Finding HMAS Sydney

1
Australian Science in the search for Kormoran and
Sydney
byKim Kirsner and John Dunn
2

• Summary
• This story is about Australian science, and the
  contribution that two Australian scientists made
  to the search for Kormoran and Sydney.
• Working in collaboration with the Finding Sydney
  Foundation (FSF) from 2001, Kim Kirsner and John
  Dunn identified the position of Kormoran down to
  2.7 nautical miles, in 2004. They also defined
  appropriate search boxes for both Kormoran and
  Sydney in 2005.
• The material was published in FSF submissions to
  the Commonwealth, the states, the RAN, and
  corporate and private donors between 2004 and
  2007. Kirsner and Dunn also published a critical
  article on the WEB (Kirsner Dunn, 2004).
• In this account Kirsner and Dunn outline the
  steps used to achieve these objectives. The
  procedure is transparent and can be applied to
  other search projects.
• The Heat Map depicted in the following images was
  the final step in our analysis of the location of
  Kormoran.

3
Kirsner and Dunn (2004) Heat Map based on
mathematical decision model. The model
integrated information from nine independent
sources.
Kirsner Dunn (2004) Predicted position of
Kormoran 26 04 South 111 02 East
Error Distance between Predicted and Observed
positions for Kormoran 2.7 nautical miles
26 South
FSF / Mearns (2008) Observed position of
Kormoran 26 06 South 111 04 East
111 East
26 30 South
4
The Ships
5
Kormoran
6
Kormoran was a German raider, heavily armed for
close range combat with lightly armed merchant
vessels but disguised as an allied merchant ship.
Kormoran left Germany on December 23rd 1940, and
sank eleven merchant ships prior to her
engagement with HMAS Sydney. Kormoran was
steaming north along the coast of Western
Australia when she sighted HMAS Sydney, and
turned west toward to avoid combat.
7
HMAS Sydney
8
HMAS Sydney was armed and equipped for long range
combat but with little or no advantage over
Kormoran at the range at which the battle
unfolded. HMAS Sydney was en route from Sunda
Strait to Fremantle when contact was made with
Kormoran. Sydney followed and gradually closed in
on Kormoran until, at a range of less than one
nautical mile, Kormoran opened fire and both
vessels were destroyed.
9
Oceanography Workshop (1991)
10
In 1991 Mike McCarthy and Kim Kirsner coordinated
a workshop to determine the most likely area for
Kormoran. The key speakers were oceanographers or
search and rescue experts. Their analyses
converged on the area supported by the Kormoran
survivors, near 26 South 111 East Oceanography
cannot be used to define a precise site
uncertainty in the direction and velocity of
current and wind is too large. Despite the fact
that the oceanographic evidence provided no
support for a wreck near the Abrolhos Islands,
200 nautical miles from 26 South 111 East, the
McCubbin Parliamentary Inquiry (2001) failed to
reject map dowsing and oral history claims for
the Abrolhos, and the RAN subsequently
implemented searches in the area The following
figure shows the submissions by the professionals
to the1991 Oceanography Workshop
11
Map of the region
26 South
111 East
12
26 South
111 East
13
26 South
111 East
14
26 South
111 East
15
26 South
111 East
16
Cognitive Science
17
What is Cognitive Science? Cognitive Science is
concerned with memory, decision-making,
neuroscience, linguistics, artificial
intelligence, attention, skill acquisition and a
host of related problems including human error.
Working together and independently, Kim Kirsner
and John Dunn have published more than 150
refereed articles and chapters in the area of
Cognitive Science, and held approximately 15
grants from the Australian Research Council. Kim
Kirsner is an elected Fellow of the Australian
Academy of Social Science.
18
Why wreck-hunting? Wreck-hunting is complicated.
Navigation records usually provide the best
source of information. In the case of Kormoran
however the critical information survived the
battle and a week at sea in the heads of the crew
- the domain experts - and memory and forgetting
were critical. In other cases oceanography might
be more important, and even oral history might be
relevant. Because most of the information in the
archives compiled in 1941 depended on memory,
Kirsner and Dunns expertise was appropriate.
Henceforth we will refer to this material as the
Kormoran Database. Where the Kormoran Database is
concerned, Kirsner and Dunn are domain
experts. The material from the diaries including
that of Detmers was less reliable than the
Kormoran Database because it was constructed
months or even years after the battle. Detmers
for example produced no fewer than five different
positions across the RAN interrogations and diary.
19
Cognitive Science The Kormoran Database
20
The 1941 interviews and interrogations
constituted the primary source. Why? First, they
were reported in November and December 1941,
months or possibly years prior to the preparation
of the diaries and coded dictionaries. Second,
as survival was critical following disembarkation
from Kormoran, it may be assumed that critical
information was distributed among the five
lifeboat crews, and not the property of Detmers
alone Third, unless the reports associated with
Detmers coded dictionary involve different dot
patterns with the same information and there is
no suggestion that this is the case - it may be
assumed that their reliability reflects only the
dot copying skills of Detmers or Detmers
secretary
21

