Annals of Tourism Research, Vol. 37, No. 2, pp.

315–332, 2010
0160-7383/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
Printed in Great Britain
www.elsevier.com/locate/atoures
doi:10.1016/j.annals.2009.10.011

COMMUNICATIONS IN ADVENTURE
TOUR PRODUCTS
Health and Safety in Rafting and Kayaking
Ralf Buckley
Griffith University, Australia

Abstract: Critical incidents involving communications on health and safety issues amongst
staff and clients are examined from 388 days of participant observations in 63 non-motorised
waterborne adventure tours in 19 countries. Good communications were critical in maintain-
ing client satisfaction and rescuing them from life-threatening dangers; poor communica-
tions put participants at unnecessary risk or led to dissatisfaction. Communications codes,
carriers and cultural contexts were all essential aspects. Communications are an important
component of adventure tourism products, and deserve further research attention.
Adventure tourism provides opportunities to test communications theories under adverse
circumstances. Keywords: workplace, safety, risk, discourse, language. Ó 2009 Elsevier
Ltd. All rights reserved.

INTRODUCTION
Communications is a complex phenomenon, and communications
involving tourists have been examined from many different perspec-
tives. This contribution aims to extend our understanding by focusing
on the role of operational workplace health and safety communications
within retail adventure tourism products. Tourism is a commercial
industry as well as a component of individual human behaviour and hu-
man social frameworks. The perspective presented here is that commu-
nications are an essential core component of the products which this
industry sells and tourists buy, and that this aspect of tourism commu-
nications therefore deserves greater research attention than it has at-
tracted historically. Additionally, communications is an information
transfer process as well as a social phenomenon, and the aspect exam-
ined here involves the fundamental interplay of code and carrier, with
culture and context as mediators rather than the primary focus. Com-
munications during critical health and safety incidents in commercial
adventure tours provide opportunities for powerful tests of human com-
munications theories at a fundamental level, stripped of most nuances.

Ralf Buckley is Director of the International Centre for Ecotourism Research (Email
<[email protected]>), and Research Director of the Climate Response Program at
Griffith University in Australia. His books include Adventure Tourism Management, Ecotourism
Principles and Practices (2009), Climate Response (2007), Adventure Tourism (2006), Environmental
Impacts of Ecotourism (2004) and Case Studies in Ecotourism (2003).

315
316 R. Buckley / Annals of Tourism Research 37 (2010) 315–332

For commercial adventure tour operators, the outdoors is both leg-

ally and operationally an industrial workplace, where health and safety
issues and communications are as important as in any other sector. In-
deed, safety and security are important throughout the entire tourism
industry. Workplace communications have been studied extensively in
many industries and professions, but rather little in tourism. Safety is of
particular concern in activities involving risk of injury, and hygiene in
activities involving risk of illness. Both of these apply for many commer-
cial adventure tourism products, especially multi-day tours involving
inexperienced clients undertaking unfamiliar activities in remote re-
gions. Here, therefore, I examine the roles and significance of opera-
tional communications on health and safety issues for one particular
adventure tourism subsector, namely those products based on rafting,
kayaking and seakayaking. These activities are referred to collectively as
paddlesports. The general operational features of this particular sub-
sector have been outlined by Buckley (2006, 2009).
The data for this analysis are derived from 388 days of records and
experiences from 63 different paddlesports tour products. This is 40
times more than the most extensive previous analysis of communica-
tions in whitewater rafting, that of Arnould, Price, and Otnes (1999).
Only a very small subset of these data can be presented in any detail
in a single publication. The analysis is therefore carried out in two suc-
cessive steps. The first describes the overall patterns, using all the data
but at a coarse resolution. The second focuses on communications in a
set of 20 individual and critical health and safety incidents. These rep-
resent only a very small proportion of the total, around 0.3%. It is this
small component, however, which best illustrates the critical opera-
tional role of these communications in commercial tour products;
where the most reliable data are available; and which provides practical
tests of particular aspects of communications theory.
To establish the theoretical context for these analyses, relevant previ-
ous literature is first reviewed in three successively more detailed stages.
The broad context of communications studies is considered briefly in
order to illustrate the diversity of approaches applied. Publications on
communications in tourism are reviewed in more detail to illustrate that
relevant research, both in tourism communications per se and in related
disciplines, has concentrated in particular subfields. Finally, the litera-
ture on adventure tourism specifically is examined to determine how,
and to what degree, it includes considerations of communications.

COMMUNICATIONS THEORIES AND TOURISM

Communications Theories
Communications theory is a complex field, and there is no one ac-
cepted approach which we can apply in tourism (Slack, 2006). The
most basic functional approach to individual communications (Las-
swell, 1948) is to consider them as transfers of information, either
deliberate or involuntary. Information may originate from single or
multiple sources and be transferred to single or multiple recipients,
 R. Buckley / Annals of Tourism Research 37 (2010) 315–332 317

which may be humans, other living creatures, or automated devices.

Information received is not necessarily identical to that transmitted.
Two particular aspects of this process have received particular theoret-
ical attention. The first is the means of information transfer, both phys-
ical transmission (carrier); and information included (signal).
Information can only be received if carrier and signal are within the
recipient’s detection capabilities; and in addition, if the recipient has
a consistent mechanism to discriminate between different signals in or-
der to respond to them in different ways (code). The second aspect is
the social context for codes. To transfer information, source and reci-
pient must use the same code, which may be very complex; and the re-
cipient must possess the code before the signal is transmitted. For
human communications, sharing of code depends on shared context
and culture. In conversation, information received may depend on
words, grammar, accent, intonation, volume, tone, timbre, speed,
and so on; on simultaneous visual signals such as expressions, posture,
movements and gestures, relative positions, clothing, makeup, hairstyle
and accessories; and on simultaneous tactile or olfactory signals. The
codes which the recipients use to obtain information from these multi-
ple simultaneous signals may also include contextual components such
as location, others present, prior relationships, and history of past
events including previous communications.
Different communication theorists have focused on different compo-
nents of these overall systems. In the social sciences, there has been a
strong focus on aspects of social and cultural context, such as language,
gender, age, emotions, and social hierarchies. If source and recipient
have highly congruent contexts and codes, a large amount of informa-
tion can be transmitted through a very simple signal: the tremor of an
eyelid across a crowded room, for example. Where two individuals with
very little shared context and code are attempting to communicate,
however, the opposite occurs. When two researchers from different
countries or academic disciplines, who speak little of the same lan-
guage or jargon, are trying to discuss a difficult concept, they must sim-
plify each communication as far as possible to improve the chance that
it will be received, and build up a large sequence of reciprocal commu-
nications so as to gradually develop a more complex shared code which
allows them to communicate successively more complex concepts.
Several recent texts illustrate the enormous range of different social
perspectives currently applied in communications theory and research.
Fiske (1990) discussed models, meanings, signs, codes, symbols, semi-
otics, structuralism, empiricism and ideology. Each of the 27 contribu-
tors to Shepherd, John and Striphas (2006) argued respectively for
communication to be viewed in different ways. 1 Berger (2006), quot-

