a formal statement of logic which is beyond dispute. The point being made was that, at least as between the Kokatha and Adnyamathanha Peoples, no inference could be drawn from their native title rights and interests in relation to the land adjoining the western and eastern boundaries respectively of Lake Torrens, as each such inference would be defeated by the other. This is necessarily correct given that both had native title rights and interests up to the boundary of the lake, those native title rights and interests related to the whole of the land on each side of the lake, and both claimed native title rights and interests over the whole of the lake exclusive of any other claimant. Given this, the primary judge's observation in [98] involves a logical necessity beyond reproach. It also said nothing about what inferences should otherwise be drawn on all of the evidence. 269 This point being made was also part of a general concern which the primary judge held, which re-appears in different guises throughout the judgment but no more clearly than in the reference to "cut-throat poker" at [47], that the parties' conduct of their cases involved risk. On the one hand, "the further the individual cases of the Applicants were advanced, the more there would be evidence directed also to showing that another Applicant, or Applicant group, did not have the native title rights which were asserted" (at [45]). On the other hand, "it could not be assumed that the Court could find that there are joint or co-existing rights held by some differently defined group or society or community over Lake Torrens" (at [45]). At best, there was "no reason why, depending on the evidence, the Court could not find in favour of one or more of the Applicants over a part or parts of Lake Torrens" (at [45]).