• What is in the Kormoran Database?
• The Database included the following reports based
  on interrogations with survivors conducted
  between November 26th and December 10th 1941
• Eighteen reports referred to 26 S 111 E
• Ten reports referred to either 26 S or 111 E
• Six reports referred to distances from land,
  60, 120 and 150 nm
• Two reports stated that one lifeboat sailed
  150 nm NE from Kormoran to the coast
• One report stated that the battle occurred 160
  nm SW of NW Cape
• Four reports referred to 130 nm SW of Shark
  Bay
• Fourteen reports referred to 2634, 26 32,
  26 31 or 26 30 S 111 E.
• Additional reports involve outliers and
  singletons

22
A glance at the Kormoran database (3 of the DB)
SOURCE   26S 111E 26S 110E 26S 108E 26S 11E 26S 24S 111E 26S 111-21E 26S 111-40E
Ahlbach 26S 111E 1              
Bretschneider 26S 111E 1              
Bretschneider 25S 111E                
Bunjes 160 nm SW of North West Cape (Use Cape Cuvier)                
Bunjes 120 NM FROM THE COAST (DRAW FROM 25S TO 27S)                
Bunjes 26S (draw line from 110E to 112E         1      
23
Chaos The eyewitness accounts in the Australian
Archives include hundreds of descriptions of
Kormoran, her voyage and the battle, Seventy of
these reports include information about the
location of the battle The reports referred to a
vast area, from the latitude of Carnarvon in the
North, to Fremantle in the South, and from 60 nm
to 300 nm from the coast. How reliable are they?.
24
Geographical description of the Kormoran
Database?
The chart hints at both the spread and the
concentration of the reports from the Database.
The reports tend to concentrate in the vicinity
of 26 South 111 East, particularly if
consideration is restricted to crew who were in
a position to know However the reports ranged
from the latitude of Carnarvon to the latitude of
Fremantle, and from 60 to 300 nm offshore
26S 108E
26 South
26S 11E
120 nm SW of Freo
130 nm W of Perth
100 nm W of Freo
125 nm SW of Freo
20 nm SW of Freo
111 East
25
Zipf's Law Do the reports reflect one carefully
rehearsed story, as many critics claimed, or
do they reflect ignorance, as if no-one had any
idea at all, or do they reflect expert knowledge,
about location, accompanied by random
error? Zipf's law states that given some corpus
of natural language utterances (or memory
reports), the frequency of any word is inversely
proportional to its rank in the frequency table.
The most frequent word will occur approximately
twice as often as the second most frequent word,
which occurs twice as often as the fourth most
frequent word, etc. The same relationship occurs
in many other rankings, including memory reports.
In the memory case, it is assumed that the spread
of the reports involves random error
26
Concentration of reports involving 26South 111
East From Kirsner, Norman Dunn, 2003
27
The frequency distribution is consistent with the
assumption that the Kormoran database consists of
random errors around a single position From
Kirsner, Norman Dunn, 2003, Finding Sydney
Foundation, 2005
28
Cognitive Science How reliable is the Kormoran
Database?
George Kingsley Zipf  Lecturer at Harvard
University (1902 1950)
29
Zipfs Law Zipf's law states that given some
corpus of natural language utterances (or memory
reports), the frequency of any word is inversely
proportional to its rank in the frequency table.
Thus the most frequent word will occur
approximately twice as often as the second most
frequent word, which occurs twice as often as the
fourth most frequent word, etc. The same
relationship occurs in many other rankings,
unrelated to language, such as the population
ranks of cities in various countries, corporation
sizes, income rankings, and memory reports -
given a single source of information. In the
memory case it is assumed that the spread or
reports involves random error.
30
Use of Zipfs Law to assess reliability?
Blue triangles are reports from Kormoran
survivors
Log Frequency
One frequent report - 26 S 111 E
Many infrequent reports, like 120 nm SW of
Fremantle
Log-Log coordinates
Log Rank
31
Red triangles are reports from The War of the
ghosts, a memory study by Bartlett (1932)
Log Frequency
Grey circles are from simulation based
on Kormoran survivors reports
The data conform to Zipf's law to the extent that
the plot is linear in Log-Log coordinates
Log Rank
32
Comment The data conform to Zipf's law to the
extent that the plot is linear in Log-Log
coordinates. The function relating Log Frequency
and Log Rank is linear, and the results are,
therefore, consistent with the assumption that
the survivors were telling the truth Further
support for this interpretation came from the
fact that Bunjes, an anti-Nazi, provided three
different types of information that produced
approximately the same solution. More evidence,