1
Relationality, ritual, transcendence, constructive, a practice, collective memory, vision,
embodiment, raced, social identity, ‘‘techné’’, dialogue, autoethnography, storytelling,
complex organising, structuring, political participation, deliberation, diffusion, social
influence, rational argument, counterpublic, dissemination, articulation, translation, com-
municability and failure.
318 R. Buckley / Annals of Tourism Research 37 (2010) 315–332

ing a wide variety of contemporary and historical authors and analysts,

suggested that communications may be viewed from at least 50 differ-
ent perspectives, including signs or structure, codes or culture, art or
metaphor, dialog or narrative, fashion or fairy tales, myths or music,
postmodernism or psychoanalysis.

Communications in Tourism Studies

Tourism studies and communications studies are both situated lar-
gely within the social sciences and share some of the same means of en-
quiry. The interactions between these two fields of research have been
examined at a broad level in well-known works such as those of MacCan-
nell (1976, 1999), Cohen and Cooper (1986), Bruner (1989, 2005),
Dann (1996) and Jaworski and Pritchard (2005). The main focus in
all of that work has been on the subtle and complex social, cultural
and sociolinguistic aspects for tourism-related communications, rather
than the more fundamental aspects of carrier, code and context. There
is a subtheme on tourism as a set of performances (Edensor, 2001; Noy,
2008), and another on the role of tourism experiences in influencing
tourists’ perceptions of their own identities (McCabe & Stokoe, 2004;
Noy, 2004). All of this research treats tourism more as a social phenom-
enon than a commercial industry, and examines links between tourism
and communications as part of the social fabrics in different parts of the
world.
Within the literature on tourism as a commercial industry, commu-
nications have been considered mostly within the literatures of market-
ing (Buckley & Araujo, 1997; Clarke, 1996; Dann, 1996; Pitt, Opoku,
Hultman, Abratt, & Spyropoulou, 2007; Wu, Wei, & Chen, 2008) and
interpretation (Ap & Wong, 2001; Ballantyne, 2007; Moscardo, 1999;
Pearce, 1984; Reisinger & Steiner, 2006; Ryan & Dewar, 1995; Salazar,
2005). The literature on service quality mentions communication un-
der ‘‘relational quality’’ (Tsang & Ap, 2007). Many farm tourism oper-
ators in New Zealand, for example, ‘‘saw the issue of communication as
central to the success of their operation’’ (Pearce, 1990, p. 348). There
are also smaller literatures on communications during tourism crises
(Henderson, 2003; Ritchie, 2004); amongst tourism planning stake-
holders (Beesley, 2005; Bruyere, Beh, & Lelengula, 2008; Clarke,
1996); and amongst tourists and their relatives (Murphy, 2001; Noy,
2004; Pearce & Foster, 2007; Riley, 1988; Swarbrooke, Beard, Leckie,
& Pomfret, 2003; White & White, 2007). Analyses of tourism research
such as those by Xiao and Smith (2006a,b) do not identify communi-
cations as a significant research topic; and according to Ribeiro
(2009), the literature of tourism semiotics is thin on either scientific
methods or managerial viewpoints. In particular, even though interper-
sonal workplace communications have been studied extensively in
other industries and professions (Goffman, 2005; Holmes & Stubbe,
2003), including issues such as age (McCann & Giles, 2007), gender
(Domagalski & Steelman, 2007) and emotion (Miller, 2007), these as-
pects have received rather little attention in tourism research.
 R. Buckley / Annals of Tourism Research 37 (2010) 315–332 319

In adventure tourism specifically, texts such as Swarbrooke et al.

(2003) and Pigram and Jenkins (2006) make little mention of commu-
nications, and nor do recent contributions such as Buckley (2007) or
Williams and Soutar (2009). Descriptions of adventure guiding such
as Beedie (2003), Borrie and Roggenbuck (2001), Fluker and Turner
(2000) and Pomfret (2005) mention direct communications only in
passing. Health and safety issues were mentioned for some of the pad-
dlesports case studies reported by Buckley (2006), and Cater (2006)
summarised a number of safety failures in the commercial raft tour sec-
tor, but neither focussed specifically on communications.
The role of communications in adventure tourism has been ad-
dressed from an ethnographic perspective, with particular reference
to the emotional aspects of interactions between guides and clients,
by a series of authors including Arnould and Price (1993), Arnould
et al. (1999), Jonas (1999), Holyfield (1999) and Sharpe (2005). Inter-
estingly, the majority of these studies have focused specifically on white-
water rafting. Based on 10 days’ observations of commercial rafting
tours on the Green and Yampa Rivers in the Colorado system, Arnould
and Price (1993) and Arnould et al. (1999, p. 38) proposed that these
experiences could be seen as a ‘‘magical consumption system’’ which
‘‘mirrors traditional magical systems in intent and effect’’ and involves
respect, ritual and rhetoric. A large part of their proposal relates to com-
munications, including ritual rules: for example, that the word ‘‘wind’’
should never be mentioned. These two rivers are mostly ‘‘float’’ trips
which have long stretches without significant rapids. As one of the
guides quoted by Arnould et al. (1999, p. 48) said of the principal rapid
on the Yampa: ‘‘it’s like 30 seconds . . . [compared to] five days in this
canyon’’. Even so, however, these authors noted that ‘‘customer com-
ments about the rapids echo themes of existential and physical chal-
lenge, excitement, difficulty and anticipation’’.
Similar themes appear in the work by Sharpe (2005) on emotional la-
bour by adventure tour guides. Emotional labour was defined as
‘‘organisationally sanctioned emotions in interactions with clients’’,
and the display of emotion, real or simulated, is a significant compo-
nent of communications between guides and clients. Sharpe (2005, p.
33) cited previous work by Goffman (2005, p. 237) referring to ‘‘fateful
action’’, defined as actions which are both ‘‘problematic and conse-
quential’’. Such actions are at the heart of many adventure experiences.
Jonas (1999) noted that the job of whitewater raft guides involves ‘‘mak-
ing and facing danger’’. Based on 10 days experience as a trainee guide
on the Chattooga River and 47 interviews, including an occasion when
she fell out in one of the major rapids, Holyfield (1999) described cer-
tain aspects of communications between commercial guides and clients,
and noted that guides are expected to display both bravery and calm.
Each of these studies indicated that both verbal and non-verbal com-
munications are important in shaping clients’ experiences, and that cli-
ents’ expectations include the types of behaviours adopted by guides.
None of them, however, addressed the role of direct communications
in health and safety. The key aim of this contribution, therefore, is to
test whether, and examine how, such health and safety communications
320 R. Buckley / Annals of Tourism Research 37 (2010) 315–332

