PNAS article

Mon Dec 14 00:29:23 UTC 2009

I can't attach this as a pdf given listserv rules; hopefully you can make sense of the raw text

Mark

Campbell's monkeys concatenate vocalizations
into context-specific call sequences
Karim Ouattaraa,b,c, Alban Lemassona,1, and Klaus Zuberbu¨ hlerd,1
aLaboratoire EthoS ''Ethologie Animale et Humaine,'' Unite´ Mixte de Recherche 6552, Centre National de la Recherche Scientifique, Universite´ de Rennes 1,
Station Biologique, 35380 Paimpont, France; bCentre Suisse de Recherches Scientifiques, Taï Monkey Project, 01 BP1303, Abidjan 01, Co^ te d'Ivoire;
cLaboratoire de Zoologie et de Biologie Animale, Universite´ de Cocody, 10 BP770, Abidjan 10, Co^ te d'Ivoire; and dSchool of Psychology, University of St.
Andrews, KY16 9JP Saint Andrews, Scotland
Edited by Charles G. Gross, Princeton University, Princeton, NJ, and approved October 26, 2009 (received for review July 20, 2009)
Primate vocal behavior is often considered irrelevant in modeling
human language evolution, mainly because of the caller's limited
vocal control and apparent lack of intentional signaling. Here, we
present the results of a long-term study on Campbell's monkeys,
which has revealed an unrivaled degree of vocal complexity. Adult
males produced six different loud call types, which they combined
into various sequences in highly context-specific ways. We found
stereotyped sequences that were strongly associated with cohesion
and travel, falling trees, neighboring groups, nonpredatory
animals, unspecific predatory threat, and specific predator classes.
Within the responses to predators, we found that crowned eagles
triggered four and leopards three different sequences, depending
on how the caller learned about their presence. Callers followed a
number of principles when concatenating sequences, such as
nonrandom transition probabilities of call types, addition of specific
calls into an existing sequence to form a different one, or
recombination of two sequences to form a third one. We conclude
that these primates have overcome some of the constraints of
limited vocal control by combinatorial organization. As the different
sequences were so tightly linked to specific external events, the
Campbell's monkey call system may be the most complex example
of 'proto-syntax' in animal communication known to date.
alarm call  nonhuman primate  referential communication 
semantic  syntax
One way of studying language evolution is to compare the
communicative abilities of humans and animals. Parallels
with human language can be found at various levels, both in
terms of production and comprehension. Particularly relevant
are cases of social influences on vocal development (1), cases of
infant babbling (2), and other types of vocal learning (3-5). In
some species, there is evidence for population-wide convergence
effects in the form of culturally transmitted dialects [e.g., starlings
(6), whales (7), and Japanese macaques (8)]. In terms of
pragmatic use, there is good evidence that call production can be
influenced by specific audiences (9, 10). In terms of comprehension,
primates and possibly many other species are able to
assign meaning to different calls if there is a strong relation
between a call's acoustic morphology and its eliciting context
(11-14). In some species, there is some evidence for hemispheric
specialization when processing conspecific calls [e.g., horses (15),
Campbell's monkeys (16), rhesus macaques (17), starlings (18),
and sea lions (19)].
Despite all this evidence, there is currently a wide consensus
that human language differs from animal communication in a
profound way, because the essence of human language is its
complex grammatical organization, something that is widely
lacking in all animal communication systems (20-22). Despite
considerable effort devoted to the topic, the evolutionary origins
of this key property of language have remained relatively elusive
(23). This is not to say that animal communication does not
follow certain combinatorial principles. Gibbons, whales, and
songbirds, for example, combine finite and stereotyped sound
elements to form more complicated structures (1, 24-27). In
some cases, these structures possess hierarchical organization,
although very little is known about the relationship between
acoustic structure and communicative function. A typical finding
is that if the structure of a sequence is artificially altered, for
example by changing the composition or order of elements, the
signal tends to loose its communicative function (28-30). Another
relevant point is that nonhuman primates are perfectly
capable of discriminating human speech composed in different
ways [e.g., tamarins (31)] and of comprehending simplified
nonverbal forms of human syntax [e.g., apes (32-35)].