if required, came from the extra-ordinary number
of survivors who pointed to 26 South 111 East.
Many of these crew were in a position to
know. The analysis does not address questions
about the individual reports. It demonstrates
that the set of reports is reliable as a
system The following figure shows how we used
and interpreted Zipfs argument
33
Cognitive Science Selection of Source Arguments
34
Comment The following analysis of source
selection is a summary. The detailed arguments
will be published elsewhere. Integration
produced a heat map, where probability follows
heat, and the central position was 26 04
South 111 02 East, just 2.7 nautical miles from
the position of wreck of Kormoran. The material
was provided to the FSF in 2004. The FSF
reviewed, adopted and published our prediction
for Kormoran in 2005 and 2007. The material was
used in presentations to the Commonwealth, the
States, the RAN and corporate and private
donors. Further collaboration involving the
authors and Director Bob King produced the search
boxes for Kormoran and Sydney. The material was
also provided to and discussed with Mearns, by
the authors in 2004, and by the FSF, in 2005.
35
Discard outliers
26S 108E
26S 11E
120 nm SW of Freo
130 nm W of Perth
100 nm W of Freo
125 nm SW of Freo
20 nm SW of Freo
36
Accept reports referring to 26 South 111 East.
Discard reports referring to only one of these
values
37
Accept 26 South 111 East
38
Accept reports that one life-boat sailed 150 nm
NE from Kormoran The calculated position was
included in the analysis
39
Accept signal (from Kormoran) referring to 26
S 111 15 E Longitude was included in the
analysis, and alternative latitudes were tested
to determine the best fit with the balance of the
sources.
40
Three reports involved distance from land.
Accept reports that battle occurred 120 nm from
land. Discard reports that battle occurred 60 nm
or 150 nm from land
41
Accept report that battle occurred 160 nm SW of
NW Cape as 160 nm SW of Cape Cuvier
42
Accept reports from Detmers referring to 2634,
26 32S, 26 31 and 26 30 S and 111 E
43
Accept Detmers report that Kormoran would (have
been) off Shark Bay four hours after contact with
HMAS Sydney
44
Accept reports that battle occurred 130 nm
South-West of Shark Bay
45
Include Circle of Equal Distance for two
life-rafts calculated by Dunn and Kirsner
(2001) The analysis was originally implemented to
test the navigation claims made by Whittaker
(that action took place off the Abrolhos
Islands. The analysis was mathematical and
involved no assumptions about wind or
current. Given equal velocity for 84 107 hours
by two life-rafts, what is the range of positions
from which they could have originated? The arc
passes within 2 nm of the wreck of Kormoran
46
Summary of reports accepted for
analysis Integration of these reports involved a
mathematical decision model that gave equal
weighting to all of the sources An alternative
analysis involving weighting of the individual
sources was not implemented.
47
Cognitive Science Integration based on Human
Judgement (Kirsner Dunn, 1998)
48
Kirsner Dunn (1998) Submission to McCubbin
Parliamentary Inquiry
Kirsner and Dunn (1998) Position given as,
Kormoran sank a few miles to the north of 26
15 South 111 00 East.
49
FSF / Mearns (2008) Observed position of Kormoran
given as 26 06 South 111 04 East
Kirsner and Dunn (1998) Position given as,
Kormoran sank a few miles to the north of 26
15 South 111 00 East. Error 11 nm
50
FSF / Mearns (2008) Observed position of Kormoran
given as 26 06 South 111 04 East
Kirsner Dunn (1998) Error Distance between
Predicted and Observed 11 nm
Kirsner and Dunn (1998) Predicted Position given
as, Kormoran sank a few miles to the north of
26 15 South 111 00 East.
51
The FSF referred to Kirsner Dunn (1998) in
discussion but it was not recommended to or
published by the Foundation. The FSF reviewed
and adopted Kirsner and Dunn (2004), a solution
based on a mathematical decision model.
52
Cognitive Science Integration based on
Mathematical Decision Model (Kirsner and Dunn,
2004)
53
Heat Map Integration (Kirsner Dunn, 2004)
The FSF reviewed and adopted Kirsner and Dunn
(2004), a solution based on a mathermatical
decision model.
54
Kirsner Dunn (2004) Heat Map used to define
location of Kormoran
Kirsner and Dunn (2004) was distributed to and
reviewed by Mearns (December, 2004) and by Mearns
and the FSF (May, 2005a) The critical arguments
were adopted, extended and published by the FSF
(2005b, 2007).
111 East
26 South
26 30 South
111 East
55
Kirsner Dunn (1998) Cognitive Reconstruction
with mathematical decision model
Kirsner and Dunn (2004) Predicted position of
Kormoran given as 26 04 South 111 02
East. The argument for this position was reviewed
and adopted by the FSF.
111 East
26 South
26 30 South
111 East
56