may be critical to the successful operations of a suite of commercial

adventure tour products.

Methods
The methodological approach adopted was participant observation
(de Walt & de Walt, 2002; Spradley, 1980), focussing on critical inci-
dents (Flanagan, 1954; Moscardo, Taverner, & Woods, 2006). The 63
tours studied took place in 19 different countries and ranged from a
day to a month in duration, from tropical to High Arctic latitudes, from
beginner to expert clients, and from US$50 to over $5,000 in price.
Most participants were 20–50 years old, with an equal gender balance
and a variety of ethnic, socioeconomic, linguistic and cultural back-
grounds. Some of the products were seakayak tours, either self-sup-
ported or in a few cases with boat support. Others were river raft or
kayak tours, with the raft tours using either oar or paddle rafts (Buckley,
2006), and in most cases also accompanied by at least one safety kaya-
ker. Some of them were in cold, remote areas with high levels of real
risk; others in warm, accessible areas where real risks were lower.
A total of 388 days’ operations were observed. Bearing in mind that a
critical safety incident, such as a client being thrown from a raft and
trapped underwater, can occur in a single second, it is worth noting
that this comprises over 15 million seconds of observations. It com-
pares with a total of 10 days’ observations on two rivers by Arnould
et al. (1999), and 10 days on one river by Holyfield (1999). A set of
20 incidents from 11 countries are reported here, illustrating a range
of events when communications were critical to health and safety.
These events, summarised in Table 1, make up <0.3% of the total per-
iod observed, i.e. <45,000 seconds. For reputable adventure tour oper-
ators, critical incidents are relatively rare. During those incidents,
however, health and safety were at significant and in some cases imme-
diate risk. The results reported here examine the role of communica-
tions in such situations. Table 1 also shows generally applicable
parameters such as water temperature and remoteness, which influ-
ence the overall degree of risk associated with the tour products con-
cerned (Buckley, 2007). Warm-water rivers pose no immediate
threats from hypothermia. Immersion in cold-water rivers can produce
hypothermia within a few minutes. Very cold rivers are only a degree or
two above freezing, and produce immediate cold shock on immersion,
and a severe risk of hypothermia without very prompt rescue. Arctic
ocean waters, being salt, may be up to 4 degrees below zero. For the
river-based cases, all save the Brazilian case incorporated whitewater
rapids at Class IV or Class V on the five-class international scale of dif-
ficulty (Buckley, 2006). All the examples used were multi-day tours with
overnight accommodation in river-bank or ocean-shore campsites. Cli-
ent-to-guide ratios ranged from <1:1 up to 6:1 for whitewater river
tours, and up to 11:1 for ocean seakayak tours.
For the majority of the raft tours studied, the author took part as
safety kayaker. This is a particularly useful position from which to
 R. Buckley / Annals of Tourism Research 37 (2010) 315–332 321

Table 1. Critical Incidents

Case No. Watercraft Country River flow Water temperature How remote

1 paddle rafts + 1 kayak Australia low cold med

2 oar rafts + kayaks China high med high
3 oar rafts + kayaks Tibet high v. cold high
4 oar rafts + 1 kayak China high cold med
5 oar rafts + 3 kayaks Tibet high v. cold high
6 paddle rafts + 1 kayak Uganda high warm high
7 oar rafts + kayaks USA high cold low
8 paddle rafts + 2 kayaks Russia med cold low
9 seakayaks Canada n/a Arctic med
10 seakayaks Norway n/a Arctic high
11 seakayaks USA n/a cold low
12 seakayaks Canada n/a Arctic med
13 dugout canoe Brazil med warm high
14 oar raft + kayak Nepal high warm med
15 oar raft + kayak Tibet high v. cold high
16 kayaks NZ med low low
17 oar rafts USA high cold med
18 seakayak USA n/a cold med
19 paddle raft + kayaks Nepal high warm med
20 oar raft + kayaks Tibet high v. cold high

conduct research observations. Safety kayakers do not have the routine

operational responsibilities and authority of guides, but are nonethe-
less treated by the clients as members of the staff. Their key role is
to assist in river rescues, if rafts flip or clients or guides fall out. They
are thus expected to watch out continually for safety aspects and incip-
ient risks. They also provide a backup to the guides, e.g. in explaining
health and safety information to clients who did not fully take in the
instructions at the initial briefings. In other tours, the author was pres-
ent as a kayaking client, but recognised by guides and other clients as
relatively experienced, albeit not particularly skilled. In either capacity,
there were excellent opportunities to observe and take part both in
health and safety communications and in more general conversation.
The author was thus present in person at each of these critical inci-
dents, and can report relevant evidence from direct observation and
experience. For the particular incidents presented here, there are also
documented reports made either at the time or shortly thereafter,
using a variety of mechanisms. Several of these incidents were docu-
mented using more than one such approach. The mechanisms in-
clude: high-definition videography for subsequent television
broadcasting (at least 3 incidents); still photography and low-definition
video later distributed on DVD (at least 3); articles, most of them illus-
trated, in newspapers and magazines (at least 6); written descriptions
in journals and other materials produced by the tour operators and/
or participants (at least 3); general descriptions in academic books
322 R. Buckley / Annals of Tourism Research 37 (2010) 315–332

(at least 10); descriptions and/or illustrations on websites, including

both tour operator websites and social utility sites such as Facebook
Ò (at least 3); and participant notebooks, journals and private corre-
spondence (at least 10). Only 19 of the overall set of 63 tours observed
are represented in the set of 20 critical incidents reported here. The set
reported was selected because it includes a variety of issues, circum-
stances and countries, and because the incidents concerned were re-
corded in some way.