Adnyamathanha and Barngarla Peoples' appeals 270 The Adnyamathanha and Barngarla Peoples contend that the primary judge misused the native title determinations referred to above, particularly the determination in favour of the Kokatha People and, in so doing, disregarded or wrongly discounted the weight given to many aspects of their evidence. The State of South Australia and the Kokatha People deny this contention. 271 The native title determinations are each an "approved determination of native title" as referred to in s 253 of the NTA. An approved determination of native title, by s 13(3) of the NTA, is a determination by the Federal Court of native title made on an application under s 13(1) of that Act. Relevantly, under s 13(1)(a) applicants on behalf of the Kokatha People, the Adnyamathanha People and the Barngarla People had applied for "a determination of native title in relation to an area for which there is no approved determination of native title" as provided for in s 61(1) of the NTA. Their applications, from which Lake Torrens had been deferred as a result of the overlap, succeeded. Determinations under s 225 of the NTA were made consequential upon reasons for judgment in Starkey v State of South Australia [2014] FCA 924 in respect of the Kokatha People, Adnyamathanha No 1 Native Title Claim Group v South Australia (No 2) [2009] FCA 359 in respect of the Adnyamathanha People, and Croft v State of South Australia [2015] FCA 9; (2016) 325 ALR 213 and Croft on behalf of the Barngarla Native Title Claim Group v State of South Australia (No 2) [2016] FCA 724 in respect of the Barngarla People. 272 Section 225 of the NTA provides that: A determination of native title is a determination whether or not native title exists in relation to a particular area (the determination area) of land or waters and, if it does exist, a determination of: (a) who the persons, or each group of persons, holding the common or group rights comprising the native title are; and (b) the nature and extent of the native title rights and interests in relation to the determination area; and (c) the nature and extent of any other interests in relation to the determination area; and (d) the relationship between the rights and interests in paragraphs (b) and (c) (taking into account the effect of this Act); and (e) to the extent that the land or waters in the determination area are not covered by a non-exclusive agricultural lease or a non-exclusive pastoral lease-whether the native title rights and interests confer possession, occupation, use and enjoyment of that land or waters on the native title holders to the exclusion of all others. 273 By s 223 of the NTA: (1) The expression native title or native title rights and interests means the communal, group or individual rights and interests of Aboriginal peoples or Torres Strait Islanders in relation to land or waters, where: (a) the rights and interests are possessed under the traditional laws acknowledged, and the traditional customs observed, by the Aboriginal peoples or Torres Strait Islanders; and (b) the Aboriginal peoples or Torres Strait Islanders, by those laws and customs, have a connection with the land or waters; and (c) the rights and interests are recognised by the common law of Australia. 274 In accordance with s 225 of the NTA, each determination determines the persons holding the common or group rights comprising the native title in the determination area (relevantly, the Kokatha People in relation to the Kokatha determination area, the Adnyamathanha People in relation to the Adnyamathanha determination area, and the Barngarla People in relation to the Barngarla determination area). It also determines the nature and extent of the native title rights and interests in relation to the determination area (relevantly, only the Kokatha People were determined to have native title rights and interests in relation to the Kokatha determination area, only the Adnyamathanha People were determined to have native title rights and interests in relation to the Adnyamathanha determination area, and only the Barngarla People were determined to have native title rights and interests in relation to the Barngarla determination area). The determinations so determined each of the other matters required by s 225. 275 A little time was wasted during the appeals by speculation that the fact that each of the determinations was expressed to be "non-exclusive" might mean that the determinations left open the possibility that some Aboriginal peoples other than the Kokatha, Adnyamathanha and Barngarla peoples might have native title rights and interests in the respective determination areas of these Peoples either capable or incapable of recognition under the NTA. 276 As to the first possibility, s 61A(1) of the NTA provides that a native title determination application must not be made in relation to an area for which there is an approved determination of native title. Accordingly, it is clear that the non-exclusive nature of the native title rights and interests recognised in the determinations is because there are other non-native title rights and interests, rights and interests also recognised as provided for in s 225(c) of the NTA. 277 As to the second possibility, the concepts of "native title" and "native title rights and interests" are constructs created by the NTA. By the terms of s 223(1) they are constructs which depend on two components. The first component, by s 223(1)(a) and (b), depends on rights and interests "possessed under the traditional laws acknowledged, and the traditional customs observed, by the Aboriginal peoples or Torres Strait Islanders". The second component depends on the recognition of those rights and interests by the common law of Australia. As there have already been determinations of native title which recognise the rights and interests possessed under the traditional laws acknowledged, and the traditional customs observed, by the Kokatha People in relation to the Kokatha determination area, the Adnyamathanha People in relation to the Adnyamathanha determination area, and the Barngarla People in relation to the Barngarla determination area, there is no scope for the common law of Australia to recognise any other native title rights and interests in those determination areas. 278 The possibility which is left open by the determinations, however, is that before the determinations were made it may or may not have been an historical fact that Aboriginal peoples, other than the Kokatha People in relation to the Kokatha determination area, the Adnyamathanha People in relation to the Adnyamathanha determination area, and the Barngarla People in relation to the Barngarla determination area, possessed rights and interests under the traditional laws acknowledged, and the traditional customs observed, by those Aboriginal peoples which, by those laws and customs, had a connection with the land or waters of the determination areas. Those rights and interests, moreover, might have been capable of being recognised by the common law of Australia. This possibility involves a number of elements which depend on the characteristics of a determination under s 225 of the NTA. 279 It may be accepted that each determination recognises the native title of the Kokatha People in relation to the Kokatha determination area, the Adnyamathanha People in relation to the Adnyamathanha determination area, and the Barngarla People in relation to the Barngarla determination area. Of necessity given the common law of Australia, the determination thereby recognises that from sovereignty and until the making of the determination the Kokatha People in relation to the Kokatha determination area, the Adnyamathanha People in relation to the Adnyamathanha determination area, and the Barngarla People in relation to the Barngarla determination area, possessed rights and interests under their respective traditional laws and customs by which those people had, and as at the date of the determination continued to have, a connection to their respective determination areas. In this respect, the determination determines as a juridical fact a past (from sovereignty) to present (to the date of the determination) state of affairs. 280 It may also be accepted that each determination recognises that no Aboriginal peoples other than the Kokatha People in relation to the Kokatha determination area, the Adnyamathanha People in relation to the Adnyamathanha determination area, and the Barngarla People in relation to the Barngarla determination area possessed rights and interests under their traditional laws and customs by which those people had a connection to any of the determination areas. In this respect, however, the determinations determine as a juridical fact only a present and future (on and from the date of determination) state of affairs. They say nothing about the possibility, as a matter of historical fact, of rights and interests of any other Aboriginal people under traditional laws and customs by which those people had a connection to any of the determination areas before the date on which each determination was made. No doubt if a determination recognises the rights and interests of only one group of Aboriginal people then any possible rights and interests under traditional laws and customs of other Aboriginal people in relation to land within the determination area must have been found not to have continued to exist at some time before the date of the determination. But the necessary effect of the determination under the NTA is only that as at the determination date no other Aboriginal people had rights or interests under traditional laws and customs in relation to the determination area. 281 It is this possibility, of the Adnyamathanha People and the Barngarla People having had rights and interests under their traditional laws and customs by which those people had a connection to the Kokatha determination area before the Kokatha determination was made, which the Adnyamathanha and the Barngarla Peoples said the primary judge wrongly discounted by reason of the Kokatha determination. This, said the Adnyamathanha and the Barngarla Peoples, was in error because the Kokatha determination did not and could not say anything about the rights and interests of the Adnyamathanha and the Barngarla Peoples under their traditional laws and customs by which those people had a connection to the Kokatha determination area before the Kokatha determination was made. The Kokatha determination determined only that: (a) From sovereignty to the date of the determination the Kokatha People possessed rights and interests under the traditional laws acknowledged, and the traditional customs observed, by the Kokatha People and, by those laws and customs, the Kokatha People had and have a connection with the determination area. (b) As at and from the date of determination no Aboriginal people other than the Kokatha People possessed rights and interests under traditional laws acknowledged, and the traditional customs observed, by Aboriginal peoples which, by those laws and customs, have a connection with the determination area. 282 Consistent with these propositions, it was fundamental to the cases of the Adnyamathanha and the Barngarla Peoples that their evidence of rights and interests under their traditional laws and customs by which they had a connection to the Kokatha determination area before the Kokatha determination was made had to be weighed along with all other evidence to determine their claims to native title to Lake Torrens. It could not be disregarded or devalued because of the Kokatha determination. On their cases, if the Adnyamathanha and the Barngarla Peoples could establish, as a matter of historical fact, the existence of their traditional laws and customs under which they had rights and interests in relation to the Kokatha determination area before the Kokatha determination was made this would support their current claim to native title in relation to Lake Torrens because they would have proved connection under traditional laws and customs to both the east and the west of Lake Torrens, it being common ground between the anthropologists that Lake Torrens itself must have been the subject of traditional rights (even if such rights were shared between Aboriginal peoples). 283 The Adnyamathanha and the Barngarla Peoples were permitted to adduce evidence to this effect by the primary judge (some over objection by the Kokatha People) but, they said, his Honour wrongly discounted this evidence in whole or in material part because he assumed or considered that the Kokatha determination established that no Aboriginal people other than the Kokatha People from sovereignty had any rights or interests under traditional laws and customs by which they were connected to the Kokatha determination area. This was in error because it elided the past, present and future effects of a determination. 284 The past effect was a determination that from sovereignty to the date of the determination the Kokatha People possessed rights and interests under the traditional laws acknowledged, and the traditional customs observed, by the Kokatha People and, by those laws and customs, the Kokatha People had and have a connection with the determination area. 285 The present and future effect was a determination that from the date of determination no Aboriginal people other than the Kokatha People possess rights and interests under traditional laws acknowledged, and the traditional customs observed, by Aboriginal peoples which, by those laws and customs, have a connection with the determination area. 286 However, there was no past effect of a determination that from sovereignty to immediately before the date of determination no Aboriginal people other than the Kokatha People possessed rights and interests under traditional laws acknowledged, and the traditional customs observed, by Aboriginal peoples which, by those laws and customs, have a connection with the determination area. 287 According to the Adnyamathanha and the Barngarla Peoples, while the primary judge, in some parts of the reasons for judgment, expressed the principles about the way in which a determination operates correctly, in other parts, particularly when he came to deal with the evidence and the weight which it should be given or inferences which should be drawn from it, he proceeded on the incorrect basis that the Kokatha determination meant that he was bound to accept that from sovereignty to the date of determination no Aboriginal people other than the Kokatha People possessed rights and interests under traditional laws acknowledged, and the traditional customs observed, by Aboriginal peoples which, by those laws and customs, have a connection with the determination area. The determination did not have this effect. Rather, according to the Adnyamathanha and the Barngarla Peoples, whether they had proved such a connection or not was required to be assessed in the usual course having regard to the whole of the evidence. If, by that process, it was found that the Adnyamathanha and the Barngarla Peoples did have a connection under their traditional laws and customs to the Kokatha determination area from sovereignty until some time before the Kokatha determination was made, then that fact would itself be relevant to the assessment of their claim to the immediate adjoining land, Lake Torrens, particularly when weighed along with their own determinations in relation to the land immediately to the east and south of Lake Torrens. This opportunity, however, was said to be denied to them by a process of reasoning which miscarried. 288 There was an unwarranted focus in the submissions on the effect of a determination under s 225 of the NTA being in rem. I am unable to see that the characterisation of a determination under s 225 as operating in rem advances this argument. There is no doubt that a determination under s 225 of the NTA binds the world at large and does not operate only as between the parties to the litigation (CG (dec'd) (on behalf of Badimia People) v Western Australia [2016] FCAFC 67; (2016) 240 FCR 466 at [44] and the cases cited therein). The question in the present case is, what is it that a determination determines? This is not answered by recourse to the law in respect of orders which operate in rem rather than in personam. It is answered by the NTA, in particular s 223. Section 223 supports the contention of the Adnyamathanha and the Barngarla Peoples about the effect of a determination under s 225. For the Kokatha determination the Kokatha People necessarily established that they possessed rights and interests under the traditional laws acknowledged, and the traditional customs observed, by the Kokatha People and that, by those laws and customs, the Kokatha People had and have a connection with the determination area. For the laws to be "traditional", moreover, they had to exist pre-sovereignty and continue thereafter (Members of the Yorta Yorta Aboriginal Community v Victoria [2002] HCA 58; (2002) 214 CLR 422 at [33], [46] and [47]). The Kokatha determination itself also necessarily established that as at and from the date of determination no Aboriginal people other than the Kokatha People could claim any native title in the Kokatha determination area (subject only to the capacity for an application to be made to revoke or vary an approved determination of native title as provided for in s 13(5) which, under s 61(1) may be made by a limited class of persons, not including a native title claim group). The Kokatha determination did not establish, however, that the Adnyamathanha and the Barngarla Peoples did not have rights and interests under their traditional laws and customs by which they had a connection with the Kokatha determination area pre-sovereignty or at any time thereafter until the date of the determination itself. 289 The Adnyamathanha and the Barngarla Peoples (as well as the State and the Kokatha People) are right to identify various parts of the primary judge's reasons which disclose that his Honour was well aware of the effect of a determination under s 225. Given the primary judge's extensive experience in matters arising under the NTA, this is not unexpected. Accordingly, the primary judge said this, which is beyond dispute: 8 The determination in Kokatha Part A recognised the Kokatha People as the holders of native title over that claim area, running extensively west from the western borders of Lake Torrens, subject to those parts of the claim area where other acts have extinguished native title, either wholly or in part (including, by agreement the areas the subject of the ODA agreement registered as an ILUA under the NTA, and as explained in [32]-[37] later in these reasons). The recognised and extensive native title rights recognised are non-exclusive, because of the other interests in the claim area as recorded. The other interests do not include interests of any Adnyamathanha or Barngarla People except as recognised in the ODA agreement and as noted the area of the ODA agreement is excluded from the claim area. It is well to the west of Lake Torrens. The prescribed body corporate under s 57(2) of the NTA is the Kokatha Aboriginal Corporation. 9 The determination in Adnyamathanha No 1 recognised the Adnyamathanha People as the holders of native title over three claim areas, including relevantly the areas immediately to the east of Lake Torrens. Again, the rights recognised are extensive, and subject to the areas where other acts have extinguished native title either wholly or partially. The rights are non-exclusive other than four areas where, by operation of s 47A of the NTA, the extinguishment is to be disregarded. These areas of exclusive native title rights are not directly relevant to the present issues. There are no other terms of that determination which are said to inform the resolution of the present issues. 10 The determination in Barngarla No 2 recognised the Barngarla People as the holders of native title over that claim area, including relevantly the areas immediately to the south of Lake Torrens. Again, the rights are extensive but by reason of other acts those rights have been extinguished either wholly or partially, so the rights are relevantly to the present issues non-exclusive. There are no other terms of that determination which would inform the resolution of the present issues. 11 Obviously, at sovereignty, the native title rights and interests so recognised would not have been extinguished either in whole or in part by the acts which have occurred subsequent to sovereignty. 290 This orthodoxy is also apparent in the primary judge's observations at [52] and [57]: 52 It was common ground as between the parties, that no expert proposed to give evidence directly to attack any finding of fact which is fundamental to or underlies any existing determination and that the focus of the experts' reports was on who, if anyone, occupies or is entitled to native title rights over Lake Torrens. That is, as the Court expressed it in the course of the hearing, the parties accepted the primary findings of fact necessary to support the three determinations on the areas of land surrounding Lake Torrens (as outlined in Kokatha Part A, Adnyamathanha No 1 and Barngarla) and, of course, the three determinations in Kokatha Part A, Adnyamathanha No 1, and Barngarla No 2 handed down on 23 June 2016. … 57 Similar submissions were also made in respect of certain lay evidence, to the extent that it was said to contradict the findings of the Court in Kokatha No 1, Adnyamathanha No 1 and Barngarla. That evidence was also received on the basis that no weight would be given to evidence that was inconsistent with those determinations. 291 All parties to the appeal accepted that the primary judge's approach, as recorded in these paragraphs, was appropriate. The primary judge did not have before him an application to revoke or vary an approved determination of native title. As such, he was bound to accept that the determinations existed. As explained above, he was thus bound to accept the facts necessary to support each determination, which are those made necessary by the terms of s 223 of the NTA. 292 The same point may be made about the primary judge's statements as follows: 167 It is not disputed that each of these native title claim groups satisfy the statutory test laid down in s 223(1) of the NTA. That is, each group acknowledges and observes a body of traditional laws and customs under which that group possesses rights and interests in and had a connection with, an area of land and waters. That has been the subject of the three determinations referred to earlier in these reasons. 168 The issue in this proceeding is whether all of, or any part of, the rights and interests possessed by any of the three groups, extend to and include all of, or any part of, Lake Torrens, including Andamooka Island, at sovereignty and have since been maintained and exercised by the successors to the relevant group or groups to the present time, albeit appropriately adapted, so as to satisfy s 223(1)(b) requiring the present connection to Lake Torrens (or a part or parts of it) under those traditional laws and customs. There is no issue concerning s 223(1)(c). 293 Sections 223(1)(a) and (c) could not be in dispute because the groups in question were the same as those which had been found by the determinations to have rights and interests possessed under the traditional laws acknowledged, and the traditional customs observed, by them, which were recognised by the common law of Australia. As the primary judge recorded at [12]: While the named applicants who comprise each of the claimant groups are no longer identical to the applicants who were authorised by the claimant groups and in whose favour the determinations were made in Kokatha Part A, Adnyamathanha No 1 and Barngarla No 2, they identify as being part of the same societies as those recognised as the holders of native title rights and interests under the NTA in Kokatha Part A, Adnyamathanha No 1 and Barngarla, respectively. 294 The issue in dispute arose under s 223(1)(b) which required each group to establish that they "by those laws and customs, have a connection with the land or waters" the subject of their competing claims (that is, to Lake Torrens including Andamooka Island). 295 At [169] the primary judge recorded certain facts which had been agreed between the Kokatha People, the Adnyamathanha People and the State. These facts included the following: Adnyamathanha 1. The Adnyamathanha people, and its society, are as identified in the determinations of native title made by [the] Federal Court of Australia in March 2009 and February 2014, and are a relevant society for the purposes of s 223 of the Native Title Act 1993. 2. The Adnyamathanha comprise various traditionally closely related groups, including the Kuyani, Pirlatapa, Wailpi and Yadliyawara. 3. The Adnyamathanha people acknowledge and observe traditional laws and customs in the Adnyamathanha people, which are rooted in those normative rules which existed prior to the assertion of sovereignty by the British Crown. Those traditional laws and customs have a continuous and unbroken existence from sovereignty until the present day, and satisfy the requirement for such traditional laws and customs in s 223 of the Native Title Act 1993. 