Content uploaded by Amotz Zahavi
Author content
All content in this area was uploaded by Amotz Zahavi on Nov 21, 2017
Content may be subject to copyright.
Nordic Society Oikos
Altruism as a Handicap: The Limitations of Kin Selection and Reciprocity
Author(s): Amotz Zahavi
Reviewed work(s):
Source:
Journal of Avian Biology,
Vol. 26, No. 1 (Mar., 1995), pp. 1-3
Published by: Blackwell Publishing on behalf of Nordic Society Oikos
Stable URL: http://www.jstor.org/stable/3677205 .
Accessed: 10/01/2012 06:11
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
Blackwell Publishing and Nordic Society Oikos are collaborating with JSTOR to digitize, preserve and extend
access to Journal of Avian Biology.
http://www.jstor.org
Point-of-View
Altruism as a handicap
- the limitations
of kin selection and
reciprocity
Amotz
Zahavi,
Inst.
for Nature Conservation
Research,
Tel-Aviv
University,
Tel-Aviv
69978, Israel.
The phenomenon
of "helping
at the nest" has exposed
ornithologists
to the
problem
of the evolution of altruism,
i.e. why should a non-breeder invest in the fitness of
breeders rather than
trying
itself to breed. Until the 1960s
altruism was explained by many biologists through
mod-
els of "group
selection"
(GS), i.e. the investment of the
non-breeders
in the
breeding
of their
group
members was
justified by its contribution
to the group.
GS was later
rejected by most
evolutionary biologists (Maynard
Smith
1964, Lack 1966, Williams 1966), and another
model,
that of "kin-selection"
(KS) (Hamilton
1964) was sug-
gested
as an alternative solution
to the problem
of altru-
ism. In KS models
the investment
of the altruist is justi-
fied by the benefit to its relatives.
However,
the many
studies on group
living birds
which have resulted
from
the growing interest
in the dilemma
of altruism,
have
revealed that
the
helpers
are
in fact often
not related to the
breeders
(Ligon and Ligon 1990, Reyer 1990, Zahavi
1990). Trivers
(1971) suggested
an additional
model -
"reciprocal
altruism"
(RA)
- to interpret
altruistic
adapta-
tions
among
non-relatives.
This latter model
suggests
that
the investment of the altruist is compensated
by a gain
from a reciprocal
investment
by other
group
members.
However,
data from several
field studies have indicated
that
in many
cases the act of the non-related
altruist
was
not
reciprocated.
In these cases neither KS nor
RA could
explain helping
at the nest. It was evident,
however,
in
some of these studies (Ligon and Ligon 1990, Reyer
1990, Zahavi 1990) that many of the helpers
were en-
hancing
their own chances to breed. Helping by non-
related
birds, which is not reciprocated,
can therefore
only be interpreted
as a simple
selfish investment.
How-
ever,
these
findings
did
not
stimulate
further
research
into
the possibility
that "helpers"
in general gain direct
ad-
vantages
by their
apparently
altruistic
investment.
Weaknesses of previous models
In the following I note the inherent
instability
in the
models based on GS, KS or RA and propose what I
consider
to be a more
general
and
stable
model,
in which
helping
is considered
as a selfish
behaviour,
i.e. its initial
presentation
as an altruistic
adaptation
was misleading.
The model also has implications
for interpreting
other
altruistic
phenomena,
even among the eusocial insects,
but these will not be discussed
here.
Group
selection.
GS is not an illogical
model: it pays to
invest in a group,
if the benefit of having a successful
group
exceeds the investment of the individual members
in the general
welfare of the group.
However,
at present
GS is generally
not used
by most
evolutionary biologists,
because it has been shown to be vulnerable to social
parasitism
(Maynard
Smith
1964).
The reason is that,
in a
GS model,
a member
of the group
that does not invest
in
the group's
welfare
(a social parasite),
gains
as much
as
the individuals
that do invest in the group,
without
in-
curring
the cost of helping.
Kin selection.
According
to KS theory,
altruism is based
on a model
of individual
selection
in which the gene for
altruism is the selected unit (Dawkins
1989).
The theory
claims that the frequency
of the gene for altruism
in-
creases in the population
as a result of the altruistic
behaviour,
even though
it decreases the reproduction
of
the altruist
itself. Is this
really
so? The best
way to expose
the fallacy of this claim is to tell a variant
of a story
attributed
to J.B.S.
Haldane,
who suggested
(1955)
that
if
one of two brothers
walking
beside a river,
were to fall
into
it and be in danger
of drowning,
it would
be reason-
able
for the other brother
to risk his life somewhat to save
the drowning
brother,
since by taking such a risk (i.e.
decreasing
his fitness), he may save his brother and
increase
the frequency
of genes similar
to his own in the
following generation.