CRITICAL HEALTH AND SAFETY COMMUNICATIONS

General Communications Context
Guides and clients communicated with each other continually
throughout these tours. Only a very small proportion of these communi-
cations, <1% of the total time, relate to critical health and safety issues. In
most tours, the majority of conversation is social, and does not affect
operations except as it influences group cohesion. This type of commu-
nication corresponds to ‘‘small talk’’ (Coupland, 2000) or ‘‘idle talk’’ in
the Heideggerian sense. Against this social background, there is a much
smaller set of operational communications which are important for day-
to-day routine, but not critical in the sense of those presented here. Most
far-reaching in their effects are communications between guides, espe-
cially those involving operational decisions. Second are communications
from guides to clients, especially those involving instructions. Third are
those from clients to guides, principally involving enquiries. And finally
there are communications amongst clients, largely seeking confirmation
or clarification of communications with guides (Table 2).
The communications context for the critical health and safety inci-
dents described here may be summarised as follows. Most tours of this
type include an initial safety briefing, commonly combined with basic
paddling instructions for those clients unfamiliar with the craft con-
cerned (Holyfield, 1999). These initial briefings cover topics such as:
the proper use of helmets and lifejackets; when and why to wear shoes;
how to hold on to rafts and how to help maintain stability; safe swim-
ming positions to adopt if one falls out; and what to do during a rescue
either by a raft, by a safety kayak, or using a thrown rope. At particularly
risky points in the tour there may also be more specific safety briefings.
For example, the guides may halt the rafts at the side of a river in order
to inspect a rapid before running it, and to point out particular hazards
and give specific instructions in case a raft overturns or a client falls
out. Similarly, in a seakayak tour the guides may group up the kayaks
temporarily, in order to give specific instructions for manoeuvres such
as landing on a beach through surf. There is also a standard set of rou-
tine operational hand, paddle and whistle signals between guides, for
safe navigation in difficult sections.
Most multi-day raft, kayak and seakayak tours involve camping on
small river or ocean beaches, and in some areas these same campsites
are used repeatedly by many different tour groups. In addition, partic-
ipants in these commercial tour groups commonly share camp crockery,
 R. Buckley / Annals of Tourism Research 37 (2010) 315–332 323

Table 2. Major Subjects of Routine Communications

Subject Examples Relative frequency for particular subject

Guide to Guide to Client Client

guide client1 to guide1 to client

Related to functional operations of the tour

safety throw rope, avoid danger \\ \\\ \ \
logistics lift boat, carry gear \\\ \\ \ \
logistics, campsites, stops, \\\ \ \\ \\
planning day’s journey, cooking
equipment expected weather, \ \\ \\\ \\
clothing needs, tents, tarps
hygiene washing, fires, human waste \\ \\\ \ \\
hygiene don’t drop litter, butts \ \\\ \ \\
hygiene wash hands, filter water, \ \\\ \\ \\
toilet practice
observation wildlife, rapids \ \\\ \\ \\
relationships, existing or new \2 \ \ \\
on-tour
Not related to functional operations of the tour
social general chatter \ \ \ \\\
politics current events \ \ \ \\
interests outside the tour \ \\ \ \\\
relationships, status, opinions \ \ \\ \\
outside tour
employment type of work \ \\ \ \\
travel other than this tour \\\ \ \ \\\

1
Communications guide to client mostly instructions; client to guide mostly questions.
2
Refers to public discussion only; much more out of clients’ hearing.;

cutlery and cooking utensils. Sanitation and hygiene are thus key
operational issues, both to prevent participants falling ill, and to main-
tain the aesthetic appearance of campsites for customer satisfaction.
Paddlesports tour operators worldwide have therefore developed stan-
dard practices and protocols for: constructing and using camp toilets;
washing and sterilising hands and dishes; separating and storing differ-
ent categories of garbage; and keeping campsites clean (Buckley,
2002). None of these are complicated, but they are new to many cli-
ents, and they need to be followed quite precisely from the very begin-
ning of the tour. Guides must communicate relevant information in
ways which minimise embarrassment for clients but which are unam-
biguous. Clients often discuss these procedures with each other for fur-
ther clarification and explanation.

Communications During Critical Incidents

The circumstances surrounding each of the 20 critical incidents
listed in Table 1 are outlined below, and the key features of the
324 R. Buckley / Annals of Tourism Research 37 (2010) 315–332

Table 3. Characteristics of Communications

Case Importance Urgency Who Clarity Value Language Culture Mode H&S Client
No. Outcomes Satisfaction

1 very immediate G to C high high same same speech good high

high
2 high immediate G&C high high same same speech good high
3 medium medium C to C low high different different speech good high
and
signs
4 high high G &C medium high several several speech good high
5 high low G &C medium high several several speech good high
6 high immediate G to G very low low same similar none poor low
7 high high G to G high high same same speech good high
8 high medium G &C low mixed several several speech good high
or
signs
9 low low G to C very low low same same signal, n/a low
speech
10 high low G to C high high same same speech good high
11 low medium C to G medium medium same same speech good high
12 low medium C to C very low low same same speech poor low
13 high medium G to C high very same different speech good high
high and
signs
14 medium low G &C medium high same several speech good high
15 low low G &C medium low same same speech good high
16 high low G &C high high same same speech good high
17 medium low G to C medium medium same same speech good high
18 high low G to C medium medium same same speech good high
and
signs
19 low low G&C low low same several speech poor low
and
signs
20 low low G&C medium medium different different signs good high

communications in each incident are summarised in Table 3. Incident

1 occurred on a cold-water river where the rafts and kayaks had re-
grouped in a small pool above a runnable but potentially deadly rapid.
A kayaking client miscalculated water flow and was washed under a raft,
pinned against a rock, and had to bail out of the kayak, floating to the
surface immediately above the rapid. The guides performed a very
quick and technically competent rescue, in seconds, using thrown
safety lines. The lead guide said quietly: ‘‘I hope you realise what a neat
piece of bagging [i.e. rope throwing] that was.’’ This was a very effec-
tive communication, leading the client to pay greater attention to
safety on the rest of the tour, but without causing undue embarrass-
ment or alarm.
Incident 2 took place on a 12-day expeditionary first descent of a
high-volume river with several large rapids. At the most dangerous of
these, most of the kayakers portaged the rapid. One client kayaker
ran the rapid successfully and waited below for safety, whilst one of
the rafts attempted a run with one guide and one client. The raft
was thrown high in the air and capsized, and the occupants were
thrown clear and washed down the rapid. The guide washed through
in a more dangerous position, and the kayaker first pulled him into
 R. Buckley / Annals of Tourism Research 37 (2010) 315–332 325