4. Key features of the Adnyamathanha society are the division of the Adnyamathanha people into two traditional matrilineal moieties, Matheri and Arraru; and belief in the Mura (in relation to areas covered by the determinations of native title made by the Federal Court of Australia in March 2009 and February 2014), for which the Adnyamathanha have responsibility. Kokatha 5. The Kokatha people are a recognisable group forming part of the wider Western Desert society. 6. The Western Desert society has continued to exist from sovereignty through to the present day as a body united in and by its acknowledgement and observance of a body of traditional laws and customs under which the member[s] of that society possess rights and interests in and have a connection with land and waters. 7. The Kokatha people possess rights and interests in and have a connection with, the land and waters immediately to the west of the claim area under the traditional laws and customs of the Western Desert society which the Kokatha people continue to acknowledge and observe. 8. The non-exclusive rights and interests possessed by the Kokatha people in relation to the land and waters immediately to the west of the claim area are as set out in paragraph 8 of the determination made by Allsop CJ in Starkey v State of South Australia [2014] FCA 924. 9. The Kokatha people include the descendants of the named ancestors listed in paragraph 7(a) of the determination in Starkey v State of South Australia [2014] FCA 924. 296 Thereafter, until [202], the primary judge dealt with the determinations and related objections to evidence in a manner which the State and Kokatha People contend discloses that his Honour did not misunderstand the effect of or misuse the determinations. Thus, at [171] the primary judge said: To the extent that evidence is sought to be led which is directly inconsistent with the necessary findings in Kokatha Part A, I have not accorded it any weight. 297 The primary judge recorded the State's submissions that: 183 The State submitted that the effect of a determination in rem precludes a party in another proceeding putting in issue the ultimate findings or ultimate facts determined in the in rem determination. It submitted that any evidence adduced for the purposes of identifying indigenous groups with a particular interest in an area at sovereignty can only be used for the purposes of identifying indigenous groups with native title in Lake Torrens at sovereignty, that is only for that confined or different purpose, but not so as to challenge any fact underlying an existing in rem determination. … 185 Consequently, the State said that it is permissible to adduce evidence in this proceeding which identified the Kuyani People as having been located on the north western area of land adjacent to Lake Torrens for the limited basis of establishing the Kuyani's potential connection to the Lake. However, it is impermissible to adduce that same evidence for the purpose of contending, or otherwise controverting the finding in Kokatha Part A, that the Kokatha People were not in exclusive possession of the Kokatha Part A area. It is noted that, even if the Court found that there were shared rights in country at sovereignty, if the persons said to have the benefit of those "shared rights" no longer asserted those rights, then those with the remaining rights must be held to have held them exclusively from sovereignty: see Banjima FC [Banjima People v State of Western Australia [2015] 231 FCA 456] at [48]-[55]. 298 While the Adnyamathanha and the Barngarla Peoples argue that error had crept in to that part of [185] in which the primary judge referred to the Kokatha People's "exclusive possession of the Kokatha Part A area", I disagree. The primary judge was merely repeating a submission. Read fairly in the context of the whole of the State's submissions, the State meant only that the Kokatha were the only Aboriginal people who, from the date of determination onwards, had established native title rights and interests in relation to the Kokatha determination area. This correctly reflects the effect of the Kokatha determination. In any event, that the primary judge correctly understood the effect of the submission is apparent from his subsequent observations in these terms: 186 In my view, having regard to the approach of the Court in Banjima FC, a finding that at sovereignty more than one group, and particularly more than one of the Applicant groups, might have held some rights and interests in the land which has now been determined to be the land of the Kokatha, Adnyamathanha or Barngarla People respectively would not necessarily be inconsistent with the determinations in Kokatha Part A, Adnyamathanha No 1 or Barngarla. To the extent that the ethno-historical record supports a finding that at sovereignty any applicant group or groups were located on both the eastern and western side of Lake Torrens (or a portion of the eastern and western sides of Lake Torrens), that is said to identify the claim area (or a portion of it) as lying within the territory of that group. 187 Indeed, in their closing submissions, the First Applicant accepted that the Kuyani and Barngarla People, along with the Kokatha People, may have been present in some parts of the area west of the western shoreline of Lake Torrens prior to European settlement but that subsequently the Barngarla and Kuyani left those lands and ceased to have any continuing connection with that area. They accepted that such a factual scenario would be consistent with the determination and the findings in Kokatha Part A that only the Kokatha People now possess native title rights and interests in the land that is contiguous with the western shoreline of the Lake. 188 That acknowledgment is consistent with the evidence given by Willis. Willis accepted that, at effective sovereignty, in addition to the Kokatha, the Kuyani were clearly on the western side of the Lake. As noted above, he said that progressively through to the end of the 19th Century and the beginning of the 20th Century the Kokatha presence was growing at the same time that the Kuyani and the Barngarla presence was retreating and that by 1940 there were almost no Aboriginal people other than the Kokatha on the western side of the Lake. 189 However, I accept the contention of the First Applicant that certain aspects of the expert anthropological evidence referred to does not merely assert that there were shared rights as between the Kokatha and the Kuyani at sovereignty, but rather it asserts that native title rights and interests in the area immediately to the west of Lake Torrens at the time of sovereignty were held by only the Kuyani or the Barngarla, and it also says that at sovereignty most, if not all, Kokatha People resided or were widely thought to be associated in some way with country further to the west, and not with areas immediately to the west of Lake Torrens. From that premise, it is said that it is unlikely that the eastern border of Kokatha country extended west into Lake Torrens at the time of sovereignty, because the Kokatha People were not at sovereignty in the western side of Lake Torrens. In my view, such a premise is not consistent with the decision in Kokatha Part A. 190 The premises upon which expert opinion evidence is based are fundamental, as expert evidence "is only as helpful as the evidence and assumptions on which it is based": Anikin v Sierrai (2004) 79 ALJR 452 at [28]. Consequently, to the extent that the expert anthropological views are premised upon the Kokatha People not having native title rights and interests in the area immediately to the west of Lake Torrens at sovereignty, I do not place weight on it. That is not to question the scholarship and integrity of any of the expert anthropologists. But, as a matter of record, the premise referred to is fundamental to the Kokatha Part A determination, and the Court must proceed on the basis of it. Nor is it to be critical in any way of those who supported the determination in Kokatha Part A. There was clearly an appropriate body of information, including expert anthropological information, which enabled the State with its responsibilities on behalf of the community to support that determination. Clearly, too, that information was sufficient to satisfy those other Aboriginal persons or groups who had become respondents to the then Kokatha Uwankara Claim to withdraw any objections they had to the determination in Kokatha Part A, that is to give recognition to the Kokatha People as the holders of the native title rights and interests over the area immediately to the west of Lake Torrens. 299 It was argued for the Adnyamathanha and the Barngarla Peoples that his Honour must be understood to have mistakenly considered that the Kokatha determination meant that Kokatha occupied the Kokatha determination area when the determination meant only that they had established continuing (from sovereignty) native title rights and interests in relation to that area as specified in s 223 of the NTA. I do not perceive these paragraphs as disclosing any confusion in the primary judge's mind between the concepts of native title rights and interests (which may not involve the occupation of land in the Western sense of that term) and of occupation. The primary judge refers in [189] only to the premise of the Kokatha not being "in" the land to the west of Lake Torrens as being inconsistent with the Kokatha determination. In this part of his reasons, it seems to me the primary judge was dealing with the concept of mere physical "presence" and not occupation. 300 This is reinforced by the fact that the primary judge was dealing with anthropological evidence from Professor Sutton that the land to the west of Lake Torrens, the subject of the Kokatha determination, was in fact and remains Kuyani (a group of the Adnyamathanha) and Barngarla country. Professor Sutton, as the primary judge recorded at [178], considered that the Kokatha People's post-sovereignty work history on pastoral lease areas had led them to claim connection with the area to the immediate west of Lake Torrens when this connection did not exist before or as at sovereignty. As such, Professor Sutton saw the westward expansion of Kokatha connection as "attempted annexation". It is this evidence the primary judge, at [189] and [190], considered could be given no weight because it was inconsistent with the Kokatha determination. 301 It is not necessary to decide the appeals of the Adnyamathanha and the Barngarla Peoples on the basis of what was said at [189] and [190], but I accept that it is possible these parts of the primary judge's reasons go too far. The primary judge must be understood as having decided that no weight could be given to Professor Sutton's opinions as recorded at [178] because they were inconsistent with the Kokatha determination. The main thing which Professor Sutton said which was inconsistent with the Kokatha determination is that the area of the Kokatha determination to the immediate west of Lake Torrens still remains Kuyani and Barngarla country. By reason of the Kokatha determination, at least insofar as native title rights and interest are concerned, that land is Kokatha country. By "Kokatha country" all that is meant is that, as the determination recognised, the Kokatha People had maintained a pre-sovereignty connection under their traditional laws and customs with that land from which they derived rights and interests in relation to that land. This did not mean, however, that the land was not also Adnyamathanha and Barngarla country at sovereignty or that the land did not continue to be Adnyamathanha and Barngarla country until some time before the Kokatha determination. Nor did the Kokatha determination make the pastoral history of the area or the work history of Kokatha People irrelevant or inappropriate to be given weight, to the extent that history could inform the existence or otherwise of a pre-sovereignty connection to Lake Torrens which had continued. The Kokatha determination did not mean that such evidence from Professor Sutton should or could be disregarded or discounted. 302 Nevertheless, if nothing more had been said than what appears in these paragraphs I would have rejected the challenges to his Honour's process of reasoning because, despite the suggestions at [189] and [190] that his Honour might have taken a broader view of inconsistency with a determination than justified by the terms of s 223 of the NTA, the subsequent observations at [191] represent a clear return to orthodoxy. The primary judge said this: There was substantial accord between the expert anthropologists on a range of related facts. They all broadly accepted that: (1) at some time at or before sovereignty there was a large population of Kuyani or Barngarla People on the western side of Lake Torrens (and, I add, also Kokatha People - although not all the experts accepted that); (2) the historical record demonstrates a "migration" of Western Desert Peoples in the period up to the 1880s, coming from the north-west, into Mount Eba and other nearby locations and subsequently, more intensely into the area west of Lake Torrens; and that (3) by the 1940s Western Desert people had come to numerically predominate in the area west of Lake Torrens. There is no reason not to place any weight on the collective position of the anthropologists on those matters. 303 The Adnyamathanha and the Barngarla Peoples may appear to be on firmer ground in respect of [192] in which the primary judge said this: Hence, the key differences between the experts in relation to the area to the west of Lake Torrens are principally twofold. The first is the times when the presence of the Kuyani or Barngarla People decreased in that area, and the times when the Kokatha People's presence in that area increased. I have ruled that it is a determined fact that the Kokatha People occupied, and held native title rights in, the area immediately to the west of Lake Torrens at sovereignty. The second is the difference between the experts as to what the increased presence of the Kokatha People to the west of the Lake signified, that is was it representative of increased usage by the Kokatha People of lands and waters that were, and had always been, Kokatha country, or did it signify a case of migration by the Kokatha People into an area post sovereignty which had previously been Lakes Group country. 304 The error is again said to be the reference to the Kokatha having "occupied" the Kokatha determination area when this is not the effect of the Kokatha determination which concerns only the existence of native title rights and interests in relation to the area, not occupation of the area. This reference is in error but, standing alone, it cannot be characterised as material. The primary judge knew that he was dealing with native title rights and interests, not occupation. At [95] he recorded the following, which is inconsistent with the notion that the primary judge was wrongly searching for evidence of occupation rather than connection under traditional laws and customs: Where evidence of Aboriginal presence at sovereignty is unavailable (as is commonly the case), such an inference is more readily drawn where there is a preponderance and weight of available evidence and in the absence of contradictory evidence: see e.g. Yarmirr v Northern Territory (1998) 82 FCR 533. Where it is appropriate to do so, weight should be accorded not only to physical use of the claim area but also to evidence of spiritual connection: e.g. see Griffiths v Northern Territory (2007) 165 FCR 391; [2007] FCAFC 178 (French, Branson and Sundberg JJ) at [127] where spiritual sanctions visited upon unauthorised entry supported a finding of exclusive possession. 305 The inference I draw, that the reference to "occupation" in [192] is a stray word of no significance to the primary judge's reasoning process, is supported by [193] in which the primary judge said: As any evidence or submissions in relation to migration must be considered in accordance with the determination in Kokatha Part A, that at least by sovereignty, the Kokatha People had rights and interests in that country, to the extent that the dates adopted by an expert or experts indicate that the Kokatha People annexed the land to the west of Lake Torrens only post-sovereignty, or are otherwise inconsistent with the findings in Kokatha Part A, no weight has been placed on that evidence. I have also considered the expert opinion evidence which relies in part on those dates, with particular caution. If the necessary premise for an expert opinion is not consistent with the determination in Kokatha Part A, the opinion itself cannot be given weight. 306 Subject to the note of caution given above about the extent of any inconsistency which a determination under s 223 of the NTA is capable of creating, there is no error apparent in this paragraph. 307 There is an important observation at [194] in these terms: As the reasons below disclose, I have placed significant weight on the underlying ethnographic material, where that material is available either in its original form, or in the relevant expert reports where the contents of that material is described in detail by the anthropologists. 308 This is important because it indicates that if the primary judge did make an error when dealing with the ethnographic material the error may be material because he placed significant weight on his interpretation of the material. I do not consider that his Honour was bound to adopt the anthropologists' interpretation of the ethnographic material (which was suggested by the Adnyamathanha and the Barngarla Peoples) in preference to his own interpretation, but in applying his own interpretation he was bound to give effect to the determinations only to the extent required by s 223 of the NTA. 309 At [205] the primary judge said: The challenge which confronts the Court, in the light of those findings, is how the ethnography is capable of properly informing the "at sovereignty" position in the areas surrounding the Lake beyond the findings in Kokatha Part A, Adnyamathanha No 1 and Barngarla (and there is some dispute about what it does, in fact, demonstrate), especially where it may in part not be consistent with those findings. And then the further step must be addressed as to how that evidence may be properly used to inform findings in respect of Lake Torrens itself. 310 This was said to involve error in that the ethnographic material consistently placed the Kokatha People further to the west of Lake Torrens than the Kokatha determination area until the 20th century. However, the error, if it exists, is nothing more than a misreporting of the ethnographic evidence. The primary judge was right in [203] when he said that the ethnographic record relates to the land around Lake Torrens rather than to the Lake itself. He was also right in [203] when he said the claim areas, apart from Andamooka Island, were not suitable for long-term occupation, a fact which the primary judge appreciated had affected the ethnographic records which involved attempts to "record and map the areas of different Aboriginal groups" (also at [203]). 311 Another important fact, recorded at [207] is this: Each of the anthropologists accepted that it was most likely that Lake Torrens would have been subject to traditional rights and interests by an Aboriginal society, or societies, at sovereignty and that it is likely that members of country groups closest to Lake Torrens would likely have had stronger rights and interests in the nearby portions of the Lake, its islands and springs, than others. In that respect the anthropologists disagree with the views of Norman Tindale, who regarded salt lakes as unoccupied areas, as demonstrated by his 1974 mapping and explanatory statement. 312 At [209] and [210] the primary judge returned to the concept that the Adnyamathanha and the Barngarla Peoples were present on the land to the immediate west of Lake Torrens (the Kokatha determination area) at least at sovereignty, noting this: 209 The Kokatha People submitted that the Kuyani People and the Barngarla People, along with the Kokatha People, may have been present in some parts of the area west of the western shoreline of Lake Torrens prior to European settlement, but subsequently the Barngarla People and the Kuyani People left those lands and ceased to have any continuing connection with that area. They said, correctly, that such a factual scenario would be consistent with the determination and the findings in Kokatha Part A that only the Kokatha People possess native title rights and interests in the land that is contiguous with the western shoreline of the Lake. 210 That is supported by the evidence given by Willis. He accepted that, at effective sovereignty, in addition to the Kokatha, the Kuyani were definitely on the western side of the Lake. He said that progressively through to the end of the 19th Century and the beginning of the 20th Century the Kokatha presence was growing at the same time that the Kuyani and the Barngarla presence was retreating and that by 1940 there were almost no Aboriginal people other than the Kokatha People on the western side of the Lake. 313 These paragraphs leave open the date from which the Adnyamathanha and the Barngarla Peoples may be said to have ceased to have a continuing connection with the land to the immediate west of Lake Torrens. As noted above, the Kokatha determination, given the terms of s 223, depends on no Aboriginal people other than the Kokatha having rights and interests in relation to the determination area under their traditional laws and customs at (and from) the date of determination. 314 At [247] the primary judge noted that RH Mathews (1900) had said: Touching the Kooyeeunna [Kuyani or Adnyamathanha] about Red Lake, and extending thence southerly down the western side of Lake Torrens is occupied by the remnants of the Hillary, Kakkarurra, Yallingarra and other friendly tribes, meeting the northern limit of the Parnkalla [Barngarla] and Nauo … Lying to the west of the Parnkalla [Barngarla], Hillary, and Arrabuna Nations, the country is occupied by several tribes, including the Kookatha [Kokatha]…The Kookatha [Kokatha] tribe extends from the Stuart Range, taking in Lake Phillipson, Mr Eba Station, Wilgena Station, Mount Finke and Lake Bring. 315 At [250] the primary judge said: …I do not regard the map and paper as probative in any material way of any one of the applicant groups holding native title rights in the area immediately to the western side of Lake Torrens, or on Lake Torrens itself. Given the decision in Kokatha Part A, I am unable to ascribe any real significance to the reference to "the Hillary Kakkarurra, Yallingara … tribes" to the west of Lake Torrens if it excludes the Kokatha People. 316 Again, the worst that might be said of this paragraph is that, if taken in isolation, it is ambiguous. The primary judge might have meant that he was going to disregard the evidence of Barngarla People occupying the land to the immediate west of Lake. If so, this would be in error because the Kokatha determination does not exclude the possibility that other Aboriginal people, at sovereignty and thereafter until some time before the Kokatha determination, possessed rights and interests under their traditional laws and customs which connected them with the area. Alternatively, the primary judge might have meant only that he could not give weight to the evidence only to the extent that such evidence suggested that, at sovereignty and thereafter, the Kokatha People themselves did not have rights and interests in the area under their traditional laws and customs which connected them with the area. This would not involve any error. 317 A similar ambiguity exists in respect of the primary judge's conclusions about the work of Elkin (1930, 1938 and 1944). The primary judge recorded at [257] Elkin's work in respect of the Kwiani (Kuyani): The Wailpi (or Adnya-matana) of the Flinders Range, the Kwiani on their west, the Yadliaura on their east, and the Pankala (Bangala) on the south and west of Lake Torrens and the northern part of Eyre's Peninsula. 318 At [260] the primary judge said: It is not necessary to consider the varying descriptions of the areas around Lake Torrens in detail. They may be a reflection of the particular matters being considered by Elkin in those reports or they may also reflect the relative and agreed increase in the population of Kokatha People living in the area to the west of Lake Torrens by the 1940s. They do not indicate that the Kokatha presence extended eastwards into Lake Torrens. 319 The Adnyamathanha and the Barngarla Peoples note that while it is true that Elkin's work does not "indicate that the Kokatha presence extended eastwards into Lake Torrens", this is not an accurate summary. The work, they say, clearly shows the Adnyamathanha and the Barngarla Peoples around Lake Torrens with the dividing line between the Lakes kinship system and the Western Desert kinship system well to the west of Lake Torrens. Again, it is not apparent from [260] in isolation that the primary judge was disregarding the evidence of the connection of the Adnyamathanha and the Barngarla Peoples under traditional laws and customs to the land around Lake Torrens (which would involve error) or merely observing that, whatever else it might show, Elkin's work did not assist the Kokatha (which would not involve error of the kind asserted by the Adnyamathanha and the Barngarla Peoples by reference to the Kokatha determination). 320 At [270] the primary judge said this: The review of that material to this point does not point in any significant way to supporting any of the three Applicants' claims. 