The instability
of the model is clearly apparent
if the
same
story
is told
with three or more
brothers,
rather
than
two, walking along
the river. It is obvious
that
if one of
them
jumps
to the
rescue,
the other
sibling
(who does not
risk
himself),
gains
as much
as the
one who
risks
himself,
but
without
incurring any
cost. Thus,
in KS models,
as in
GS models,
the total
gain
of the selfish
brother
(the
social
parasite),
is higher than that of the altruist.
Eshel and
Motro
(1988), trying
to untangle
the problem
of multiple
1 JOURNAL OF AVIAN BIOLOGY 26:1 (1995) 1
potential rescuers, suggested an unrealistic
model, in
which there is no interaction or communication
among
the potential
rescuers,
and
consequently,
no way for the
brothers to predict
which of them, if any, is likely to
rescue the drowning
sibling.
Models of KS are in fact models of GS among
kin.
They are equally
unstable.
Although
the investment and
gain
in GS models are not
presented
by gene
frequencies,
any
investment is ultimately
turned into
gains
or losses in
fitness.
Thus,
because
of the
potential
advantage
to social
parasites
in models of KS and GS, both are equally
unstable over evolutionary
time.
The similarity
between GS and KS has recently
been
discussed
by Wilson and Sober
(1994). They suggested
that this similarity
justifies the use of GS, while I con-
sider both
GS and
KS to be equally
inadequate
to explain
social evolution. It is certainly
not
reasonable
to reject
GS
and accept
KS.
Reciprocal
altruism. It is easy to show that reciprocal
altruism
(RA) is also inherently
unstable.
Trivers
(1971)
suggested
that
RA could
be stable
in cases in which there
are mechanisms that
ensure
reciprocation.
He suggested
that among higher
animals,
such as man and monkeys,
the social parasite,
i.e. the individual that
does not reci-
procate,
will be punished.
However,
like any other
char-
acter,
such a mechanism is costly to the individual
that
possesses
it (Zahavi 1981).
Sigmund
(1993)
discussed
the
difficulty
in enforcing
reciprocation
and
regarded
it as a
major
problem
for models of RA. Thus,
as in GS and KS
models,
selfish
individuals that
do not
invest
in punishing
a social parasite,
gain more than those that invest in
maintaining
reciprocation
within
their
group.
An alternative model
Our studies of the Arabian
Babbler
Turdoides
squam-
iceps, a group-breeding songbird,
suggest a completely
different interpretation
for the motivation of birds to
invest in helping
their
group:
an interpretation
that
may
provide
a general
solution to the problem
of altruism.
It
suggests
that the investment
in the welfare of the group,
or of its members,
functions to advertise the quality
and
motivation
of the helper.
The advertiser
gains from its
investment
by increasing
its "social prestige".
Helping
may thus be considered as a simple selfish character.
Earlier
(Zahavi
1976, 1990) I used the term
"social
sta-
tus"
instead of "social
prestige",
however,
"social
status"
is often
used as a synonym
for "social
rank".
To empha-
size the difference
- the rank order of babblers
does not
change as long as they remain
members
of the same
group,
whereas their
social
prestige
may
change
(without
a change
in rank),
i.e. other
individuals
may
compromise
with them
to a greater
or lesser
extent,
as a consequence
of their
performance.
Social
prestige
functions like
a
peacock's
tail
or the
song
of a songbird.
It attracts
collaborators
and deters
rivals.
Prestige may be gained by investing
in wasteful
charac-
ters (Zahavi 1977a, 1987) as well as by investing in
"altruistic"
activities. The investment
involved in the
altruistic activities
serves
as an honest
signal
of the
ability
of the babbler
to help its group and its quality as a
collaborator;
it also reflects
its quality
as a rivar
in intra-
group
conflicts.
It is interesting
to note
that,
contrary
to what
might
be
ex.ected from all three
models
in which the advantage
to
the
helper
is indirect
(GS,
KS & RA), babblers
are
highly
motivated
to invest
in their
group,
and the population
is
not infected
by social
parasites.
On the contrary,
individ-
uals compete
with each other
to invest
in the interests
of
the group,
often interfering
with the helping
of others.