Kirsner and Dunn (2004) Predicted position of
Kormoran given as 26 04 South 111 02 East
111 East
FSF (2005) Recommended search box for Kormoran
26 South
26 30 South
111 East
57

Kirsner and Dunn (2004) Predicted position of
Kormoran given as 26 04 South 111 02 East
111 East
FSF (2005) Recommended search box for Kormoran
26 South
FSF / Mearns (2008) Observed position of Kormoran
given as 26 06 South 111 04 East
26 30 South
58

Kirsner and Dunn (2004) Predicted position of
Kormoran given as 26 04 South 111 02 East
111 East
FSF (2005) Recommended search box for Kormoran
26 South
FSF (2008) Observed position of Kormoran given as
26 06 South 111 04 East
26 30 South
Distance between Predicted and Observed positions
of Kormoran 2.7 nautical miles
111 East
59
History of Search Definition(2001-2008)
60
Finding Sydney Foundation Maps and Plans (June,
2001)
61
Finding Sydney Foundation (HMAS Sydney Search Pty
Ltd) Kormoran Recommended Search Position (June
4th, 2001)
FSF / Mearns (2008) Kormoran Observed Position
Used with permission of the FSF
62
Mearns Search area and positions recommended
during presentation to West Australian Maritime
Museum (November, 2004)
63
Mearns (2004a, 2004b)
Mearns / FSF (2008) Position of Kormoran
Positions and search area recommended by Mearns
during presentations to ? The West Australian
Maritime Museum (November, 2004a) ? Directors of
the FSF (November, 2004b)
64
Finding Sydney Foundation Meeting between Mearns
and Kirsner (November, 2004)
Alan Puckett
65
Convenor and Recorder Director Keith Rowe
(December, 2004)
Positions nominated by Kirsner and Dunn (2004)
and Mearns (2004), together with the Observed
position of Kormoran
Mearns 1
Kirsner Dunn (2004)
26 South
Kormoran Observed
Mearns 2
Mearns 3
111 East
66
Finding Sydney Foundation Map and Search Plans
(2005)
67
Finding Sydney Foundation (2005)
Finding Sydney Foundation (2005) Map prepared and
used by FSF in presentations to the Commonwealt,
the states, the RAN, and private and corporate
donors
Used with permission of FSF
68
Red positions / boxes published by Kirsner Dunn
(2004) or attributed to Kirsner Dunn (FSF,
2005).
69
Red positions / boxes published by Kirsner Dunn
(2004) or attributed to Kirsner Dunn (FSF,
2005).
70
Red positions / boxes published by Kirsner Dunn
(2004) or attributed to Kirsner Dunn (FSF,
2005).
71
Red positions / boxes published by Kirsner Dunn
(2004) or attributed to Kirsner Dunn (FSF,
2005).
72
FSF (2008) Observed position of Kormoran
73
Finding Sydney Foundation Map and Search Plans
(2007)
74
Finding Sydney Foundation (2007)
Finding Sydney Foundation (2007) Map prepared and
used by FSF for submission to the Commonwealth
75
(No Transcript)
76
(No Transcript)
77
(No Transcript)
78
In-Water Search for Kormoran
Alan Puckett
79
Kirsner Dunn (2004) Heat Map used to define
location of Kormoran
80
Kirsner Dunn (2004) / FSF (2005) Predicted
position and Search Box for Kormoran
Kirsner Dunn (2004) / FSF (2005) Predicted
position for Kormoran Search box for Kormoran
81
Mearns (2008) Search Box
Mearns (2008) In-water Search box for Kormoran
82
Mearns (2008) Search Track 1
Mearns (2008) In-water Search box for Kormoran
83
Mearns (2008) Search Track 2
Mearns (2008) In-water Search box for Kormoran
84
Mearns (2008) Search Track 3
3
Mearns (2008) In-water Search box for Kormoran
85
Mearns (2008) Search Track 4
Mearns (2008) In-water Search box for Kormoran
86
Mearns / FSF (2008) Discovery of Kormoran
Mearns (2008) In-water Search box for Kormoran
FSF / Mearns (2008) Location of Kormoran
87
Cognitive ScienceSearch for Kormoran
Performance Review
Alan Puckett
88
How precise is 2.7 nautical miles (nm)? The
oceanographers and search and rescue solutions
ranged from 15 to 40 nm from Kormoran. The
final solution offered by the historians was 11
nm from Kormoran, close enough to support
discovery in an in-water search. The solutions
published by Mearns (Mearns, 2004a) or attributed
to Mearns (FSF, 2005 2007) prior to the in-water
search ranged from 17 to 45 nm from Kormoran.
The performance of the Cognitive Scientists
reveals a traditional learning curve (See
Speelman Kirsner, 2005). Thus, ? Using a
Search and Rescue model Kirsner (Kirsner, 1991,
Kirsner, 1992 and Kirsner and Hughes (1993)
identified positions 20 nautical miles from
Kormoran. ? Using cognitive reconstruction
without a mathematical decision model, Kirsner
and Dunn (1998a) identified a position 10
nautical miles from Kormoran (a few miles north
of 26 15 South 111 East). ? Using cognitive
reconstruction with a mathematical decision
model, Kirsner and Dunn (2004) identified a
position 3 nautical miles from Kormoran.
89