calmer water, and turned to go back for the raft client. The guide, how-
ever, gasped: ‘‘No! Get me to the bank!’’ The kayaker shouted to the
raft client: ‘‘Are you okay?’’ The client replied calmly, so the kayaker
took the guide to the bank before returning. This provides an example
of critical communications in an unusual situation, where a client was
not only rescuing a guide, but rescuing a guide before a client.
Incident 3 occurred on a trip with Chinese, Japanese and English-
speaking participants, where communications had to be translated by
the few multi-lingual participants. Two clients fell out of a raft and were
rescued by a safety kayaker. The clients did not immediately under-
stand the kayaker’s spoken instructions, so the rescue was slow. The
procedure had been explained during an earlier safety briefing, but
as confirmed by a Chinese-language magazine article published subse-
quently, it had not been understood.
Incident 4 involved a technically complex rescue of a capsized raft
caught in an eddy against a cliff immediately below a large rapid, as
dusk approached. The position was such that other rafts could not
reach it, but only kayaks. Three clients, two guides and one kayaker
were involved. Some spoke only Chinese, some only English, and two
were fluent in both. Communications, including translations, were crit-
ical firstly to alert the safety kayaker of the capsize, secondly to free and
reassure a client who was trapped partially under the raft, and thirdly
for the technical mechanics of the rescue, which required a quadru-
ple-purchase rope-and-pulley lift from a rock crag, with coordinated ef-
forts by 5 of the 6 people. It was after dark when the rescue was
completed, by the light of a floodlight from across the river.
Incident 5 occurred on a two-week commercial expeditionary first
descent of a high-volume, cold-water river with some severe rapids
and terrain, in a cold and very remote area. After a series of difficult
rapids requiring time-consuming portages, the group halted above a
long and dangerous rapid with active rockfalls onto the only potential
portage route. An attempt to explore downstream on foot was thwarted
by sheer canyon walls. In fact, a later expedition found dangerous and
inescapable rapids downstream, which would have placed the group in
life-threatening circumstances if it had continued. After discussion, the
trip leader decided to abort the expedition and retreat to the nearest
roadhead overland, which took six days. The participants were disap-
pointed, but trusted the expedition leader’s judgment, and the group
as a whole was able to extract itself safely. Good communications be-
tween the leader and the rest of the group, in three languages, were
critical.
In Incident 6, the clients were inexperienced, and there were five
paddle rafts but only one safety kayaker. The group portaged the rafts
around the upper half of the most difficult rapid, and ran the lower
half. The rapid ended in a large wave above a wide circulating pool,
neither of them deadly but sufficient to capsize the rafts and spill
the clients into the pool. They were swept in circles, some dropped
their paddles, and several were gasping rather desperately for breath.
Despite this, the rafts followed each other down the rapid without
allowing time for previous rafts to be rescued. Every raft capsized,
326 R. Buckley / Annals of Tourism Research 37 (2010) 315–332

and the safety kayaker had to rescue 19 individual clients, some of

them severely distressed, as well as a number of paddles. The guides
were overconfident and incautious, and had no communication sys-
tem, so each guide made an independent unilateral decision to start
their run, with no overall safety strategy.
Incidents 7 and 8 involved risks associated with heat and cold. Inci-
dent 7 took place on a high-volume river with a warm climate but cold
water. A raft capsized, and a kayaker took one of the participants to the
river bank, but was unable to see whether the section downstream was
safe. He therefore waited for instructions from the trip leader, forget-
ting that the cold water itself posed a danger to the swimmer. The lea-
der shouted urgently to get the swimmer out of the water, and the
rescue was completed by a raft at the riverbank downstream. Incident
8 occurred on a raft trip in Siberia, where all guides and clients except
two were Russian, and few spoke any other language. At one campsite
the guides built a banya, a steam bath system consisting of a pile of
rocks heated for several hours by a large fire, and covered by a heavy
double-skin tent sealed down onto sand. The guides and clients
climbed naked into the tent and whipped themselves with wet birch
branches, diving into the river when the steam heat became intolera-
ble. The space between tent wall and the rocks was very narrow, and
the rocks were large and hot enough to inflict extremely severe burns
if anyone fell onto them, e.g. whilst squeezing past others to get out. To
avoid such injury it was critical for participants to communicate any
intention to move—either in Russian or failing that, in sign language.
Incidents 9 and 10 took place during seakayak tours in the High Arc-
tic. In Incident 9, two single guide kayaks and two double client kayaks
were crossing an inlet on a calm sunny day. Risks were low, especially
since participants were wearing drysuits. One of the client kayaks was
dragging behind, and the other paddled a little ahead and waited at
a small iceberg. Instead of simply sending the second guide to catch
up with them, the lead guide fired a marine flare to signal an urgent
halt. The group paddled together to the nearby shore, and the lead
guide then berated the clients, to the considerable surprise of the rest
of the group. As it happened, these clients were carrying the lead
guide’s rifle, since it would not fit in his own kayak. The lead guide said
angrily: ‘‘Give me my rifle, and you can paddle off on your own.’’ The
clients were confused and muttered amongst themselves. One said: ‘‘I
think he is saying . . . . . . . . .’’ but faded to silence. The lead guide said
threateningly: ‘‘Don’t go there with me!’’ The second guide raised his
eyebrows, and the clients were silent. This was very poor communica-
tion, highly inappropriate, and a source of considerable client dissatis-
faction with no gain in safety. Incident 10 provides a strong contrast. It
occurred on a High Arctic tour involving seakayaking from an expedi-
tion cruise ship. There were 10 inexperienced clients in five double
kayaks, and one guide and one experienced client in single kayaks.
The guide handed this last client a small waterproof GPS unit and
asked: ‘‘Can you enter a waypoint every time we stop or change course?
If that fog closes in I’ll need to radio a course to the ship so they can
come and pick us up.’’
 R. Buckley / Annals of Tourism Research 37 (2010) 315–332 327