Dominants
interfere
with the "altruistic"
activities of sub-
ordinates
in feeding the nestlings
(Carlisle
and Zahavi
1986),
in allofeeding
between
adults,
in sentinel
activities
(Zahavi,
unpubl.obs.),
in mobbing
(Carmeli
1988,
Anava
1992), and in the defence
of the common
territory
(Za-
havi, unpubl.
obs.). Dominants
often prevent
subordi-
nates from helping the group. They are most likely to
interfere
with older and experienced
individuals,
which
are one rank
lower than
themselves,
and are more
tolerant
towards
the investment
of young
and
inexperienced
birds
of lower ranks.
Competition
and interference
among
indi-
viduals acting as helpers
have been observed
in other
studies
of cooperatively
breeding
birds
(Reyer
1990),
but
these
phenomena
were
not
considered
by the observers
as
a challenge
to the theories
of indirect
selection.
These
kinds
of interference
cannot be explained by any
model of indirect
selection,
according
to which
individu-
als gain more when other
individuals,
rather than
them-
selves, invest in the altruistic
acts. However, if, as I
suggest,
helping
confers
a direct
advantage
on the
helper,
the interference
and
competition
for opportunities
to help
can easily
be explained by individual
selection,
and
there
is no need for alternative
models to explain
the phenom-
enon of helping.
Investment
in "altruistic"
activities
stabilises the
social
collaborations
by advertising
the motivation
of the altru-
ists to cooperate,
hence it may
erroneously
be considered
as a mechanism that
has been
selected
to ensure
reciproc-
ation. Reciprocation,
however, cannot explain the in-
terference
with the altruistic
acts of others,
or the
fact that
altruism
can also function
as threat.
The
evolution
of cooperation
in groups
of two, such
as
sexual
mates,
is no easier
to explain
than
that
of larger
cooperations.
Although
some sexual
partners
do abandon
their mates, or do not invest much in their offspring,
frequently
mates invest more than is required
of them.
Sexual
mates,
like individuals in a group
of babblers,
tend
to compete for the investment in the cooperation:
in-
cubating
birds
are
often
reluctant
to be replaced
by their
incoming
mate,
and
frequently
the incoming
bird has to
push
its mate
aside
in order to replace
it (Zahavi,
unpubl.
obs.). I suggest
that
even in collaborations
of two, a large
part
of the investment
can be explained
as an advertise-
ment of the
quality
of the
investor
and of its motivation to
2 JOURNAL OF AVIAN BIOLOGY 26:1 (1995)
continue collaborating, in order to decrease the partner's
tendency to cheat or desert.
It has been suggested that the solution to the "prison-
er's dilemma" could explain the problem of the evolution
of cooperation in general (Axelrod 1984). Hence it has
attracted
much attention among biologists. I suggest that
neither the dilemma, nor its solutions, has much to do
with the real biological world. In the real world, individu-
als assess the qualities and the motivations of their poten-
tial partners and invest in advertising their own qualities
and motivations - before they enter into any collab-
oration. They continue to test each other as long as they
continue to cooperate (Zahavi 1977b). In the prisoner's
dilemma the collaboration is neither preceded by commu-
nication, nor are the prisoners able to talk to one another
during the game. Hence, clever as the "dilemma" and its
solutions are, they have very little, if any, relation with
the world of biology and social behaviour.
It has often been claimed that the observations that
individuals of many species tend to cooperate with their
kin, support
the theory of KS. However, this is not neces-
sarily so. When a new group is formed, members invest
time, energy and risk in advertising their qualities and
assessing the qualities and motivations of their potential
collaborators before they decide with whom to collab-
orate, and how much to invest in the collaboration with a
particular partner. Collaboration with kin may save a
large amount of this investment, because in a group of kin
the individuals know each other, and their rank order and
social prestige have been established long before the
establishment of the new collaboration. In babblers, new
groups composed of non-related individuals, take much
longer to settle down to breed than groups composed of
kin (Zahavi, unpubl. obs.). The fact that it is easier to
collaborate with kin may have an effect on the gene
frequency of the following generation. I term this 'kin
effect'. However, if this effect would cause individuals to
compromise their own direct fitness, in order to increase
their "inclusive fitness", the path would be open for the
success of social parasites and for the disintegration of
the social system.
To conclude, I suggest that all three models, GS, KS
and RA, sophisticated as they seem to be, are not useful
for interpretating
social interactions, because natural se-
lection has evolved a much more stable solution to the
problem: individuals invest in their collaboration in order
to increase their social prestige. They do so by advertis-
ing their qualities and their motivation to collaborate. The
benefit to the group is a consequence of, rather than the
factor that selects for the investment. It is now generally
accepted that a handicap involving waste, such as the
peacock's tail, is a logical and stable strategy to advertise
quality. Altruistic handicaps may operate equally well.