Oceanographers, Marine Surveyors and Search and
Rescue experts
Distance between predicted and observed
Distance (nm) between predicted and observed
positions of Kormoran for solutions offered by
Oceano-graphers, Marine Surveyors and Search and
Rescue experts. The relevant publications are for
Steedman McCormack (1991), Hughes (1991),
Penrose Klaka (1991), Anderson (1997), Hughes
(2001) and Griffin (2008)
90

Historians
Distance between predicted and observed
Distance (nm) between predicted and observed
positions of Kormoran for solutions offered by
the Historians. The relevant publications are
from Olson (1995) and Olson, Hore, Goldsmith and
Vickridge (2001)
91

Mearns
Mearns (2004-2007) Positions published by Mearns
(Mearns, 2004a, 2004b), or attributed to Mearns
by the FSF (2005, 2007).
Distance between predicted and observed
Distance between predicted and observed
92
Distance between predicted and observed position
for Kormoran for Kirsner / Kirsner and Dunn

• Distance (nm) between predicted and observed
  positions of Kormoran for solutions offered by
  Kirsner and his colleagues.
• The contributions fall into three phases based
  on
• ? Search and Rescue models (1991-1993)
• ? Cognitive Reconstruction without a mathematical
  decision model (1997-1998)
• Cognitive Reconstruction with a mathematical
  decision model (2004)
• SH Sam Hughes

93
Distance between predicted and observed position
for Kormoran for Kirsner / Kirsner and Dunn
In 2004, Kirsner and Dunn recommended 26 04
South 111 02 East as the target for Kormoran,
and they made no adjustment to that position
despite extensive criticism of their work. The
FSF adopted this position, and included it in
numerous presentations to the Commonwealth, the
states, the RAN, and private and corporate donors
between 2005 and 2007.
94
Phase 3 (2004) Based on Cognitive Science with
Mathematical Decision Model - Position is the
only one offered to or used by the Finding Sydney
Foundation
95

• Skill Acquisition
• Kirsner and Dunns contribution to search
  definition reflects skill acquisition, as they
  moved down the learning curve, from novices in
  1991 to experts in 2004. The phases are described
  below
• ? Phase 1 (1991-1993) Work based on Search and
  Rescue model involving collaboration with Sam
  Hughes (Kirsner, 1991 Kirsner, 1992a 1992b
  Kirsner and Hughes, 1993)
• ? Phase 2 (1997-1998) Based on Cognitive
  Science without a Mathematical Decision Model
  (Kirsner, 1997a Kirsner Dunn, 1998a)
• ? Phase 3 (2004) Based on Cognitive Science
  with a Mathematical Decision Model (Kirsner
  Dunn, 2004)
• Learning curves follow similar functions for
  laboratory as well as real world tasks such as
  probability of death in aerial combat, surgical
  skill, industrial production rates, and a host of
  other tasks.
• See Speelman and Kirsner (2005) for a review of
  Learning Curves

96
The Search for HMAS Sydney
97

• In 1991, McCarthy proposed that the search for
  HSK Kormoran should precede the search for HMAS
  Sydney because more information was available
  about the location of Kormoran.
• In 1997 Rear Admiral Holthouse (ret), an engineer
  with combat experience, cast doubt on the
  Abrolhos Island claim for HMAS Sydney, arguing
  that the absence of survivors and signals left
  little room for a 20-40 hour voyage after the
  engagement.
• HMAS Sydney was on fire after the battle. The
  Kormoran survivors reports provided estimates of
  the movement of HMAS Sydney following the battle,
  and enabled Kirsner (1997) to plot changes in the
  rate of separation between the vessels from 1800
  hours to the last reported sighting some four
  hours later. A figure depicting the rate of
  separation is shown on the following page
  (Kirsner, Norman and Dunn, 2003).
• The following image shows the product of this
  analysis. The 1997 analysis provided a prediction
  for the position of the wreck of HMAS Sydney
  relative to Kormoran. The search area for HMAS
  Sydney was specified in FSF (2005)