Incidents 11, 12 and 13 involved encounters with aquatic wildlife. In

Incident 11, the guides lead a group of inexperienced seakayakers into
a bay with a pod of orcas or killer whales. Some of the whales, which are
significantly larger than the kayaks, swam very close and in a few cases
underneath the kayaks, occasionally emitting bubbles which caused the
kayaks to shake in the water. The clients were greatly impressed and
some were frightened. The guides were able to reassure them success-
fully and the incident was the highlight of the tour. In Incident 12, a
High Arctic seakayak tour group encountered narwhal, the tour’s key
attraction and advertising icon. One kayak with two clients paddled
quietly towards the narwhal to see them at close range. When a narwhal
started to surface next to the kayak, however, one client panicked and
suddenly began to paddle backwards. This scared the narwhal into a
crash dive, to the great disappointment of the other client. This could
have been avoided through better communications between the clients
beforehand. Incident 13 involved a significantly higher actual risk,
associated with piranha, caiman and giant predatory catfish. It took
place on a small tributary of the Amazon, accessed by river boat and
then in a dugout canoe. The area was uninhabited because it was above
a set of rapids, and to get above the rapids required wading for about
1 km, pushing a dugout canoe. The river supports large populations of
piranha and in most areas it would be very dangerous to enter the
water. The local guide was able to communicate with the clients, in
English, that it was safe to walk in shallow rapids though not in deep
still water. The group was thus able to navigate successfully to the wil-
derness stretch upstream of the rapids.
Incidents 14, 15 and 16 all involved discovery of corpses in the rivers
concerned. In incident 14, a group of inexperienced clients became
excited and concerned when they discovered a bloated human corpse
floating in the river after the annual monsoon. The guides reassured
them that the tour’s drinking water treatment was adequate to prevent
health risks. In Incident 15, a much more experienced group found a
fully-clothed human skeleton in the water, but discussed it only in
terms of cultural differences and photo opportunities. In Incident
16, a kayak group found a dead dog in the same pool where they
had been taking drinking water. This led to some discussion of health
risk, but in fact nobody fell ill. Incidents 17 and 18 involved the need to
communicate non-standard toilet procedures designed for the particu-
lar environments concerned. In one case the instructions were straight-
forward but unexpected, requiring interpretation and reinforcement.
In the other case the procedure recommended involved a certain de-
gree of technical complexity, causing some embarrassment. In each
case, however, the guides were able to explain the procedures and rea-
sons in a straightforward but light-hearted way, and the clients ac-
cepted both.
Incidents 19 and 20 involved communications amongst guides and
clients with regard to campsite litter, especially cigarette butts. One
case involved a large international group, most of them inexperienced
backpackers, on a routine multi-day raft tour in a developing country.
Even though the company had a no-litter policy, the guides gave no
328 R. Buckley / Annals of Tourism Research 37 (2010) 315–332

relevant instructions, and both guides and clients simply dropped butts
onto the ground at campsites. This created significant dissatisfaction
for non-smoking clients, who picked up the cigarette butts every morn-
ing. In the other case, a multilingual but experienced group on an
expeditionary raft tour in a developing nation, some clients initially dis-
carded butts, but the other clients discreetly persuaded them not to do
so, so client satisfaction was greatly improved.

DISCUSSION AND CONCLUSIONS

The data reported here have implications both for the analysis of tour-
ism and the analysis of communications. These implications are exam-
ined below in successively greater breadth. The closest previous
analogues are those of Arnould et al. (1999) and Holyfield (1999). The
former found that guides use a formalised set of behaviours, including
communications, to create the client experience. They did not address
the direct role of communications in the transfer of health and safety
information. The latter described pre-embarkation briefings on routinely
repeated single-day trips, and her own emotional response to an inadver-
tent swim. Sharpe (2005) found that guides who repeated the same tour
throughout an entire season became emotionally exhausted. The much
larger set of data used here indicates that repetitiveness may be a key com-
ponent in emotional labour. This deserves further research, since results
would be applicable throughout the service industries.
From the perspective of adventure tourism product construction
(Buckley, 2006, 2007) these results indicate that operational health
and safety communications are indeed a key component of commer-
cial retail tour products. Adventure tours are temporary assemblages
of strangers who must carry out coordinated actions, including unfa-
miliar procedures, under circumstances which may involve risks to life,
limb or liver. In such conditions, communication is as critical as equip-
ment or staff. These results could be expanded: by examining other
activities; by considering culture, ethnicity, age, gender, socioeconomic
background and ideology; or by taking a broader social situations ap-
proach (Argyle, Furnham, & Graham, 1981; Pearce, 1990).
From a workplace communications perspective, the data presented
here indicate five key issues (Table 3): clarity and comprehension, va-
lue and appreciation, language barriers, cultural differences, and com-
munication modes (West, 2007). Guides need to be sure that clients
have both (a) understood the technical aspects of what to do, and
(b) appreciated the importance of doing so. This is difficult if there
are differences in language, in cultural norms, or in experience and
attitude to risk (Cater, 2006). Non-verbal demonstrations, role model-
ling and signals are important, as well as the spoken word (Buckley &
Littlefair, 2007). Instructions from guides to clients involve asymme-
tries of power (Gibson, 2003), and may contain emotional elements
(Beesley, 2005; Shepherd, St. John, & Striphas, 2006; West, 2007).
If guiding is performance which includes simulated emotion, then
the clients as audience expect only positive emotions such as enthusiasm
 R. Buckley / Annals of Tourism Research 37 (2010) 315–332 329