321 The Adnyamathanha and the Barngarla Peoples say that this exposes error which is explicable only by reference to a misuse of the Kokatha determination because the material which the primary judge had reviewed all showed the Adnyamathanha and the Barngarla Peoples surrounding Lake Torrens with the Kokatha moving in from the west in the 20th century. Again, however, [270] is at worst ambiguous when read in isolation. It may be that the primary judge was disregarding the evidence of the Adnyamathanha and the Barngarla Peoples surrounding Lake Torrens in the ethnographic record until a 20th century westward movement by the Kokatha People. If this was so, I consider there would be error as the Kokatha determination did not require this evidence to be disregarded. It may also be that the primary judge meant only that because the Kokatha must also have been connected with the land to the west of Lake Torrens under their traditional laws and customs (a necessary fact underlying the Kokatha determination), the records showing the presence of the Adnyamathanha and the Barngarla Peoples surrounding Lake Torrens were neutral in the sense that the ethnographic record showing the presence of the Adnyamathanha and the Barngarla Peoples had to be understood against the indisputable fact (that is, indisputable by reason of the Kokatha determination) that the Kokatha People were also present in that area from sovereignty. 322 The primary judge then said this: 271 If one starts with the premises of the three claim groups at settlement occupying the areas as determined (and so virtually all the areas around Lake Torrens), until the work of Mountford and Tindale in the late 1930s and early 1940s, the material does not focus on the "occupation" or the exercise of traditional laws and customs with respect to the area of Lake Torrens, including Andamooka Island, by any particular group of Aboriginal people. 272 Although, as noted, the terrain of Lake Torrens is a very harsh and confronting one, Sutton said (uncontroversially, according to the other expert anthropologists), that it was very unlikely that such a geographical feature would not attract or have Indigenous traditional owners under classical conditions. Nevertheless, the ethnographic observations to that time do not record any instance or observations indicating in any persuasive way who those traditional owners might have been. 273 If there is any relevant focus to be gained from that material, it is that the movement of the Kokatha People eastwards towards Lake Torrens (accepting that by settlement they were the traditional owners of the area immediately to the west of Lake Torrens) was an ongoing one. 274 There is some ethnographic material from the next decades which places the Barngarla People on the western side of Lake Torrens up to about the Carrapateena Arm, and the Kuyani People on the western side of Lake Torrens north of the Carrapateena Arm, including those adjacent parts of the Lake itself (Pratt 1967, Elkin 1976), although Elkin's later work also places the Kokatha in that vicinity. To the extent that such material would exclude the Kokatha People as the traditional owners of the area to the west of Lake Torrens, both north and south of the Carrapateena Arm, it must be given no weight for the reasons already given. 323 These paragraphs involve the same potential ambiguity and another alleged error. The potential ambiguity is that, as noted, the ethnographic record before the 20th century consistently shows the Adnyamathanha and the Barngarla Peoples surrounding Lake Torrens. Accordingly, the statement in [272] that the "the ethnographic observations to that time [the 1930s] do not record any instance or observations indicating in any persuasive way who those traditional owners might have been" is wrong if it means that the observations did not involve the Adnyamathanha and the Barngarla Peoples surrounding Lake Torrens. If, however, the statement means that the observations showed the Adnyamathanha and the Barngarla Peoples surrounding Lake Torrens but are not persuasive because, given the Kokatha determination, the Kokatha People must also have had a presence in the same area, then the statement is not wrong. 324 From the many orthodox statements in the judgment noted above an inference of error by reason of mere ambiguity would not lightly be drawn. However, the ambiguity is further exposed by the statements in [273] in parentheses that "accepting that by settlement they [the Kokatha] were the traditional owners of the area immediately to the west of Lake Torrens" and in [274] referring to the Kokatha as "the traditional owners of the area to the west of Lake Torrens". The concept of "traditional owners" does not appear in the NTA but is commonly understood to mean the Aboriginal people who have rights and interests in land under their traditional laws and customs. On this basis, the primary judge was correct to describe the Kokatha as "traditional owners" of land to the west of Lake Torrens because this is a fact necessary to the Kokatha determination. However, and as discussed, this does not mean that the Kokatha were the (in the sense of the only possible) traditional owners of that land pre-sovereignty or "at settlement" (assuming this means, as it appears to in the reasons for judgment, at the time white settlers occupied the land). There was a wealth of evidence in the ethnographic record which may well have indicated that the Adnyamathanha and the Barngarla Peoples also had rights and interests in the land under their traditional laws and were present on that land at least until the 1940s. Again, read in isolation, the fact that the primary judge in [272] referred to the Kokatha as the traditional owners of the area to the west of Lake Torrens at settlement might indicate error but the reference to the material being given no weight only to the extent it would exclude the Kokatha People as the traditional owners, but for the use of the definite article at least, would suggest to the contrary. It would be a long bow to find error on the basis of the stray use of a definite article. 325 The same difficulties confront the submissions of the Adnyamathanha and the Barngarla Peoples about [304] and [331] which contain statements again characterising the ethnographic materials as not supporting any one party more than the other when, in fact, that record shows the Adnyamathanha and the Barngarla Peoples surrounding Lake Torrens until a 20th century migration westwards by the Kokatha People. That the primary judge so understood the ethnographic record appears from [341] in which his Honour said: On that material, in my view, there is little to support the claims of the Kokatha People to be the sole traditional owners of Lake Torrens either at sovereignty or at any later time, at least up to the late 1980s, in the ethnographic material. I have considered that material both as it appears on its face, and with the insight which certain of the anthropological experts sought to attribute to it. 326 This statement also weighs against the contention of error put by the Adnyamathanha and the Barngarla Peoples in respect of the primary judge's use of the Kokatha determination. 327 However, things change at [343] to [345]. It is at this point that earlier statements, which may be ambiguous but which I would not conclude indicate error, are further exposed. In these paragraphs it is clear that, despite all of the orthodox statements which precede (and, indeed, follow) these paragraphs, the primary judge's reasoning process has miscarried in two respects. First, it is in these paragraphs that the primary judge expressly places weight on the lack of evidence that the Adnyamathanha and the Barngarla Peoples occupied Lake Torrens, when Lake Torrens was not capable of occupation (as the primary judge knew) and occupation is not a prerequisite for the establishment of native title under s 223 of the NTA. Second, it is also in these paragraphs that the primary judge expressly states that the Kokatha determination means that the Adnyamathanha and the Barngarla Peoples were not the traditional owners of the land to the immediate west of Lake Torrens "at and from settlement" as that land "was and is the country of the Kokatha People". Lest there be any confusion about what the primary judge meant by "settlement", at [345], the primary judge repeated both propositions, as to occupation and the effect of the Kokatha determination at [345] in these terms and also by reference to sovereignty: The ethnographic material then, if assessed on the basis of movement south and west from the northern and eastern boundaries of Lake Torrens for the Adnyamathanha People and if assessed on the basis of movement north and west from the southern boundary of Lake Torrens, does not provide evidence of their occupation and use of Lake Torrens so as to contribute towards a state of satisfaction that they, or either of them, were and are the traditional owners of Lake Torrens or particular parts of it at sovereignty or continuously since sovereignty. 328 From this I infer that the primary judge was using "settlement" to mean what was referred to in [141] as "effective settlement" (that is, by Europeans), from which it could be (and was) inferred that the position was the same at sovereignty. 329 I do not see any basis upon which it might properly be considered that these paragraphs are merely ambiguous. The primary judge, in these paragraphs, can only be understood as saying that the Kokatha determination meant that the Adnyamathanha and the Barngarla Peoples could not have had any rights and interests in relation to the land to the west of Lake Torrens under their traditional laws and customs at any time from sovereignty. As discussed above, this is not the effect of the Kokatha determination. Further, the primary judge's reference to that land as land which "was and is country of the Kokatha People" can only be understood as a statement that, because of the Kokatha determination, it must be the case that the Kokatha People alone, from sovereignty and at all times until the Kokatha determination, were the only Aboriginal people who had or could have had rights and interests in relation to the land to the west of Lake Torrens under their traditional laws and customs. This is also not the effect of the Kokatha determination. Finally, the primary judge's references to "the occupation" of Lake Torrens indicates that his Honour was focusing on the occupation of the area when native title does not depend on occupation, and cannot so depend in respect of an area which is effectively uninhabitable. Given the undisputed anthropological evidence that it was very unlikely that a feature such as Lake Torrens would not have had traditional owners despite it being unsuitable for occupation, to which the primary judge referred at [207], [272] and [386], his focus on occupation of the claim area was apt to give rise to error. 330 These paragraphs also form part of the context in which the earlier paragraphs of the judgment, which I have described above as ambiguous, must be assessed. In the light of [343] to [345] it is difficult not to see those ambiguities in a different light, indicative of erroneous approaches to the effect of the Kokatha determination and the relevance of evidence (or its lack) of occupation of Lake Torrens which belie the orthodox statements of principle. 331 This indication of an erroneous approach to two critical issues is reinforced by [364] and [365] of the reasons. In those paragraphs the primary judge said: 364 The difficulty, in my view, with taking the linguistic evidence too far is the extent to which the pattern it is said to disclose is related more or less consistently to the western side of Lake Torrens (as well as to the areas to the south and north) in terms which point to a conclusion inconsistent with Kokatha Part A. To the extent that that picture therefore shows significant use of Barngarla words or word derivations, on the western side of Lake Torrens and appears to exclude any Kokatha presence, the evidence cannot be received to support the proposition that the presence of Barngarla People in those areas shows or tends to show that, at settlement, they were the traditional owners of all that land. If the pattern or patterns were confined to areas to the south, east or north of Lake Torrens, they may have some probative significance. But their significance must be taken from the pattern. The pattern can, and does, show significant Barngarla presence and probably significant Kuyani presence in the areas around Lake Torrens. That that was the state of affairs accepted by Willis in any event, and was the general effect of the evidence. Having regard to the decision in Kokatha Part A the pattern cannot show that that presence demonstrated that the Barngarla (or Kuyani) People had at, and therefore since, sovereignty native title rights and interests under their traditional laws and customs in the area west of the western boundary of Lake Torrens. Once that step is taken, it is hard to draw any inference from the linguistic analysis which informs the identification of the Aboriginal People who, by their traditional laws and customs, held native title rights and interests over Lake Torrens or part of it. 365 Indeed, given that evidentiary pattern of word use and word derivation in the area to the west of Lake Torrens, and the Kokatha Part A determination in respect of that area, the suggested explanation for the absence of Kokatha word usage and word derivation in place names in the area, namely that in Western Desert societies such word usage or word derivation is kept confidential, appears the more likely. If that step is taken, the pattern demonstrated by the linguistic evidence does not really advance the case of the Barngarla People (or the Adnyamathanha People) in any event. 332 In [364] the primary judge must be understood as saying that the linguistic evidence of Barngarla presence to the west of Lake Torrens could not be "received" (although it was received in the sense of being admitted, this must mean that the evidence was not given any weight) because, in common with the evidence showing "significant Barngarla presence and probably significant Kuyani presence in the areas around Lake Torrens", the Kokatha determination meant that such evidence "cannot show that the Barngarla (or Kuyani) People had at, and therefore since, sovereignty native title rights and interests under their traditional laws and customs in the area west of the western boundary of Lake Torrens". To the contrary, the Kokatha determination did not exclude the possibility that the Adnyamathanha and the Barngarla Peoples, in addition to the Kokatha People, from sovereignty and up to some time before the date of the Kokatha determination had rights and interests under their traditional laws and customs in the area west of the western boundary of Lake Torrens. The Kokatha determination meant only that, as at and from the date of the determination, they had no such rights and interests capable of recognition under the NTA. 333 For the primary judge to not consider or refuse to give weight to evidence of the pattern of Aboriginal presence and language to the immediate west of Lake Torrens because it supported or tended to support the connection of the Adnyamathanha and the Barngarla Peoples under their traditional laws and customs to that land from sovereignty until some time before the Kokatha determination merely because of the Kokatha determination was in error. It effectively doomed the claims of the Adnyamathanha and the Barngarla Peoples in relation to Lake Torrens to fail. The potential importance of the evidence to which no weight was given is plain. If a proper understanding of the evidence meant that: (a) the pattern of Aboriginal presence and language indicates a pre-sovereignty connection of the Adnyamathanha and the Barngarla Peoples to all of the land surrounding Lake Torrens; (b) the Kokatha determination recognises the Kokatha connection to the land to the west of Lake Torrens; (c) the Kokatha determination establishes that any connection of the Adnyamathanha and the Barngarla Peoples to the land to the west of Lake Torrens did not exist as at and from the date of the determination; (d) the Adnyamathanha and the Barngarla determinations recognise the Adnyamathanha and the Barngarla connection to the east and south of Lake Torrens and establish that any connection of other Aboriginal Peoples (including the Kokatha) to that land did not exist as at and from the determinations; (e) the anthropologists agreed, as they did, that it was very unlikely that there were not pre-sovereignty rights and interests under traditional laws and customs in relation to Lake Torrens and that proximity of presence to the lake was likely to be related to the existence and the strength of those rights and interests (see [207], [272] and [386]), then it is possible that the claims of the Adnyamathanha and the Barngarla Peoples to Lake Torrens might have been inferred to be stronger than those of the Kokatha, at least if there was no evidence of pre-sovereignty Kokatha presence to the east of Lake Torrens in the determination areas of the Adnyamathanha and the Barngarla Peoples. By giving the Kokatha determination the effect of excluding from consideration the evidence of the presence of the Adnyamathanha and the Barngarla Peoples and of their language to the west of Lake Torrens from sovereignty or from settlement, the main planks of the claims of the Adnyamathanha and the Barngarla Peoples to Lake Torrens were removed. 334 The effect of the primary judge excluding evidence from consideration on the inferences he drew is also exposed in [365] where his Honour inferred that the absence of the Kokatha language to the west of Lake Torrens, given the Kokatha determination, was more likely to be a result of the confidentiality of word use in Western Desert society than to advance the cases of the Adnyamathanha and the Barngarla Peoples. Given that the Kokatha determination did not mean that the Adnyamathanha and the Barngarla Peoples did not have rights and interests in the land to the west of Lake Torrens from before sovereignty until the date of the Kokatha determination, the evidence of Barngarla language use in that area and lack of evidence of Kokatha language was capable of advancing the cases of at least the Barngarla People and, possibly, the Adnyamathanha People as another group of the Lakes society. 335 It may have been that the inference the primary judge drew (of Kokatha confidentiality) was still open, but the question whether or not to draw that inference had to be answered on the basis of the evidence of the whole and not by disregarding evidence supporting the connection of the Adnyamathanha and the Barngarla Peoples to the land to the west of Lake Torrens before the date of the Kokatha determination. It is only if that evidence is disregarded that it is possible to say, as the primary judge did, that the evidence "does not really advance" the cases of the Adnyamathanha and the Barngarla Peoples in any event. Otherwise it must be that this evidence was capable of advancing their cases because it was at least open to infer from it that from sovereignty to some time before the Kokatha determination the Adnyamathanha and the Barngarla Peoples had rights and interests to the land surrounding Lake Torrens. When taken with the anthropological evidence of the high likelihood of traditional rights and interests in relation to Lake Torrens, the existence, extent and strength of which were determined by proximity to the Lake, the linguistic evidence was significant to the cases of at least the Barngarla People and most probably the Adnyamathanha People. 336 The same error is apparent in the way in which the primary judge dealt with the evidence of Michael McKenzie, an initiated Adnyamathanha man, who gave evidence of mura (stories) he had learned about what he considered to be Adnyamathanha country which included Lake Torrens (at [478]). At [479] the primary judge said this: To an extent, it is necessary to be cautious about giving weight to his evidence where it relates to the area to the west of Lake Torrens and is or may be inconsistent with the findings underlying, or the orders made, in Kokatha Part A. Those informants included Max Thomas. The fact that his evidence about Lake Torrens itself, and including Adnamooka Island, is in part sourced or learned from those who gave him history or stories not consistent with Kokatha Part A also causes me to pause before placing much weight on his evidence directly concerning the claim area. 337 The problem with this paragraph is that, while it is qualified by the opening words, it is apparent from what follows that the primary judge considered that stories which Mr McKenzie had been told about Lake Torrens by people who believed that the land to the west of Lake Torrens and the subject of the Kokatha determination was also Adnyamathanha country and not Kokatha country could not be given much weight because those stories were inconsistent with the Kokatha determination. Even if the limited effect of the Kokatha determination is put to one side, it is not the case that stories Mr McKenzie had been told by his elders and others about Lake Torrens being Adnyamathanha country were adversely affected because the elders or others also thought the land to the west of Lake Torrens was Adnyamathanha country and not Kokatha country. Apart from this, as discussed, evidence of the connection of the Adnyamathanha People to the land to the west of Lake Torrens before the date of the Kokatha determination was not inconsistent with the Kokatha determination. Such evidence was capable of supporting the claim of the Adnyamathanha People to Lake Torrens and was entitled to the weight which it would have been given apart from the Kokatha determination. The evidence was devalued by the primary judge on the basis of an erroneous approach to the effect of the Kokatha determination. 338 The primary judge's treatment of the evidence of Rosalie Richards, a non-Aboriginal woman who was married to Leroy Richards, who was both Adnyamathanha and Barngarla, and of their daughter Amanda Richards, discloses the same problem. Rosalie Richards' evidence was about what her husband had told her about the extent of Adnyamathanha and Barngarla country. Contrary to [557] of the primary judge's reasons, her evidence of Barngarla country, at some time before the Kokatha determination, having extended to the west of Lake Torrens was not necessarily inconsistent with the Kokatha determination. The inconsistency, if any, would have been confined to evidence, to the extent it was, to the effect that this land was exclusively Barngarla country or continued to be Barngarla country as at the date of the Kokatha determination and thereafter. This confined scope of inconsistency would have left intact the evidence of the existence of a Barngarla connection with the land to the west of Lake Torrens before the Kokatha determination. The primary judge, however, said the evidence "cannot be taken as meaning the area of the Kokatha Part A determination" and, it must be inferred, did so because of a mistaken view about the effect of that determination. It is also apparent from the primary judge's treatment of Rosalie Richards' evidence about Lake Torrens as described in [559] that his Honour took the same view about the effect of the Adnyamathanha and Barngarla determinations as he did for the Kokatha determination. Accordingly, evidence that the rights and interests of the Barngarla People under traditional laws had extended into the area of the Adnyamathanha determination was effectively discounted by the primary judge in [559] and [565]. As with the Kokatha determination, the Adnyamathanha determination does not mean that the Barngarla People may not have had rights and interests in relation to that land under traditional laws and customs from sovereignty until some time before the date of the Adnyamathanha determination. Evidence to this effect was rationally capable of supporting the claim of the Barngarla People to Lake Torrens. 339 The primary judge also gave no weight to the evidence of Eric Paige about the rights and interests of the Barngarla People under their traditional laws and customs in relation to land within the Kokatha determination area. Mr Paige was both an initiated Western desert man and a Barngarla man. At [671] the primary judge said that he had "addressed earlier in these reasons the weight that can (or cannot) be given to this story, showing a Barngarla connection from Iron Knob to Roxby Downs in the light of Kokatha Part A". The explanations can only be inconsistency of that evidence with the Kokatha determination but, as discussed, the inconsistency is more confined than the primary judge considered it to be. His Honour appears to have taken the same approach at [675] to certain evidence of Graham Richards, who is both Adnyamathanha and Barngarla. 