98
Each circle shows one report from a survivor
indicating the time and the estimated distance of
HMAS Sydney. The general direction offered with
these reports was South-East. from Kirsner
(1997) and Kirsner, Norman and Dunn, 2003
99
Finding Sydney Foundation (2005)
Finding Sydney Foundation (2005) Map prepared and
used by FSF for presentations to the
Common-wealth, the states, the RAN, and the
private and corporate donors
100
Kirsner Dunn (2004) Predicted position of
Kormoran
101
Kirsner Dunn (2004) Predicted position of
Kormoran
FSF (2005) Recommended search quadrant for HMAS
Sydney given predicted position of Kormoran.
102
FSF (2008) Observed position of Kormoran
Kirsner Dunn (2004) Predicted position of
Kormoran
FSF (2005) Recommended search quadrant for HMAS
Sydney given predicted position of Kormoran.
103
Kirsner Dunn (2004) Predicted position of
Kormoran
FSF (2008) Observed position of Kormoran
FSF (2005) Adjusted search quadrant for HMAS
Sydney given Observed position of Kormoran.
FSF (2005) Recommended search quadrant for HMAS
Sydney given Predicted position of Kormoran.
104
Kirsner Dunn (2004) Predicted position of
Kormoran
FSF (2008) Observed position of Kormoran
FSF (2005) Adjusted search quadrant for HMAS
Sydney given Observed position of Kormoran.
FSF (2005) Recommended search quadrant for HMAS
Sydney given Predicted position of Kormoran.
FSF/Mearns (2008) Observed position of HMAS Sydney
105
Conclusions
106
Australian Science Between 1991 and 2004 Kirsner
and Dunn systematically revised and refined their
approach to search definition, and reduced their
error the distance between the predicted and
observed positions for Kormoran - from 20
nautical miles to lt 3 nautical miles. In 2004
they used seven reports from the Kormoran
survivors, one signal from Kormoran, and one
life-raft drift analysis together with a
mathematical decision model to predict the
location of Kormoran. The successful application
of transparent scientific procedures demonstrates
that Cognitive Science has a significant role to
play in wreck-hunting and other problems. The
interested reader is referred to Kirsner and Dunn
(2004) for a more detailed account of the
analyses used to predict the location of
Kormoran.
107
Precedent Precedent is the single most important
principle in scientific discovery. The position
identified for Kormoran by Kirsner and Dunn in
2004 was correct, to within spitting distance,
as Director Keith Rowe put it in an email in
2008. The material was provided on a pro bono
basis to the FSF and Mearns in 2004. It was
reviewed and adopted by Mearns and the FSF in
2005, and used by the Foundation in presentations
and submissions to the Commonwealth, the states,
the RAN and corporate and private donors in 2005,
2006 and 2007. In 2008, as the MV Geosounder
departed Geraldton, Mearns produced a search box
that included Kormoran for the first time.
108
The Finding Sydney Foundation (FSF) FSF
publications demonstrate that it had adopted an
appropriate location for Kormoran as early as
2001, the year in which the Foundation was
established. By 2004, the FSF search definition
team (Kim Kirsner, John Dunn, and Bob King) had
provided the Foundation with a precise target
just 2.7 nm from the wreck of Kormoran, together
with a search box and quadrant that included
Kormoran and Sydney respectively. The search box
recommended by the FSF team for Kormoran in 2004
was 400 square nautical miles. Kormoran is more
or less dead centre in that search box. The
search box adopted by Mearns in 2008 was 1800
square nautical miles. Mearns search box
included that of Kirsner and Dunn (2004). The
search quadrant recommended by the FSF team for
the search for HMAS Sydney in 2005 was 570 square
nautical miles. HMAS Sydney was in that quadrant.
109
Authorship
Alan Puckett
110

Kim Kirsner received bachelors and doctoral
degrees from the University of London. Following
a post-doctoral fellowship at the University of
Toronto, Kim joined the University of Western
Australia. Kim was elected Fellow of the Academy
of Social Science in 1997. Kim is currently
Professor, Centre for Experimental and
Regenerative Neuroscience at the University of
Western Australia. Kim has published numerous
books, book chapters, refereed articles and
conference proceedings on memory, aphasia,
bilingualism, implicit memory, language
production, skill acquisition and cognitive
reconstruction Contact pkirsmer_at_bigpond.net.au
111