In natural contexts, spontaneous call combinations have also
been observed in nonhuman primates, although there are only a
small number of examples. Chimpanzees combine some of their
calls in nonrandom ways, although the communicative function
of these combinations remains to be investigated (36). Bonobos
produce five acoustically distinct call types in response to
different foods, with a predictable relationship between the
caller's food preference and the relative frequency of the different
calls (37). In putty-nosed monkeys, adult males produce
two loud calls, ''pyows'' and ''hacks,'' in a range of contexts,
including predation. However, when combining the two calls in
one specific way (i.e., a few pyows followed by a few hacks), males
give a supplementary message to their group members to move
away from the current location (38, 39). In chickadees, songs
contain C and D notes, which encode different information,
while the number and acoustic variation of D notes provides
supporting information about the size and dangerousness of a
predator (40, 41). During social interactions, animals hear call
combinations even more often, especially when individuals
exchange vocalizations during social interactions. Playback experiments
have shown that baboons recognize individual callers
and respond to the combinations of calls produced by them,
rather than the individual calls, and extract meaning from them
(42, 43).
The vocal behavior of female Campbell's monkeys is relatively
well studied, particularly in captivity, and has revealed considerable
socially determined acoustic plasticity (5, 44-47). Adult
male Campbell's monkeys very rarely participate in the various
call exchanges that take place within the group, but instead
produce a range of loud and low-pitched calls that carry over
large distances, much beyond the immediate group distribution
(48, 49). These calls are usually given to serious disturbances and
in sequences, some of which have been described in earlier
research (49-51). For instance, Gautier and Gautier-Hion (48)
suggested that hack sequences functioned as predator signals
while ''boom-hack'' sequences functioned in spatial coordination.
Zuberbu¨hler (49) confirmed experimentally that artificially
Author contributions: K.O., A.L., and K.Z. designed research; K.O. performed research; K.O.,
A.L., and K.Z. analyzed data; and K.O., A.L., and K.Z. wrote the paper.
The authors declare no conflict of interest.
This article is a PNAS Direct Submission.
1To whom correspondence may be addressed. E-mail: kz3 at st-and.ac.uk or alban.
lemasson at univ-rennes1.fr.
www.pnas.orgcgidoi10.1073pnas.0908118106 PNAS Early Edition  1 of 6
PSYCHOLOGICAL AND
COGNITIVE SCIENCES
ANTHROPOLOGY
adding ''booms'' before a pure hack sequence ''cancelled'' the
predator message. Recently, we have found that the hacks were
an acoustically heterogeneous class of vocalizations, which could
be categorized into five structurally and contextually different
call types (''krak,'' ''hok,'' ''krak-oo,'' ''hok-oo,'' and ''wak-oo'')
(50) (Fig. 1). This subtle acoustic diversity is achieved by
frequency transitions and an optional suffixation principle,
which increases the males' call type repertoire considerably.
Here, we studied the alarm calling behavior of free-ranging
Campbell's monkey males in natural and experimental predator
situations. We were particularly interested in how males concatenated
their repertoire of six call types into call sequences. To
this end, we sought to describe the principles governing the
organization of sequences in terms of composition and call order
and how the different sequences related to external events.
Results
Call Sequences in Nonpredation Contexts. In nonpredation contexts,
we recorded three distinct call sequence types, (i) a pair of
''boom'' calls (B) given alone, (ii) a pair of boom calls followed
by a series of krak-oo (K), and (iii) a pair of boom calls, followed
by a series of K calls, with one to several hok-oo (H) calls
interspersed (Table 1 and Fig. 2).
The first sequence type, unaccompanied pairs of boom calls, was
not very common, although we managed to record this sequence
from seven males (n13 events).Whenproducing unaccompanied
pairs of boom calls, the common finding was that the male was not
in visual contact with the group, usually some 30-50 m away (12 of
13 events). Campbell's males prefer peripheral positions within the
group where they remain in visual contact with at least one group
member (average distance between the male and the nearest
female: 4.800.61 m; n10 per male; four habituated males). We
observed that, immediately after this boom call sequence, the group
often progressed toward the male (8 of 13 events), similar to
previous findings in putty-nosed monkeys (39). More detailed
observations were possible for the four habituated males. In 10 of
13 instances of unaccompanied pairs of booms, the group traveled
for unusually long distances of 100 m during the next 30 min at
a speed significantly higher than normal (1m/min; determined by
matched samples collected at the same time of day without male
calling; Mann-Whitney U test: N1  N2  10, Z  1.96, P  0.05).