340 The conclusions in [712] are also at least suggestive of the same error insofar as they mean that the Kokatha determination meant that at sovereignty the Kokatha "occupied" the land to the west of Lake Torrens, occupation not being a prerequisite to a determination of native title, or that no other Aboriginal people from sovereignty may have had rights and interests under traditional laws and customs in relation to that land. 341 The error is suggested again in [729] where the primary judge said: It is certainly the case that, as elsewhere noted, some Kuyani People (and some Barngarla People) at around the time of first observations around the 1850s, and possibly considerably earlier, spent some time on land to the west of the western side of Lake Torrens. That is confirmed to some degree not simply by the anecdotal lay evidence, but by some of the ethnographic material referred to and in part by some of the linguistic evidence. It does not show that, at sovereignty, the Kokatha People did not have the traditional Aboriginal rights to the country immediately to the west of Lake Torrens. 342 The cases of the Adnyamathanha and Barngarla Peoples were not that they merely had a presence to the west of Lake Torrens before the Kokatha determination but that they had rights and interests in relation to that land under their traditional laws and customs and that the Kokatha determination mean only that their rights and interests must have ceased by the date of the determination but no more. As such, the determination did not mean that the Kokatha People had "the traditional Aboriginal rights to the country immediately to the west of Lake Torrens" if, by this, the primary judge meant the only traditional Aboriginal rights to that country from sovereignty, as is apparent from other parts of the judgment. 343 The error is also suggested by [731] where the primary judge said: It is known that, at sovereignty, the Kokatha People held the native title rights and interests in the area immediately to the west of Lake Torrens. 344 Read in the context of the paragraph this statement, more probably than not, indicates that the primary judge took the Kokatha determination to mean that only the Kokatha had rights and interests in that land under traditional laws and customs from sovereignty. 345 The error is clear again in [744] where this is said: From the second or third decade of the 20th century, there is evidence of the ancestors of the McKenzie witnesses, and their more intermediate ancestors, having lived and worked in areas to the west and to the east of the Lake up to about the 1970s. Again, their activities west of the Lake, in my view, do not inform the decision to be made in this matter, having regard to the determination in Kokatha Part A. Nor is there any evidence of Fred McKenzie, the father of Malcolm McKenzie, having any association with Lake Torrens or Andamooka Island. Indeed, both he, and his father Mt Serle Bob, are in some sections of the evidence described as non-Kuyani affiliated (as the Ellis Report observes). 346 In other words, by reason of the Kokatha determination, which was made in 2014, the primary judge gave no weight to the evidence of Adnyamathanha people of their relationship to and use of land to the west of Lake Torrens up to the 1970s. This approach involves error for the reasons already given, but also discloses why the response of the State and the Kokatha People cannot be accepted. The response was that any error the primary judge made about the effect of the Kokatha determination was immaterial because the Adnyamathanha and Barngarla Peoples' claims failed by reason of lack of continuity up to the present day. If, as is the case, the evidence of the use of and relationship to the land to the west of Lake Torrens by any people other than the Kokatha People from sovereignty was wrongly disregarded or discounted because of the Kokatha determination, then it necessarily follows that the conclusions of lack of continuous connection with Lake Torrens must have been infected by that error. 347 In [758] the primary judge repeated his reluctance to give any weight to Barngarla place names to the west of Lake Torrens because of the Kokatha determination, which involves the same error. 348 The effect of the misuse of the Kokatha determination on the claim of the Adnyamathanha and Barngarla Peoples is also clear in [762] where the primary judge said about the latter's claim: …as with the Adnyamathanha People's claim, even if such a state of affairs were found to exist at the time of first European contact, it is not a case where it can be inferred that that state of affairs existed for the preceding 60 years or so to sovereignty. That is primarily because of the Kokatha Part A determination, but is reinforced by the evidence about the eastwards tending movement of the Western Desert Cultural Bloc and the accepted western movement of the Barngarla (and Kuyani) People from the western side of Lake Torrens over that period and presumably subsequent decades. The observations I have made on this topic in relation to the Adnyamathanha claim apply equally to the issue of continuity concerning the Barngarla claim. 349 As discussed, the Kokatha determination says nothing about the rights and interests of the Adnyamathanha and Barngarla Peoples under their traditional laws and customs in relation to the land to the west of Lake Torrens from sovereignty to immediately before the making of the determination. As such, it cannot have been relevant to the state of affairs for 60 years from sovereignty. 350 In [763] and also [764] the primary judge repeated that, in respect of the claim of the Barngarla People, he has "not placed any weight on material to the extent that the anthropological evidence is inconsistent with the Kokatha Part A and Adnyamathanha No 1 determinations". Given those earlier parts of the reasons for judgment, in which it is apparent that the primary judge took an over-expansive view of the effect of the determinations and thus the scope of any inconsistency between the evidence and the determinations, this discloses that the error had a material effect on the conclusions his Honour reached about the Barngarla People's claim. 351 It is true that at various places in [728] to [773] the primary judge also recorded his conclusions that the claims of the Adnyamathanha and Barngarla Peoples failed because they had not proved the continuity of their connection to Lake Torrens under their traditional laws and customs. As noted, this must have been affected by the primary judge's view that the Kokatha determination meant that evidence of the Adnyamathanha and Barngarla Peoples' connection to the land to the west of Lake Torrens under their traditional laws and customs had to be disregarded or discounted. For this reason alone these conclusions do not mean that the error in respect of the effect of the Kokatha determination was immaterial. There are other difficulties with the approach of the primary judge to this issue. Of themselves, the second, third and fourth of these matters, as discussed below, might not justify appellate intervention but, taken with the effect on the process of reasoning that the primary judge's views about the effect of the determinations must have had, they provide a further foundation for the conclusion I have reached that these appeals must be allowed. 352 First, the primary judge referred to there being a "missing" 60 years in the evidence of the Adnyamathanha and Barngarla Peoples in [731] and [762]. In [731] the primary judge, referred to the lack of evidence of Adnyamathanha people having rights and interests in Lake Torrens at the time of first European contact (referred to as settlement by the primary judge) or, working backwards from there to sovereignty, which he describes in [730], saying: In addition, as I have discussed, even if that state of affairs were shown to have existed at the time of first European contact, there is a further period of some 60 plus years to be addressed. It is known that, at sovereignty, the Kokatha People held the native title rights and interests in the area immediately to the west of Lake Torrens. It is known that, progressively, the Kuyani people who were at the time probably still in part on the western side of Lake Torrens were moving eastwards. Those two matters, or even the first, would operate to discourage any inference that, for all of the 60 years or so proceeding European contact, the Kuyani (Adnyamathanha) People held the native title rights and interests over the entirey of Lake Torrens to its western boundaries. The same observation applies in relation to the claims of the Third Applicant. 353 The "state of affairs" being described must be of Adnyamathanha people being present to the west of Lake Torrens in the area of the Kokatha determination. The 60 years must be the period between sovereignty and settlement. This is confirmed by subsequent paragraphs as follows: 762 Again, as with the Adnyamathanha People's claim, even if such a state of affairs were found to exist at the time of first European contact, it is not a case where it can be inferred that that state of affairs existed for the preceding 60 years or so to sovereignty. That is primarily because of the Kokatha Part A determination, but is reinforced by the evidence about the eastwards tending movement of the Western Desert Cultural Bloc and the accepted western movement of the Barngarla (and Kuyani) People from the western side of Lake Torrens over that period and presumably subsequent decades. The observations I have made on this topic in relation to the Adnyamathanha claim apply equally to the issue of continuity concerning the Barngarla claim. … 772 Although the ethno-historical records provide some support for the Adnyamathanha (Kuyani) People being associated at least with part of the claim area, mainly in its northern part at the time of first European contact, it is difficult inferentially to take that back to the time of sovereignty. 354 The basis upon which the primary judge considered that there was an evidentiary gap for the 60 years from sovereignty to settlement is unclear. But the relevant point is that there was no such evidentiary gap. To the contrary, the primary judge had recorded at [191] the agreement between the anthropologists that: (1) at some time at or before sovereignty there was a large population of Kuyani or Barngarla People on the western side of Lake Torrens (and, I add, also Kokatha People - although not all the experts accepted that); … 355 Given this, the conclusion that the claims of the Adnyamathanha and Barngarla Peoples should fail because there was an evidentiary gap between sovereignty and settlement involves error in the sense described in Robinson Helicopter Company Inc v McDermott [2016] HCA 22; (2016) 331 ALR 550 at [43] (references omitted) that the conclusion was "wrong by 'incontrovertible facts or uncontested testimony', or …'glaringly improbable' or 'contrary to compelling inferences'". 356 Second, while they may be directional errors, something has gone wrong in [345] in which the primary judge said: The ethnographic material then, if assessed on the basis of movement south and west from the northern and eastern boundaries of Lake Torrens for the Adnyamathanha People and if assessed on the basis of movement north and west from the southern boundary of Lake Torrens, does not provide evidence of their occupation and use of Lake Torrens so as to contribute towards a state of satisfaction that they, or either of them, were and are the traditional owners of Lake Torrens or particular parts of it at sovereignty or continuously since sovereignty 357 These comments are based on the primary judge's assessment of the ethnographic materials but it is clear that there was no western movement from the eastern side of Lake Torrens by the Adnyamathanha People as the eastern boundary of the Lake is the Adnyamathanha determination area. Further, this characterisation of the movement of Adnyamathanha People is also inconsistent with the undisputed evidence of all of the anthropologists as recorded at [191] that "at some time at or before sovereignty there was a large population of Kuyani or Barngarla People on the western side of Lake Torrens". 358 Third, the primary judge said at [738]: While I accept the heritage of the mura recounted by the Adnyamathanha witnesses, so far as I can discern, there are only two detailed stories (the Irti Vardnappa Kadni story and the Arkurru/Artunyi story) which specifically relate to the claim area. In relation to those two stories, I am cautious not to rely on post-sovereignty and contemporary evidence of Adnyamathana mura to the extent that it places contemporary Adnyamathanha rights and interests on the western side of Lake Torrens post-sovereignty. For the reasons I have given, I do not think they are persuasive evidence of the continued connection required by s 223 and 225 of the NTA as prescribed in Ward HC. Other stories, as noted in the earlier sections of these reasons, are more generic rather than related particularly to a part or parts of Lake Torrens. By way of example, the Alda/Kalta story was recorded by Mountford within the ranges. 359 In fact, the evidence of other stories given by Adnyamathanha People concerned the creation of Lake Torrens and traverses the claim area. The primary judge's discernment that only two stories did so is irreconcilable with the evidence. It is not possible to know for certain the source of the misapprehension but, at least in respect of one story, the problem is apparent on the face of the judgment. The Bivu story is described by the primary judge at [636] as one in which, as described by one witness, the ancestral hero Bivu "does not go into Lake Torrens itself". At [643], the evidence about Bivu by another witness involves Bivu being on and travelling across the Lake. At [651] the primary judge referred to the story of Vidni Murunhai which the primary judge said he had not placed much weight on because only one witness referred to it. It was not in dispute, however, that this was another name for the Bivu story. 360 Contrary to the submissions for the State, these matters cannot be explained by the primary judge having been confined by the fact that much of the evidence was subject to a gender restriction and confidential. That does not explain why the primary judge considered that only two stories of the many on which the Adnyamathanha People relied concerned Lake Torrens when all of the stories related to the Lake. 361 Fourth, at [774] the primary judge said: The conclusion reached, having regard to the anthropological evidence referred to, is somewhat counter-intuitive of an appropriate starting point. Each of the claim groups now has contemporary significant and credible spiritual connection to parts of Lake Torrens, but it is not possible, in my view, presently to prioritise one set of spiritual beliefs over the other for the purposes of a finding in terms of ss 223 and 225 of the NTA. As Sutton said in the course of his cross-examination, it is the abutment between the Lakes Cultural Bloc and the Western Desert Cultural Bloc that the Court is being requested to determine, and to use his word which, on the evidence I think is appropriate, to "re-imagine" what existed in 1788. The competing or inconsistent spiritual beliefs, which clearly exist, tend to demonstrate also the lack of the continuance of a dominant particular set of spiritual beliefs of one of the three Claim Groups over that of the others for the purposes of s 223(1)(b) of the NTA from sovereignty to contemporary times. 362 The first part of this paragraph, which refers to all three claim groups as now having "significant and credible spiritual connection to parts of Lake Torrens", is difficult to reconcile with the statement at [772] that the Adnyamathanha People had only "a remnant knowledge of sites, ceremonies and songs [in relation to the claim area] but it is broken and not coherent". The second part of this paragraph suggests that the primary judge was side-tracked by a concern that he was being required to "prioritise one set of spiritual beliefs over the other", which cannot be the case having regard to the criteria for native title established by s 223 of the NTA. 363 Despite the appeals having raised other issues, I consider that only one other contention by the Adnyamathanha and Barngarla Peoples should be assessed. Those other issues, including the contention that the primary judge ought to have adopted certain findings as a result of the determinations under s 86 of the NTA and of various other errors of fact or misunderstandings of the evidence by the primary judge, would not be sufficient to justify appellate intervention, despite it appearing that most of the contentions of factual error or misunderstanding of evidence, at least, are accurate. 364 The one other contention which is important is the contention that the primary judge erred in refusing to countenance a determination that both the Adnyamathanha and Barngarla Peoples had native title rights and interests in Lake Torrens. In my view, the primary judge was in error in refusing to countenance this possibility on the basis that it was precluded by the NTA. Undoubtedly, the possibility gave rise to case management considerations given the positions of the Kokatha People and the State, but the primary judge rejected the possibility on the basis that it was not open to consider it because of the NTA. 365 The primary judge explained the issue this way. At [44] the primary judge noted that: The three claimant applications, each as formulated, presents a claim for a determination being made in favour of one or other of the three Applicants to the exclusion of the others. That was the case presented by each of the Applicants in opening, and subject to certain admissions on the evidence and joint submissions made by the Second and Third Applicants in their oral, and supplementary, closing submissions. 366 While this was the primary case of each appellant, they also all accepted (as must have been the case) that they might not succeed in respect of the whole of their claim in any one of a number of possible ways including: (a) Their claim might be accepted over a part of the lake only and rejected over all other parts of the lake. (b) Their claim might be accepted over the part or the whole of the lake in addition to the claim of one or both of the other competing claimants. 367 The primary judge recognised this at [45] but said that what could not be done was to "create a fourth group, not one of the Applicant groups, and determine that that different group held native title over Lake Torrens or parts of it". This was correct, but I do not consider that the Adnyamathanha and Barngarla Peoples advanced any such proposal. The primary judge identified their alternative cases as follows: 103 It is clear enough that, consistently with those provisions, the Court on the present hearing could conclude that one or other of the Applicants has established native title over some part or parts of Lake Torrens. Indeed, as later noted the Third Applicant has acknowledged that the claim by the Barngarla People in relation to the northern part of Lake Torrens is not made out, so the primary determination sought is only to an area over the southern part of Lake Torrens. It is then accepted that the Second Applicant should be recognised as holding native title over the northern part of Lake Torrens. Although I have not mentioned there the First Applicant, one available conclusion would be that the Kokatha People hold native title over only part of Lake Torrens. 104 However, the Second and Third Applicants in final submissions put a further contention that, in respect of what might be described loosely as the middle third of Lake Torrens, that there should be a determination that each of the Adnyamathanha People and the Barngarla People together as one native title claim group have shared common rights. The proposed determination then describes the native title holders together as the Adnyamathanha People (as described on the Adnyamathanha Application) and the Barngarla People (as described on the Barngarla Application) with the same specified rights in the "Shared Area" to be exercised in accordance with the traditional laws and customs of the (collective) native title holders. 368 The primary judge did not consider that this further alternative was open because no such claim by any such claim group had been made under s 61 of the NTA and thus none of the procedural requirements in the NTA for claimant applications had been satisfied in respect of such a claim (see the reasons at [105] to [127]). The reason I consider this was in error is that the claims made by the Adnyamathanha and Barngarla Peoples did not exclude the possibility that another Aboriginal people might have native title rights and interests in relation to the claim area. If, as occurred in the present case, there were claims to the same land then s 67 of the NTA required the proceedings in which those claims had been made to be the subject of such orders as appropriate "to ensure that to the extent that the applications cover the same area, they are dealt with in the same proceeding". In the context of such a proceeding there is necessarily the inherent possibility that the competing claimants might both succeed as to whole or part or fail as to whole or part. If both succeed, for example, as to the whole, it is not the case that there is thereby created another new claim group. Nor does the NTA require another new claim group to be created, to be authorised to make the claim, or to be the subject of a separate application satisfying the procedural requirements of the NTA. 369 Accordingly, it is not the case that to support a conclusion of a part of Lake Torrens being subject to shared rights and interests of the Adnyamathanha and Barngarla Peoples there had to be a new native title claim group constituted and applicant authorised to make a new native title claimant application. The claim groups would remain the Adnyamathanha and Barngarla Peoples. Their claims would remain the claims as authorised. It is merely that the determination made under s 225, if the evidence supported it having regard to s 223, would be a determination that native title exists in relation to the relevant part of Lake Torrens and that the persons, or each group of persons, holding the common or group rights comprising the native title are the Adnyamathanha People as identified in the Adnyamathanha claim and the Barngarla People as identified in the Barngarla claim. As the submissions for the Adnyamathanha People put it, the NTA does not require a single application by a conjoined claim group for a determination to be made under s 225 of the NTA that more than one claimant group holds native title rights and interests in the relation to the same area. "Native title" is not necessarily unitary. It is "the communal, group or individual rights and interests of Aboriginal peoples or Torres Strait Islanders in relation to land or waters" (s 223). Nothing in the NTA suggests that only one group of Aboriginal people may hold such rights and interests in one area. To use a Western concept, a determination of native title under s 225 may be a determination of native titles held by more than one group of Aboriginal people. 370 Unlike the primary judge at [115], I do not consider that Banjima People v State of Western Australia [2015] FCAFC 84; (2015) 231 FCR 456 proposes that there may be only one native title group which holds the native title rights and interests over a particular section of country. To the contrary, the reasoning in Banjima, including at [54], recognised that there may well be more than one Aboriginal group that has native title rights and interests in relation to an area of land. In Banjima, the issue was whether those rights were exclusive of any other non-Aboriginal rights and interests. In circumstances where the rights and interests of other Aboriginal people in relation to the area (if they existed) were no longer asserted, it was held that there was no error in concluding that the rights and interests of the Banjima People were exclusive. 371 There may well have been other reasons which would have made it inappropriate for such a determination in favour of the Adnyamathanha and Barngarla Peoples to be made, including fairness to the Kokatha People and the State or inadequate evidence, but the NTA itself did not prevent the making of such a determination by reason of the nature of the claims made, their authorisation, or otherwise. The determination was a matter for the Court. In a case involving competing claims dealt with under s 67 of the NTA, nothing but the evidence and the dictates of fairness in the particular case prevents the making of a determination in which the competing claimants simultaneously succeed and fail (that is, succeed because they are found to have native title in relation to an area but fail because another Aboriginal group is also found to have native title in relation to the same area or part thereof). In such an event, it is the terms of the determination under s 225 which must identify the persons holding the rights comprising the native title(s) and the nature and extent of the native title rights and interests in relation to the determination area. The potential for inconsistency is to be resolved through the determination. As such, it is not necessary to consider s 84D of the NTA because there was no defect in the authorisation of the claims. 372 For these reasons, the appeals of the Adnyamathanha and Barngarla Peoples must be allowed. In accordance with the agreement of the parties given the grounds on which the Adnyamathanha and Barngarla Peoples have succeeded, the appropriate order is that these matters be remitted to a single judge (the primary judge having retired) for rehearing on the papers together with any further evidence that judge may allow.