John Dunn received his BSc (Hons) and doctoral
degrees from the University of Western Australia.
John has taught at the universities of
Queensland, Western Australia and Adelaide, where
he is currently Professor and Deputy Head of the
School. John has published numerous books, book
chapters, and refereed articles and conference
proceedings in visual attention, cognitive
psychology, memory including work on
methodological issues in neuropsychology,
decision-making in complex environments, and
mathematical models.
112
Acknowledgements
Alan Puckett
113
Significant credit for the successful search for
HMAS Sydney must go to Dr Mike McCarthy, Curator
at the Western Australian Maritime Museum. Mike
served the Finding Sydney family with commitment,
courtesy, and humour for more than 30 years. The
authors wish to acknowledge the contributions
made by LT COM Sam Hughes (Rtd), Bob King OAM
(Consultant, ex-Director, FSF), Ted Graham AM
(Chair and Director, FSF), Kathryn Hird
(Professor and Associate Dean, School of
Medicine, University of Notre Dame Australia) and
Kevin Durkin (Professor, School of Psychology,
Strathclyde University). The following people
played important and often critical roles in
search definition Graham Anderson, Ian Anderson,
R. Goldsmith, David Griffin, Rear Admiral David
Holthouse RAN AO, Peter Hore, Sam Hughes LCDR
RANR (Rtd) , Bob King OAM, J. Klaka, the late
Gordon Laffer, J. McCormack, Ean McDonald LCDR
RANR (Rtd), David Mearns, Wes Olsen, Alan Pearce,
John Penrose, Keith Rowe, Ray Steedman, Geoff
Vickridge, and Barbara Winter.
114
Grateful acknowledgement is made to the following
for contributions to the WEBSITE Richard Bone
(ELK Software), Mark Comeadow (IT consultant,
Cisco, London) Mark Reuten (IT Consultant,
Melbourne), Marziya Mahommedali (Multi-Media
expert, Curtin University of Technology), Doug
Robb (Information Analyst, School of Psychology,
University of Western Australia, Mike Ridout
(Communications Director, CRC for Spatial
Information) and Ming Zhu (Geo-scientist, Royal
Melbourne Institute of Technology. The in-water
search including in particular the sonar and
photographic work were implemented with
extraordinary skill and perseverance under
difficult conditions. David Mearns, Art Wright,
the Williamson team, and the crew of the MV
Geosounder are to be congratulated on their
contribution to the search. Grateful
acknowledgement is also made to the following for
their specialist contributions Kim Heitman
(Senior Legal Officer, University of Western
Australia) Ron Birmingham QC (Ex-Director, FSF)
and COMMODORE Bob Trotter OAM (Director FSF).
115

• Numerous individuals made significant
  contributions to search definition. A brief list
  must include engineers, historians,
  hydrographers, journalists, marine surveyors,
  oceanographers, search and rescue experts and
  serviceman.
• The successful search for HSK Kormoran and HMAS
  Sydney was driven by the Finding Sydney family
  (FSf). The FSf was not on the dais in
  Geraldton, in 2008 they may have made only
  fleeting or significant contributions to the
  search they did not always see eye to eye with
  each other they were not looking for medals or
  recognition but they each held an oar for a
  moment or two and rowed the FSf in the right
  direction.
• The authors also wishes to acknowledge support
  from the Australian Research Council, the
  Australian War Memorial, the Western Australian
  Maritime Museum and the University of Western
  Australia.