and encouragement, ‘‘bravery and calm’’ (Jonas, 1999); not boredom,

anger or fear. If a guide displays anxiety or even anger in a situation
with fateful consequences, this is accepted since emotional tones
provide an additional code to communicate urgency and importance.
Negative emotions are inappropriate, however, if the consequences are
not fateful or the emotions displayed are disproportionate. Some cli-
ents aim to perform as competent amateurs, mimicking the guides.
Others perform as helpless. Guides and safety assistants must distin-
guish clients who are actually in danger but communicate that they
are not, from those who are actually not in danger but communicate
that they are. For guides, communications reinforce their self-percep-
tions as having earned a privileged position through experience and
learned skills. For clients, some experience discomfort, embarrassment
and fear, creating a self-perception that they dislike the outdoors. For
most, however, successful navigation of real risks and dangers increases
their perceptions of self-worth and capability.
The data presented here reflect several key issues at the most funda-
mental level of communications theory. Often there are difficulties in
transmitting a carrier from source to recipient. The roar of the rapids
may drown out human voices; and drops, bends and rocks may obscure
the line of sight. Establishing communication relays is therefore criti-
cal. Where carriers are weak, intermittent or unreliable, and the conse-
quences of misinterpretation are severe, communication codes must
be simple and unambiguous. Connections between code and context
are also essential. Emergency signals only convey information because
those involved are expecting them. They are not part of the everyday
language of the participants. If the participants do not all speak the
same language, there may be difficulties in communicating the code
which may be needed for later emergencies. These aspects illustrate
what Striphas (2006) described as ‘‘communication as translation’’.
In addition to aspects of carrier, code and context, communications
in adventure tourism illustrate more subtle themes of language and
culture (Cohen & Cooper, 1986; Dann, 1996). Because of language dif-
ferences, tour participants may not understand safety briefings; but be-
cause of cultural differences they may not say so. Where different
groups of participants in the same tour speak different languages,
‘‘idle talk’’ can become divisive rather than cohesive. Guides commu-
nicating with clients from other cultures and languages cannot avoid
embarrassment by using euphemisms or circumlocutions.
In conclusion, the data presented here demonstrate that even
though they make up only a very small proportion of the total time in-
volved, health and safety communications between and amongst guides
and clients in commercial adventure tours nonetheless form a critical
component of these tour products. Adventure tours also provide a nat-
ural laboratory in interpersonal communications, where small groups
of people are forced to communicate on unfamiliar topics, in unfamil-
iar circumstances, over an extended period, only with each other,
whether they like each other or not. This provides an analogue for a
number of work environments, and could be used to greater advantage
for experimental studies in human communications. Adventure tours
330 R. Buckley / Annals of Tourism Research 37 (2010) 315–332

and similar products are already used as corporate training and team-
building tools, to give colleagues a chance to practice communications
under unfamiliar and stressful situations (Priest, 1997). The study of
tourism could surely be enriched by greater application of long-estab-
lished disciplines such as communications.

Acknowledgements—I thank the Coordinating Editor, eight anonymous referees and a number of
my colleagues for critical comments and constructive suggestions on several previous drafts of
this analysis, especially in regard to the communications aspects. Shortcomings remain my own.

REFERENCES
Ap, J., & Wong, K. (2001). Case study on tour guiding: Professionalism, issues and
problems. Tourism Management, 22, 551–556.
Argyle, M., Furnham, A., & Graham, J. (1981). Social situations. Cambridge:
Cambridge University Press.
Arnould, E., & Price, L. (1993). River magic: Extraordinary experience and the
extended service encounter. Journal of Consumer Research, 20, 24–45.
Arnould, E., Price, L., & Otnes, C. (1999). Making magic consumption: A study of
white-water river rafting. Journal of Contemporary Ethnography, 28, 33–68.
Ballantyne, R. (2007). Conservation learning in wildlife tourism settings: Lessons
from research in zoos and aquariums. Environmental Education Research, 13,
367–383.
Beedie, P. (2003). Mountain guiding and adventure tourism: Reflections on the
choreography of the experience. Leisure Sciences, 22, 147–167.
Beesley, L. (2005). The management of emotion in collaborative tourism research
settings. Tourism Management, 26, 261–275.
Berger, A. (2006). 50 Ways to understand communication: A guided tour of key ideas and
theorists in communication, media and culture. Landham MD: Rowman and
Littlefield.
Borrie, W., & Roggenbuck, J. (2001). The dynamic, emergent, and multi-phasic
nature of onsite wilderness experiences. Journal of Leisure Research, 33, 202–228.
Bruner, E. (1989). Review: Of cannibals, tourists, and ethnographers. Cultural
Anthropology, 4, 438–445.
Bruner, E. (2005). Culture on tours: Ethnographies of travel. Chicago: Chicago
University Press.
Bruyere, B., Beh, A., & Lelengula, G. (2008). Differences in perceptions of
communication, tourism benefits and management issues in a protected area
of rural Kenya. Environmental Management, doi:10.1007/s00267-008-9190-7.
Buckley, R. (2002). Minimal-impact guidelines for mountain ecotours. Tourism
Recreation Research, 27, 35–40.
Buckley, R. (2006). Adventure tourism. Oxford: CAB International.
Buckley, R. (2007). Adventure tourism products: Price, duration, size, skill,
remoteness. Tourism Management, 28, 1428–1433.
Buckley, R. (2009). Whitewater tourism. In M. Cooper & B. Prideaux (Eds.), River
tourism (pp. 181–196). Oxford: CAB International.
Buckley, R., & Araujo, G. (1997). Green advertising by tourism operators on
Australia’s gold coast. Ambio, 26, 190–191.
Buckley, R., & Littlefair, C. (2007). Minimal-impact education can reduce actual
impacts of park visitors. Journal of Sustainable Tourism, 15, 324–325.
Cater, C. (2006). World adventure capital. In R. Buckley (Ed.), Adventure tourism
(pp. 429–442). Oxford: CABI International.
Clarke, J. (1996). Farm accommodation and the communication mix. Tourism
Management, 17, 611–616.
 R. Buckley / Annals of Tourism Research 37 (2010) 315–332 331