Kokatha People's appeal 373 The Kokatha People's appeal does not involve any suggestion that the primary judge misused the native title determinations. Accordingly, this appeal must be dealt with separately. 374 One ground of the Kokatha appeal may be dismissed immediately. It is the contention that the primary judge, despite having heard evidence on a place called Crombie Ridge on Andamooka Island which was significant to all of the claims, mistakenly thought that the Kokatha witnesses placed Crombie Ridge outside of and to the west of the claim area. This was said to result from various paragraphs in the reasons for judgment which, on first reading and with no context, are ambiguous. 375 The paragraphs are as follows: 454 … In the course of their evidence, a statement of Eileen Wingfield of 4 November 2010 was verified as signed by her, and tendered by the Third Applicant. It included reference to a special place for women called Crombie Ridge (which is marked on Exhibit S1 as being on Andamooka Island, but was said by the Starkey brothers as being to the west of Lake Torrens). 534 He [Eric Paige] asserted cultural knowledge of two particular sites on Andamooka Island: Crombie Ridge (as marked on Exhibit S1) related to a Kalta-Widjeru story, and Crombie Sandhills (not marked on that map). I note that the Kokatha evidence referred to above also identified a site called Crombie Ridge, but placed it off Andamooka Island to the west. 593 She [Eileen Wingfield] agrees women are not allowed on Andamooka Island at all, and only wati can do so. She refers to the Seven Sisters tjukurpa or dreaming, including that the women were chased to Lake Torrens, where they sat and talked on the shore. That site is, by inference, identified as Crombie Ridge which is a restricted women's site and with restricted women's stories. As noted, there is an apparent dispute about the location of Crombie Ridge. She says the whole of Lake Torrens is significant to the Seven Sisters story, including a "big site" at a place around the Lake. 688 Andrew Starkey gave evidence of the importance of Andamooka Island and in particular the area marked as Crombie Ridge where he provided an explanation as to why the place is only visited by senior men and is significant. As noted, the Court was taken on a walk and shown some objects, the significance of which was explained in the restricted evidence of Andrew Starkey. They were not specific to either the Lake generally or to Andamooka Island. He also provided detail of ceremonies at Crombie Ridge in his restricted evidence. 376 In [609] the primary judge said: The existence of the Angkarta, Urumbulla and Wanambi stories were mentioned only in passing by Michael Starkey and there was no evidence given describing how those stories related to the claim area. 377 For the Kokatha People it was submitted that the primary judge must have believed that the Kokatha witnesses placed Crombie Ridge outside the claim area, otherwise his Honour could not have said that the stories which were told on and related to Crombie Ridge on Andamooka Island, within Lake Torrens, were not stories related to the claim area. 378 In my view, [609] stands or falls on its own. If it contains a material error, then so be it. I have dealt with [609] on this basis below. The paragraph does not, however, indicate that the primary judge believed that the Kokatha witnesses placed Crombie Ridge outside the claim area. It is clear that what the primary judge was trying to convey was that, insofar as there was women's evidence that there was an important women's area known as Crombie Ridge that important women's site, according to Kokatha men, was outside and to the west of the claim area. 379 Other alleged errors by the primary judge in respect of the Kokatha claim fall within the same class. That is, the worst that might be said is that the primary judge's reasons have been expressed in a potentially ambiguous manner or the primary judge has reached a conclusion about certain evidence contrary to that for which the Kokatha People contend, but which cannot be said to be other than reasonably open given that the primary judge had the benefit of seeing and hearing the evidence be given. These cautions apply, for example, to the Kokatha People's assertion of error by the primary judge in the way he dealt with the evidence of Max Thomas at [306] to [308] and Peter Mungkari at [438] to [442]. The mere fact that an appellate judge might have reached a different view about the significance of this evidence does not indicate error by the primary judge. 380 This said, I am nevertheless satisfied that the appeal of the Kokatha People must also be allowed. The cases were heard over many days between October 2015 and March 2016. Judgment was delivered on 9 August 2016, the primary judge having been burdened by many hundreds of pages of written submissions by the parties, in addition to oral submissions. The primary judge's reasons for judgment are lengthy. All appellants contended that parts of the judgment were difficult to reconcile and indicated that in a number of respects, including respects significant to their claims, the primary judge misunderstood or overlooked undisputed evidence. As with the claims of the Adnyamathanha and Barngarla Peoples, it is apparent that some of these contentions by the Kokatha People are accurate, the State's answer being effectively that any such error or misunderstanding was immaterial. While this may be true in relation to some matters if considered in isolation, the overall number of such matters and the importance of some of them to the claims which the primary judge dismissed make me more inclined to characterise these matters as material than might otherwise have been the case. 381 For the Kokatha People, one such contention related to reports from 1984 and 1989. The context is the primary judge's statement at [194] that he had placed "significant weight" on the ethnographic material. The 1984 report was prepared by P Fitzpatrick and T Gara and is entitled Anthropological Survey of Extractive Mineral Leases on Birthday, Four Mile and Nantilla Creeks near Port Augusta. Port Augusta is well to the south of Lake Torrens. The report refers to Kokatha mythology including the Sevens Sisters (Kunkaralinya) myth as existing at various levels including secret male and female versions. A highly secret and sensitive part of the male version of the myth involves a journey from Port Augusta to Lake Torrens and elsewhere. The report notes that the significance of the myth increases towards Lake Torrens and that Lake Torrens itself is of "major significance" in this mythology. 382 The primary judge does not refer to the 1984 report. He does refer to another report by T Gara from 1989 entitled The Kokatha Heritage Survey. In dealing with this report at [317] to [326] the primary judge recorded the statement in it that the Kokatha People today regarded their eastern boundary as the western side of Lake Torrens (p 5). At p 7 Gara noted that Kokatha informants were reluctant to divulge information of a secret/sacred nature or to have that information written down. At [325] the primary judge noted: Section 5.2 entitled "Kokatha boundaries" reviews ethnographic material (much of which is referred to above) as well as the Kokatha informants who (it is said) "regard their eastern boundary as being the western side of Lake Torrens". Those Kokatha informants included those who had worked in the Lake Torrens - Lake Gairdner area (at p 13). There is no reference to any significant site in or immediately adjacent to the western edge of Lake Torrens, including there being no reference to any significant site on Andamooka Island, although reference is made to sites somewhat west of there on Arcoona Plateau, and at or near Coorlay Swamps and Pernatty Lagoon and near Andamooka Homestead. It is noteworthy that those sites are recorded, because their relative proximity to the western side of Lake Torrens suggests that, if there were significant sites on or immediately adjacent to Lake Torrens, there is no reason why they would not have been identified. 383 At [326] the primary judge also referred to the following part of the report: Reference is also made (at p 12-13) of important Dreaming tracks and numerous "minor" tracks passing north-south "through the country west of Lake Torrens" with the observation that the Kokatha use these tracks "to define the eastern border of their land". That section of the report notes that the KPC are reluctant to divulge information about ceremonial sites on the Arcoona Plateau unless they are threatened by mining activities, but that the KPC had a high priority to document mythological sites in the Lake Torrens - Andamooka area. 384 The Kokatha People's argument is that the primary judge's treatment of Gara 1989 demonstrates that his Honour overlooked the 1984 report and thus overlooked important evidence within the 1984 report of the significance of Lake Torrens itself to Kokatha mythology. As a result, the primary judge drew a critical inference against the Kokatha People in [325] that, "if there were significant sites on or immediately adjacent to Lake Torrens, there is no reason why they would not have been identified" in the 1989 report, when, had the 1984 report been considered, the obvious and overwhelming inference would have been that the Kokatha informants in 1989 considered that the information about Lake Torrens was secret/sacred and thus, as apparent from Gara's statement at p 7 in the 1989 report, were reluctant to divulge the information. 385 I accept that the primary judge overlooked the 1984 report. Had his Honour considered the 1984 report he could not have reached the conclusion he did in [325]. To the contrary, that inference would have been "glaringly improbable" (Robinson Helicopter Company Inc v McDermott [2016] HCA 22; (2016) 331 ALR 550 at [43]) for the reason that, contrary to the 1989 report, the 1984 report did disclose Lake Torrens itself as a significant site to Kokatha People, at least by 1984. Given the significance which the primary judge placed on the ethnographic material at [194], this error must be material. 386 At [410] the primary judge, in discussing the evidence of Andrew Starkey, said that Mr Starkey "explained the significance of Andamooka Island generally, including particular physical features and during a view he showed certain stone tools and certain other very significant objects (not permanently there)". Subsequently, the primary judge at [688] said that these objects were "not specific to either the Lake generally or to Andamooka Island". Andrew Starkey described the objects as ones only certain initiated Kokatha men could touch. The objects were described as both sacred and deadly. One object was related to the Wanambi story. Andrew Starkey said that specific rituals were performed using this object on the surface of Lake Torrens. Another object was described as for the "man for the lake", the "Tjukurpa man". The primary judge's statement that the objects were not specific to the Lake, accordingly, is incorrect. Contrary to the State's submission I do not accept that the primary judge's statement at [688] should be understood as saying only that the significance of these objects was not exclusive to Lake Torrens. Nor does the highly secret nature of the evidence explain the statement at [688]. 387 Another contention for the Kokatha People concerned the primary judge's finding at [609] in a section dealing with Kokatha Tjukurpa that: The existence of the Angkarta, Urumbulla and Wanambi stories were mentioned only in passing by Michael Starkey and there was no evidence given describing how those stories related to the claim area. 388 This, it was submitted, also involved error by way of oversight. A report by R Hagen from August 1983 entitled Mythological Tracks in the Vicinity of the Olympic Dam Mining Venture was in evidence. The primary judge referred to this report at [309] in the following terms This is a restricted exhibit, so reference to it is oblique. It is entitled "Mythological Tracks in the Vicinity of the Olympic Dam Mining Venture". It reports on Dreaming travels through a triangle "bounded loosely" by Lake Torrens in the east, and then to the west. I considered each of the myths referred to. There is no clear indication, with one exception, that the tracks concern an area to the east of the western boundaries of Lake Torrens. One myth concerns the creation of Lake Torrens, including "where the Kokatha and Kuyani meet". It therefore records a mythological track which has Kokatha associations and concerns Lake Torrens itself. 389 For the Kokatha People it was submitted that [609] discloses that the primary judge overlooked the content of the report despite it being discussed at [309]. This was said to be supported by other parts of the reasons for judgment including: 719 … In 1983, Hagen went through the area with Kokatha men, including Max Thomas, but again the Dreaming stories relating to Lake Torrens and its occupation or possession by Kokatha People was not adverted to. 724 In my view, the absence of any material at all which tends to associate Kokatha interests with Lake Torrens until the relatively recent past, particularly in circumstances where it may have been expected that such Kokatha Dreaming stories as relate to Lake Torrens (according to the contemporary evidence) would have emerged in the course of those investigations particularly in the 1980s, and having regard to the fact that the maps drawn by others, including Tindale, in the earlier decades of the 20th century draw the Kokatha boundary on the western shore of Lake Torrens lead to the conclusion that, at sovereignty, the Kokatha People by their traditional laws and customs did not possess or occupy any part of Lake Torrens east of the western shore of Lake Torrens (as determined in the Kokatha Part A determination). 390 The State acknowledged that the last sentence of [719] was incorrect. The first sentence of [724] must also be incorrect because there was material, some of which the primary judge had identified at [309], which associated the Kokatha People with Lake Torrens. Given the reference to the expectation that such material would have emerged in the 1980s, the primary judge must have meant after the 1980s when he referred to "the relatively recent past" but, as noted, the evidence included ethnographic material from the early 1980s and oral evidence described as having been learnt from various elders. The knowledge of those elders must pre-date its transmission to the witnesses, at least part of which was said to have occurred in 1981 (at [413] and [430]). 391 If the statement in [609] concerns only the evidence which Michael Starkey gave, then it is not the case that Michael Starkey gave no evidence describing how the stories related to the claim area. Michael Starkey identified the Wanambi as Crombie Ridge on Andamooka Island. Otherwise, relevant oral evidence included that of Robert Starkey. The primary judge described that evidence at [417] to [426]. This description did not include any reference to the Angkarta, Urumbulla and Wanambi stories although the primary judge recorded the following: 418 He [Robert Starkey] was present during the Kokatha evidence on Andamooka Island, and contributed to the viewing of the sites and objects there. He became wati in 1993 at Amata, and continues to take part in ceremonies. 419 He confirmed the importance to him of Lake Torrens, including by reference to the Wati Nyiru Dreaming and the Seven Sisters Dreaming. The details are restricted, although there are "open" parts of those stories, and open parts of the associated dances. 392 At [423] the primary judge said: He had no knowledge of any east-west or west-east travel across Lake Torrens either as Dreaming stories or as significant historical events. 393 The transcript discloses that Robert Starkey gave evidence that Crombie Ridge on Andamooka Island in Lake Torrens was a place that Wanambi went to make rain. He pointed out two hills representing the Angkarta and Kalta Tjukurpa who had travelled to Andamooka Island along the causeway that joins the island to the mainland. He also described the travels of the Urumbulla which passed the western side of Lake Torrens. As noted, Andrew Starkey also gave evidence about the presence of the Wanambi on Lake Torrens. 394 In assessing the importance of the erroneous conclusions of the primary judge about the oral evidence, regard must also be had to the fact that at [405] he rejected a challenge to the credibility of the main Kokatha witnesses, the Starkey brothers, finding that there was no reason to doubt their honesty. The primary judge was undoubtedly right at [406] when he observed that there are a "range of reasons why evidence honestly given, and forcefully given, may not ultimately be found to prove all the critical fact or facts to which the evidence was directed". However, this does not authorise an appellate court to disregard conclusions as to a lack of evidence of critical matters which are demonstrably wrong. 395 The Kokatha pointed out other evidence of their connection to Lake Torrens in their submissions which do not appear in the reasons for judgment. While there is no obligation in reasons for judgment to refer to every piece of evidence, the critical fact in the present case is the primary judge's conclusion at [724] of the absence of any material until the relatively recent past (by which the primary judge must mean after the 1980s) "which tends to associate Kokatha interests with Lake Torrens" when the evidence to which his Honour did not refer did tend to associate Kokatha interests with Lake Torrens. This indicates not merely that the primary judge did not happen to mention certain evidence in his reasons for judgment, but that the primary judge did not consider significant evidence, forming a critical foundation for the Kokatha case, in reaching his conclusion that the case must be dismissed. 396 The answers of the State and other parties to the contentions of the Kokatha People are not persuasive. 397 First, I do not accept that the errors described above may be characterised as immaterial because, at best, they involve evidence of Kokatha connection to Lake Torrens in the early 1980s rather than thereafter and that, either way, the connection is post-sovereignty. The problem with this submission is that it fails to recognise that the primary judge reasoned, at least in material part, by reference to the apparent lack of material supporting connection in the 1980s where he would have expected such material to emerge if it existed (see [724] and [719]). The fact that such material did emerge, when taken with the overlooked oral evidence said to have been learned from elders (who, it was at least reasonably open to infer, would have learnt such things themselves before the 1980s), means that the process of reasoning miscarried. 398 Second, it is true that, apart from Dr Willis, the anthropologists did not accept the existence of a traditional Kokatha connection to Lake Torrens. It is apparent from [194] that the primary judge placed "significant weight" on his own interpretation of the ethnographic material. It is also apparent from [724] that the primary judge's expectations of what Kokatha informants would have disclosed about Lake Torrens in the 1980s had they had a traditional connection to the Lake formed a critical part of his reasoning process. As a result, any failure to consider important parts of the ethnographic material and oral evidence which, on its face at least, demonstrated such a traditional connection is likely to be, and in this case was, material. This is so irrespective of the fact that other evidence before the primary judge, of the westward migration of the Kokatha post-sovereignty and of the presence of the Adnyamathanha and Barngarla Peoples immediately to the west of Lake Torrens at and after sovereignty until at least the 1940s, were capable of supporting rejection of the Kokatha People's claim. Unless no other conclusion was reasonably open on the evidence, the issue for an appeal is not the reasoning process that a judge might have used to reach the same result, but the reasoning process actually used. It cannot be (and was not) suggested that, on the evidence, the claim of the Kokatha People was bound to fail. As such, material errors in the primary judge's process of reasoning cannot be disregarded because of the availability of other evidence which is said to support the same result. 399 Given these conclusions I do not consider it necessary to address that part of the appeal of the Kokatha People to the effect that the primary judge ought to have considered making a determination in their favour at least in respect of Andamooka Island. As discussed, partial success is a possibility inherent in any native tile claimant application. The primary judge was not bound to advert to that possibility unless it was expressly sought by the Kokatha People. Mere acceptance by the Kokatha People of the possibility of partial success is not sufficient to give rise to an error by the primary judge in the present case in not separately considering whether the Kokatha People's claim could succeed in respect of only Andamooka Island. 400 For these reasons I consider that the Kokatha People's appeal must also be allowed and the agreed orders for remittal made. I certify that the preceding one hundred and forty-three (143) numbered paragraphs are a true copy of the Reasons for Judgment herein of the Honourable Justice Jagot.