The second sequence type consisted of a pair of boom calls
followed by a K sequence (median overall  10 calls; range:
4-15 calls: n  53; Fig. 2). This combined sequence, recorded
from all four habituated males, was mainly given to falling trees
or branches with no other noticeable disturbance (85%). In the
remaining cases (15%), the sequence was given in response to
fights between other monkey species in the canopy, although this
usually led to branches falling as well.
The third sequence also consisted of a pair of boom calls
followed by a K sequence (overall median  12 calls; range:
9-16 calls; n  76), but here the sequence was interspersed with
1-7 H calls (Fig. 2). This combined sequence was recorded
from all four habituated males and two semihabituated groups
and always in response to neighboring Campbell's monkey
groups or single stranger males, suggesting that it functioned in
territorial defense. To further test this hypothesis, we investigated
whether the location of call production varied in systematic
Table 1. Vocal utterances produced by the different males over the course of the study period
Males
Sequence
composition M1 M2 M3 M4 M5 M6 M7 M8
B 4 1 2 3 1 1 1 -
B-K 29 5 13 6 - - - -
B-K-H 18 29 22 2 2 3 - -
K 2 1 2 - 1 1 1 1
K-K 6 3 4 - 1 1 1 1
K 5 5 6 1 1 - - -
K-W 2 1 2 - - - - -
K-H-W 2 1 1 - 1 - - -
K-H-H -W 9 4 8 - 1 2 2 2
M1 to 4, habituated males observed systematically, M5 to 8, semihabituated males observed opportunistically.
Frequency (kHz)
0.2 0.4 0.6
1
2
3
4
0.2 0.4 0.6
1
2
3
4
Duration (s) Duration (s)
0.2 0.4 0.6
1
2
3
4
0.2 0.4 0.6
1
2
3
4
Hok (H)
Frequency (kHz)
0.2 0.4 0.6
1
2
3
4
Duration (s)
Hok-oo (H+)
Frequency (kHz)
0.2 0.4 0.6
1
2
3
4
Duration (s)
Boom (B) Krak (K)
Wak-oo (W+) Krak-oo (K+)
Frequency (kHz)
Frequency (kHz) Frequency (kHz)
Fig. 1. Spectrograms of the six call types produced by males. Three calls could be described in terms of an acoustically invariable affix following the call stem.
, indicates that call stem is trailed by a suffix, an acoustically invariable ''oo'' utterance.
2 of 6  www.pnas.orgcgidoi10.1073pnas.0908118106 Ouattara et al.
ways. We predicted that sequences given in the periphery of a
group's home range contained significantly more H calls than
sequences given in the center, which turned out to be the case
(Fig. 3; Fisher's exact test, P  0.001).
A fourth vocal pattern in nonpredation contexts was the
production of single isolated K calls (n  9; three habituated
males). This always happened when a male was startled by
sudden movements of a nonpredatory animal, such as a duiker,
flying squirrel, or a human observer.
Call Sequences in Predation Contexts. In predation contexts, which
is when monkeys interacted with leopards or crowned eagles, we
recorded a more diverse range of call sequence types (Table 1
and Fig. 2).
General Predator Alert. A first sequence, recorded from all four
habituated males and one semihabituated male, consisted of K
calls only (median  15 calls; range: 3-25; n  18; Fig. 2). This
sequence was rare and given to any auditory sign of a predator,
typically after hearing Diana monkey alarm calls (n  15/18) or,
more rarely (n  3/18), after hearing predator vocalizations, but
never to any visual signs of a predator.