Thus, the problem of altruism is solved when the in-
vestment in the altruistic phenomenon is considered as a
handicap.
Acknowledgements
- The author wishes to acknowledge
the
Program
for Alternative
Thinking
for its financial support.
Thanks are also due to Yoram Yom-Tov and Arnon Lotem
for
their constructive
suggestions,
and to Avishag
Zahavi for her
major
contribution
in discussing
the arguments
and their
pre-
sentation.
I thank Nomi Paz for her editorial
assistance.
References
Anava,
A. 1992.
The value of mobbing
behaviour for the indi-
vidual babbler (Turdoides squamiceps). - MSc thesis, Ben-
Gurion
University
of the Negev.
Axelrod,
R. 1984. The evolution of cooperation.
Basic Books,
New York
.
Carlisle,
T. R. and
Zahavi,
A. 1986. Helping
at the nest, allo-
feeding and social status in immature Arabian
babblers. -
Behav. Ecol. Sociobiol. 18: 339-351.
Carmeli,
Z. 1988. Mobbing
behaviour in the Arabian
Babbler
(Turdoides squamiceps). - MSc thesis, The Hebrew Uni-
versity,
Jerusalem.
Dawkins,
R. 1989. The selfish gene. - Oxford Univ. Press,
Oxford.
Eshel,
I and
Motro,
U. 1988. The three brothers'
problem
- kin
selection with more than one potential
helper:
I. the case of
immediate
help. - Am. Nat. 132: 550-556.
Haldane,
J. B. S. 1955.
Population genetics.
- New Biology 18:
34-51.
Hamilton,
W. D. 1964.
The
genetical theory
of social behaviour.
- J. theor. Biol 7: 1-52.
Lack,
D. 1966. Population
studies of birds.
- Clarendon
Press,
Oxford.
Ligon,
J. D. and
Ligon,
S. H. 1990. Green
Woodhoopoes:
Life
history
traits and
sociality.
- In:
Stacey,
P. B. and
Koenig,
W.
D. (eds). Cooperative
breeding
in birds.
Cambridge
Univ.
Press,
Cambridge,
U.K., pp. 33-65.
Maynard-Smith,
J. 1964. Group
selection and kin selection.
-
Nature 201: 1145-1147.
Reyer, H.-U. 1990. Pied Kingfishers:
ecological causes and
reproductive
consequences
of cooperative
breeding.
- In:
Stacey,
P.B. and
Koenig,
W. D. (eds).
Cooperative breeding
in birds.
Cambridge
Univ.
Press,
Cambridge,
U.K.,
pp. 527-
557.
Sigmund,
K. 1993. Games of Life. Explorations
in Ecology,
Evolution
and Behaviour.
- Oxford Univ. Press,
Oxford.
Williams,
G. C. 1966. Adaptation
and natural selection: A cri-
tique of some current
evolutionary
thought.
- Princeton
Univ. Press,
Princeton,
New Jersey.
Willson,
D. S. and Sober,
E. 1994.
Reintroducing
group
selec-
tion to the human behavioral
sciences.
- Behav. Brain Sci.
17:585-654.
Trivers,
R. 1971. The evolution of reciprocal
altruism.
- Quart.
Rev. Biol. 46: 35-57.
Zahavi,
A. 1976.
Cooperative
nesting
in Eurasian
birds.
- Proc.
16th
Intern. Ornithol.
Congr.,
Canberra:
685-693.
- 1977a. The testing
of a bond.
- Anim. Behav.
25: 246-247.
- 1977b.
Reliability
in communication
systems
and the
evolu-
tion of altruism.
- In: Stonehouse,
B. and Perrins,
C. M.
(eds). Evolutionary
Ecology. McMillan,
London,
pp. 253-
259.
- 1981. Natural
selection,
sexual
selection and the
selection
of
signals. - In: Scudder,
G. G. E. and Reval, J. L. (ed.).
Evolution
today.
Carnegie-Melon
Univ.,
Pittsburg,
Ohio,
pp.
133-138.
- 1987.
The
theory
of signal
selection and some of its implica-
tions. - In: Delfino, V. P. (ed.). Proc. Intern. Symp. Biol.
Evolution. Adriatica Editrica, Bari, pp. 305-327.
- 1990. Arabian babblers: the quest for social status in a
cooperative breeder. - In: Stacey, P. B. and Koenig, W. D.
(eds). Cooperative breeding in birds. Cambridge Univ.
Press, Cambridge, U.K., pp. 105-130.
1"* JOURNAL OF AVIAN BIOLOGY 26:1 (1995) 3