WHITE J: 401 I agree with the orders proposed by Reeves J with respect to each appeal. I also agree with his reasons and, subject to one matter, do not wish to add to them. 402 In order to succeed on the appeal, the appellants had to demonstrate error by the primary judge. It was not sufficient for them to show that other inferences or conclusions may have been open on the evidence if the inferences drawn, or conclusions reached, by the primary judge were reasonably open. The task of this Court was to conduct a real review of the trial and of the primary judge's findings with a view to determining whether error was established: Fox v Percy [2003] HCA 22; (2003) 214 CLR 118 at [25]. In doing so, this Court had to recognise the advantages of the primary Judge, arising, amongst other things, from his Honour having seen and heard the evidence given. Even when the judge's findings were credibility-based, this Court could intervene if those findings were inconsistent with "incontrovertible" evidence, "glaringly improbable" or "contrary to compelling inferences" (Fox v Percy at [28]-[29]). 403 I have reviewed the evidence, particularly the evidence to which the parties referred the Court, with these principles in mind. For the reasons given by Reeves J, I consider that none of the appellants has demonstrated error of the kind warranting (or permitting) appellate intervention. I certify that the preceding three (3) numbered paragraphs are a true copy of the Reasons for Judgment herein of the Honourable Justice White.

SCHEDULE OF PARTIES SAD 249 of 2016 Respondents Fourth Respondent: BHP BILLITON OLYMPIC DAM CORPORATION PTY LTD Fifth Respondent: KELARAY PTY LTD Sixth Respondent: OZ EXPLORATION PTY LTD Seventh Respondent: OZ MINERALS CARRAPATEENA PTY LTD Eighth Respondent: OZM CARRAPATEENA PTY LTD Ninth Respondent: DOUGLAS MAXWELL GREENFIELD