116
References / Bibliography
Work in progress
Alan Puckett
117
Anderson, I. (1997). Determination of likely
position of HMAS SYDNEY and HSK KORMORAN. Fugro
Reference Number HY11076. Bartlett, F. C. (1932
reprint 1964). Remembering A study in
experimental and social psychology. London
Cambridge University Press. Detmers, T. (1959).
The Raider Kormoran. London William Kimber.
Dunn, J. Kirsner, K. (2001). Using a
'temporal traingulation' technique to recommend a
search area and starting point for the search in
the vicinity of the northern area. McCarthy, M.
(compiler). Submissions to the November 2001 HMAS
Sydney II Seminar. Finding Sydney Foundation
(2005). Search for HMAS Sydney II 2005/6
Solving Australias most enduring Maritime
Mystery. HMAS Sydney Search Pty Ltd. PowerPoint
presentation. Endorsed for use by the Finding
Sydney Foundation (2009) Finding Sydney
Foundation (2007). Status Report on Achievements
and Opportunities. Submission to the Commonwealth
of Australia. Endorsed for use by the Finding
Sydney Foundation (2009) Griffin, D. (2008).
Locating HMAS Sydney by back-tracking the drift
of two life-rafts. Final Report The search to
find andidentify the wrecks of HMAS Sydney II and
HSK Kormoran. Finding Sydney Foundation (Volume
2). Hughes, A.J. (1991). A possible solution
based on modern SAR Planning Techniques. In M.
McCarthy., K. Kirsner (compilers). The HMAS
Sydney Forum. Department of Maritime
Archaeology, Western Australia Maritime Museum
Report 52. Hughes, S. (2001). Locating HMAS
Sydney and HSK Kormoran by Computer Generated Net
Water Movement System. Submission to HMAS Sydney
Wreck Location Seminar, Fremantle,
2001.Johnstone, Crouch, Cowan, O'Driscoll (2003)
118
Kirsner, K. (1991). HSK "Kormoran" versus HMAS
Sydney - Converging operations in historical
analysis. In M. McCarthy, K. Kirsner
(compilers). The HMAS Sydney Forum. Department
of Maritime Archaeology, Western Australia
Maritime Museum Report 52. Kirsner, K. (1992).
The search for HSK Kormoran and HMAS Sydney A
Preliminary Analysis based on SAR Procedures.
The Great Circle Journal of the Australian
Association of Maritime History, 14,
88-104. Kirsner, K. (1992). HSK Kormoran versus
HMAS Sydney The 20th Century Maritime Mystery.
In R. Gardiner (Ed.). Warship 1992. Conway
Maritime Press. London Kirsner, K. Hughes, S.
(1993). HMAS Sydney and HSK Kormoran Possible
and Probable Search Areas. WA Maritime Museum
Report 71. Kirsner, K. (1997, February). The
Longest Search. Paper presented to the 55th
Anniversary Forum of the loss of HMAS
Sydney. Kirsner, K. (1997, September).
Organizational memory and the war of the
ghosts The archival search for the location of
HMAS Sydney. Paper presented to the 17th
International Conference of the Australian
Institute for Maritime Archaeology The Maritime
Archaeology of Long Distance Voyaging Iron and
Steam Shipwrecks, Western Australian Maritime
Museum. Kirsner, K. Dunn, J. (1998, May).
Feasibility of the search for HMAS Sydney and HSK
Kormoran Oceanographic and Cognitive Issues.
Submission to the Joint Standing Committee on
Foreign Affairs, Defence and Trade Defence
Sub-Committee Inquiry into the circumstances of
the sinking of HMAS Sydney. Volume 11,
2727-2742. Kirsner, K. Dunn, J. (1998, June).
A review of the oceanographic and cognitive
evidence that HSK Kormoran is at 2839S
11322E. Submission to the Joint Standing
Committee on Foreign Affairs, Defence and Trade
Sub-Committee Inquiry into the circumstances of
the sinking of HMAS Sydney. Volume 16, 4023-4058.
119
Kirsner, K. Dunn, J. (1998, June). Cognitive
problems associated with the use of oral history
data Supplement to Feasibility of the search
for HMAS Sydney and HSK Kormoran Oceanographic
and Cognitive Issues. Submission to the Joint
Standing Committee on Foreign Affairs, Defence
and Trade Defence Sub-Committee Inquiry into the
circumstances of the sinking of HMAS Sydney.
Volume 18, 4311-4313. Kirsner, K., Norman, C.
Dunn, J. (2003). HMAS Sydney Search Pty Ltd.
Presentation by Kim Kirsner, Cathy Norman and
John Dunn to the Directors of HMAS Sydney
Search. Kirsner, K. Dunn, J. (2004). The Search
for HSK KORMORAN and HMAS SYDNEY II A Cognitive
Perspective. UWA Site for Cognitive Analysis of
Archival, Historical and Memory Databases.
http//www.whereissydney.com/ Kirsner, K.
Dunn, J.C. (2008a). Kormoran Database. Submission
requested by Cole Commission of Inquiry.
Kirsner, K. Dunn, J.C. (2008b). Search
Definition in the search for Kormoran and Sydney
Triumph for Cognitive Science. Submission
requested by and submitted to the Cole Commission
of Inquiry. McCarthy, M. Kirsner, K. (1991,
November). The HMAS Sydney Forum. Department of
Maritime Archaeology, Western Australia Maritime
Museum Report 52. Mearns, D.L. (2004a). In
search of HMAS Sydney and HSK Kormoran.
Presentation at the West Australian Maritime
Museum. Mearns, D. L. (2008). Final Report.
Finding Sydney Foundation. Mearns, D.L. (2009).
The Search for Sydney. HarrperCollins Publishers.
Sydney.
120
Montgomery, M. (1981).Who sank the Sydney?
Sydney. Cassell. Olsen, W., Hore, P., Goldsmith,
R. Vickridge, G. (2001).Archival Record.
Workshop Report. HMAS Sydney Wreck Location
Seminar. Penrose, J.D. Klaka, K.P. (1991).
Notes on the movement of wreck material from the
area of the Sydney / Kormoran engagement. In
M. McCarthy, K. Kirsner (compilers). The HMAS
Sydney Forum. Department of Maritime
Archaeology, Western Australia Maritime Museum
Report 52. Steedman, R. McCormack, P. (1991).
Backtracking the lifeboats and floats a
Metaocean view. In M. McCarthy, K. Kirsner
(compilers). The HMAS Sydney Forum. Department
of Maritime Archaeology, Western Australia
Maritime Museum Report 52. Tuchman, B. W.
(1984). The March of Folly From Troy to Vietnam.
A Ballantyne Book, Random House. New York.
Winter, B. (1984).HMAS Sydney Fact, Fantasy and
Fraud. Brisbane Boolerong Press. Zipf, George K.
Human Behavior and the Principle of Least Effort.
Cambridge, Mass. Addison-Wesley, 1949. Reprint
ed., New York Hafner, 1965.
121
(No Transcript)