Cohen, E., & Cooper, R. (1986). Language and tourism. Annals of Tourism Research,
12, 5–29.
Coupland, J. (Ed.). (2000). Small talk. Harlow: Longman.
Dann, G. (Ed.). (1996). The language of tourism: A sociolinguistic perspective. Oxford:
CAB International.
de Walt, K., & de Walt, B. (2002). Participant observation: A guide for fieldworkers.
Lanham MD: Rowman AltaMira.
Domagalski, T., & Steelman, L. (2007). The impact of gender and organizational
status on workplace anger expression. Management Communication Quarterly, 20,
297–315.
Edensor, T. (2001). Performing tourism, staging tourism. Tourist Studies, 1, 59–81.
Fiske, J. (1990). Introduction to communication studies (2nd ed.). London: Routledge.
Flanagan, J. (1954). The critical incident technique. Psychological Bulletin, 51,
327–360.
Fluker, M., & Turner, L. (2000). Needs, motivations, and expectations of a
commercial whitewater rafting experience. Journal of Travel Research, 38, 380–390.
Gibson, D. (2003). Participation shifts: Order and differentiation in group
conversation. Social Forces, 81(4), 1335–1381.
Goffman, E. (2005). Interaction ritual: Essays on face-to-face behaviour. New Brunswick:
Transaction Publishers.
Henderson, J. (2003). Communicating in a crisis: Flight SQ 006. Tourism
Management, 24, 279–287.
Holmes, J., & Stubbe, M. (2003). Power and politeness in the workplace. Harlow:
Longman.
Holyfield, L. (1999). Manufacturing adventure: The buying and selling of
emotions. Journal of Contemporary Ethnography, 28, 3–32.
Jaworski, A., & Pritchard, A. (Eds.). (2005). Discourse, communication and tourism.
Clevedon: Channel View.
Jonas, L. (1999). Making and facing danger: Constructing strong character on the
river. Symbolic Interaction, 20, 343–363.
Lasswell, H. (1948). The structure and function of communication in society. In L.
Bryson (Ed.), The communication of ideas. New York: Harper.
McCabe, S., & Stokoe, E. (2004). Place and identity in tourists’ accounts. Annals of
Tourism Research, 31, 601–602.
McCann, R., & Giles, H. (2007). Age-differentiated communication in organiza-
tions: Perspectives from Thailand and the United States. Communication
Research Reports, 24(1), 1–12.
MacCannell, D. (1976). The tourist: A new theory of the leisure class. New York:
Schocken Books.
MacCannell, D. (1999). The tourist: A new theory of the leisure class (3rd ed.).
California: University of California Press.
Miller, K. (2007). Compassionate communication in the workplace: Exploring
processes of noticing, connecting, and responding. Journal of Applied Commu-
nication Research, 35, 223–245.
Moscardo, G. (1999). Making visitors mindful. Illinois: Sagamore Publishing.
Moscardo, G., Taverner, M., & Woods, B. (2006). When wildlife encounters go
wrong: Tourist safety issues associated with threatening wildlife. In Y.
Mansfield & A. Pizam (Eds.), Tourism, security and safety: From theory to practice
(pp. 209–227). Burlington: Elsevier.
Murphy, L. (2001). Exploring social interactions of backpackers. Annals of Tourism
Research, 28, 50–67.
Noy, C. (2004). This trip really changed me: Backpackers’ narratives of self-change.
Annals of Tourism Research, 31, 78–102.
Noy, C. (2008). Pages as stages: A performance approach to visitor books. Annals of
Tourism Research, 35, 509–528.
Pearce, P. (1984). Tourist-guide interaction. Annals of Tourism Research, 11,
129–146.
Pearce, P. (1990). Farm tourism in New Zealand: A social situation analysis. Annals
of Tourism Research, 17, 337–352.
Pearce, P., & Foster, F. (2007). A ‘‘University of travel’’: Backpacker learning.
Tourism Management, 28, 1285–1298.
332 R. Buckley / Annals of Tourism Research 37 (2010) 315–332

Pigram, J., & Jenkins, J. (2006). Outdoor recreation management (2nd ed.). London:
Routledge.
Pitt, L., Opoku, R., Hultman, M., Abratt, R., & Spyropoulou, S. (2007). What I say
about myself: Communication of brand personality by African countries.
Tourism Management, 28, 835–844.
Pomfret, G. (2005). Mountaineering adventure tourists: A conceptual framework
for research. Tourism Management, 27, 113–123.
Priest, S. (1997). Physical challenge and the development of trust through
corporate adventure training. Journal of Experiential Education, 21, 31–35.
Reisinger, Y., & Steiner, C. (2006). Reconceptualising interpretation: The role of
tour guides in authentic tourism. Current Issues in Tourism, 9, 481–498.
Ribeiro, N. (2009). Tourism representation and semiotics: Directors for future
research. Cogitur, Journal of Tourism Studies, 2, 7–14.
Riley, P. (1988). Road culture of international long-term budget travelers. Annals of
Tourism Research, 15, 313–328.
Ritchie, B. (2004). Chaos, crises and disasters: A strategic approach to crisis
management in the tourism industry. Tourism Management, 25, 669–683.
Ryan, C., & Dewar, K. (1995). Evaluating the communication process between
interpreter and visitor. Tourism Management, 16, 295–303.
Salazar, N. (2005). Tourism and globalization: ‘‘Local’’ tour guiding. Annals of
Tourism Research, 32, 628–646.
Sharpe, R. (2005). ‘‘Going above and beyond’’: The emotional labor of adventure
guides. Journal of Leisure Research, 37, 29–50.
Shepherd, G., St. John, J., & Striphas, T. (2006). Communication as - -: Perspectives on
theory. Thousand Oaks: Sage.
Slack, D. (2006). Communication as articulation. In G. Shepherd, J. St. John, & T.
Striphas (Eds.), Communication as: Perspectives on theory (pp. 223–231). Thou-
sand Oaks: Sage.
Spradley, J. (1980). Participant observation. New York: Holt, Rinehart and Winston.
Striphas, T. (2006). Communication as translation. In G. Shepherd, J. St. John, &
T. Striphas (Eds.), Communication as: Perspectives on theory (pp. 232–241).
Thousand Oaks: Sage.
Swarbrooke, J., Beard, C., Leckie, S., & Pomfret, G. (2003). Adventure tourism: The
new frontier. London: Butterworth-Heinemann.
Tsang, N., & Ap, J. (2007). Tourists’ perceptions of relational quality service
attributes: A cross-cultural study. Journal of Travel Research, 45, 355–363.
West, R. (2007). Introducing communication theory: Analysis and application. New York:
McGraw Hill.
White, N., & White, P. (2007). Home and away: Tourists in a connected world.
Annals of Tourism Research, 34, 88–104.
Williams, P., & Soutar, G. (2009). Value, satisfaction and behavioral intentions in
an adventure tourism context. Annals of Tourism Research, 36, 413–438.
Wu, S., Wei, P., & Chen, J. (2008). Influential factors and relational structure of
internet banner advertising in the tourism industry. Tourism Management, 29,
221–236.
Xiao, H., & Smith, S. (2006a). The maturation of tourism research: Evidence from
a content analysis. Tourism Analysis, 10, 335–348.
Xiao, H., & Smith, S. (2006b). The making of tourism research: Insights from a
social sciences journal. Annals of Tourism Research, 33, 490–507.

Submitted 27 March 2008. Resubmitted 21 December 2008. Resubmitted 29 April 2009.

Resubmitted 7 September 2009. Final Version 21 September 2009. Accepted 14 October 2009.
Refereed anonymously. Coordinating Editor: Chaim Noy

Available online at www.sciencedirect.com