Crowned Eagle Alarm. The majority of call sequences to crowned
eagles were composed of wak-oo (W) and krak-oo (K) calls,
sometimes with the addition of hok (H) and hok-oo (H) calls
(median  25 calls; range: 15 to 40; n  38; Fig. 2). By analyzing
in more details the sequences produced by the three most
habituated males, we found that high levels of urgency were
associated with a high proportion of H and H calls in the
sequence. There was a significant difference in the proportion of
H and H calls if the caller spotted a real eagle or encountered
the eagle model compared to when he only heard eagle vocalizations
or Diana monkey eagle alarm calls (Fisher's exact test,
P  0.001; Fig. 4A).
Leopard Alarms. Call sequences to leopards were always composed
of krak (K) calls, sometimes combined with krak-oo (K)
calls (median  21 calls; range: 12-35; n  26; Fig. 2). In the
sequences recorded from the three most habituated males, we
found that the level of urgency was associated with a high
proportion of K calls in the sequence. Sequences with just K calls
were given in response to real leopards and leopard models. K
calls were far more common when callers responded to leopard
vocalizations or Diana monkey leopard alarmcalls. Compared to
B K K+ H H+ W+
B
B
H+
K+
B K+
K
K K+
K+ W+
K+ H W+
K+ H H+ W+
K+
Non-predatory Leopard Crowned eagle
Cohesion
& Travel
Tree/Branch
Inter-group
Real
Model
Calls
Alarm
Sequence composition
Real
Model
Calls
Alarm
1
2
3
4
5
6
7
8
9
N total
13
--
--
--
--
--
--
--
--
--
53
--
--
--
--
--
--
--
--
--
76
--
--
--
--
--
--
--
--
--
3
--
--
--
--
--
--
--
--
6
4
--
--
--
--
--
--
--
--
8
2
--
--
--
--
--
--
--
5
11
--
--
--
--
--
--
--
--
--
--
--
11
--
--
--
--
--
--
--
--
10
--
--
--
--
--
1
3
3
3
--
--
--
--
--
4
2
2
4
13
53
76
9
17
18
5
5
28
Context
Fig. 2. Composition of call sequences in different behavioral contexts. ''Alarm'' indicates leopard or eagle alarm calls given by sympatric Diana monkeys.
**
*
*
** *
****
****
**
*
****
*
**
**
*
*
**
*
**
****
**
*
**
**
*
*
*
**
****
*
*
*
**
* *
***
*
*
**
*
***
*
*
A B
*
*
*
*
*
***
**
*
*
****
*
*
*
*
*
***
*
**
*
*
*
**
***
**
***
*
***
*
*
*
*
**
*
*
*
Fig. 3. Geographical distribution of call sequences given during intergroup encounters (A) and after falling of trees or large branches (B). The two study groups
were habituated to human observers and had adjacent territories. Each square represents a 100100marea.Wedivided each home range into two parts: center
(dark squares) and periphery (light squares). The two groups were surrounded by four other semihabituated groups. Stars indicate the location of calling
sequences (A, n  71; B, n  53).
Ouattara et al. PNAS Early Edition  3 of 6
PSYCHOLOGICAL AND
COGNITIVE SCIENCES
ANTHROPOLOGY
real leopards and leopard models, K calls were given significantly
less often to leopard vocalizations and Diana monkey leopard
alarm calls (Fisher's exact tests, P  0.001; Fig. 4B).
The Importance of Call Order. The different call sequences were not
randomly assembled but ordered in specific ways, with entire
sequences serving as units to build more complicated sequences.
As mentioned, pairs of booms alone instigate group movements
toward the calling male, while K series functioned as general
alarm calls. If combined, the resulting sequence carried an
entirely different meaning, by referring to falling wood. In all
cases, the booms preceded the K series. We also found that
another sequence, the H series, could be added to boom-K
sequences, something that callers did when detecting a neighboring
group. H series were never given by themselves.
Nonpredation events were characterized by the production of
two boom calls, which could be given alone (to indicate group
movement) or which could introduce subsequent calls (100%,
n  142 cases, all eight males). In response to tree falls, the
booms preceded a series of K calls. In response to neighbors,
the booms preceded a series of H calls, followed by a series of
K calls. In some cases, callers inserted a single K call between
the booms and H call series. Interestingly, these insertions were
only heard in group 1, both by the first male and also by his
successor (males 1 and 2), but never in group 2 (i.e., males 3 and
4).Although we only have data from four males, it is possible that
the particular way by which the K and H calls are concatenated
was an identifying specific feature of group 1. We never
heard this particular insertion in the semihabituated groups.