SAD 250 of 2016 Respondents Fourth Respondent: BHP BILLITON OLYMPIC DAM CORPORATION PTY LTD Fifth Respondent: KELARAY PTY LTD Sixth Respondent: OZ EXPLORATION PTY LTD Seventh Respondent: OZ MINERALS CARRAPATEENA PTY LTD Eighth Respondent: OZM CARRAPATEENA PTY LTD Ninth Respondent: DOUGLAS MAXWELL GREENFIELD

SAD 251 of 2016 Respondents Fourth Respondent: BHP BILLITON OLYMPIC DAM CORPORATION PTY LTD Fifth Respondent: KELARAY PTY LTD Sixth Respondent: OZ EXPLORATION PTY LTD Seventh Respondent: OZ MINERALS CARRAPATEENA PTY LTD Eighth Respondent: OZM CARRAPATEENA PTY LTD Ninth Respondent: DOUGLAS MAXWELL GREENFIELD

ralian National Railways Commission (1993) 177 CLR 472 Fox v Percy (2003) 214 CLR 118; [2003] HCA 22 Griffiths v Northern Territory (2007) 165 FCR 391; [2007] FCAFC 178 Gumana v Northern Territory of Australia (2005) 141 FCR 457; [2005] FCA 50 Harrington-Smith on behalf of the Wongatha People v Western Australia (No 9) (2007) 238 ALR 1; [2007] FCA 31 Holloway v McFeeters (1956) 94 CLR 470 Kokatha People v State of South Australia [2007] FCA 1057 Lake Torrens Overlap Proceedings (No 3) [2016] FCA 899 Mabo v Queensland (No 2) (1992) 175 CLR 1 Members of the Yorta Yorta Aboriginal Community v State of Victoria (2002) 214 CLR 422; [2002] HCA 58 Minister for Immigration and Multicultural Affairs v Jia Legeng (2001) 205 CLR 507; [2001] HCA 17 Munn (for and on behalf of the Gunggari People) v Queensland [2002] FCA 486 Rawson Finances Pty Ltd v Federal Commissioner of Taxation (2013) 93 ATR 775; [2013] FCAFC 26 Robinson Helicopter Company Inc v McDermott (2016) 331 ALR 550; [2016] HCA 22 Sampi (on behalf of the Bardi and Jawi People) v Western Australia (2010) 266 ALR 537; [2010] FCAFC 26 Seltsam Pty Ltd v McGuiness (2000) 49 NSWLR 262; [2000] NSWCA 29 Starkey v State of South Australia [2014] FCA 924 State of Western Australia v Ward (2002) 213 CLR 1; [2002] HCA 28 Warren v Coombes (1979) 142 CLR 531 Wik Peoples v State of Queensland (1994) 49 FCR 1 Yarmirr v Northern Territory (No 2) (1998) 82 FCR 533

At a glance

Court

Decision date

Before

Source

Judgment (48 paragraphs)

Parties

Legislation Cited (2)

Cases Cited (64)