Another unexpected difference was that both immigrant males
produced significantly more H calls (males 2 and 4; median 
5; range: 1-7; n  36) than their predecessors (males 1 and 3;
median  2; range: 1-5; n  40; Mann-Whitney U tests, Z 
5.33, P  0.001).
To leopards, males sometimes produced pure K sequences
(n  9/26), but if K calls were added, then typically toward the
end of the sequence. We compared the number of K and K calls
in the first and second half of the sequences and found a
significant difference (Fisher's exact test, P0.001). To crowned
eagles, most sequences began with a series of H (16 of 38
sequences) and typically ended with a series of K (36 of 38
sequences). This specific order probably reflected the decrease
in urgency or threat perceived by the caller. If W or H were
produced, then they appeared more or less randomly throughout
the sequence without any detectable patterns, while H and K
followed the ordering outlined before. There was a significant
difference between the number of H and K at the beginning
and at the end of sequences (Fisher's exact test, P  0.001; n 
38), which had the effect that the distribution patterns differed
significantly between the four call types (2 test X2  311.3;
DL  21; P  0.001).
Discussion
In earlier work, we have shown that Campbell's monkey males
use an affixation rule to increase the size of their call repertoire.
Adding the suffix ''oo'' to krak (K3K) or hok (H3H) calls
altered these calls' meanings in predictable ways (50). Here, we
describe regularities at another level, i.e., in how monkeys
combined this repertoire of six call types into nine distinct call
sequences (Fig. 2). These call combinations were not random,
but the product of a number of principles, which governed how
semantic content was obtained.
We found five main principles that governed these relationships.
First, callers produced sequences composed of calls that
already carried narrow meanings (e.g., K  leopard; H 
crowned eagle) (49, 50, 52). In these instances, sequence and call
meanings were identical. Second, callers produced meaningful
sequences, but used calls with unspecific meanings (e.g., K 
predator). Third, callers combined two meaningful sequences
into a more complex one with a different meaning (e.g., B  K
 falling wood). Fourth, callers added meaningless calls to an
already meaningful sequence and, in doing so, changed its
meaning (e.g., B  K  H  neighbors). Fifth, callers added
meaningful calls to an already meaningful sequence and, in doing
so, refined its meaning (e.g., K  K  leopard; W  K 
crowned eagle). We also found regularities in terms of call order.
Boom calls, indicative of a nonpredation context, always preceded
any other call types. H and K calls, indicators of crowned
eagles or leopards, were always produced early in the sequence
and were relatively more numerous if the level of threat was high.
In forest guenons, the single adult male of the group mainly
vocalizes in response to predators and other significant external
disturbances (5, 48, 53). Females are vocally active in social
situations (44-46) and to predators to which they produce a
diverse alarm call repertoire that encodes information about
predator type and level of threat (47). Whether Campbell's
monkeys produced these calls to intentionally inform others
about the event they have experienced cannot be decided with
our data. Some observations suggest not. For example, it is
puzzling that males produce loud calls in response to the
thundering sound of falling trees, as all other group members will
have perceived the event as well. The collapse of a large tree
provides a significant danger to arboreal animals and fatalities
are not unusual. Thus, calling males may have the urge to
advertise their uninjured state and provide an acoustic focal
point for scattered or disoriented group members, rather than
attempting to inform others about the danger.
W+ / K+
H / H+
K
K+
*** ***
A Eagle B Leopard
0
20
40
60
80
100
1 2 3 4
Mean N calls per sequence (%)
Mean N calls per sequence (%)
Fig. 4. Relative distribution of different call types within predator sequences with varying levels of predator threat. (A) eagle, (B) leopard. Fisher exact test (***,
P  0.001) were used to compare the relative contribution of obligatory (black) vs. optional (white) call types in low threat situations (1, mainly Diana monkey
alarms; 2, playback of predator vocalizations) or high threat circumstances (3, visual predator models; 4, real predator encounters).
4 of 6  www.pnas.orgcgidoi10.1073pnas.0908118106 Ouattara et al.
The core finding of this study is that the different calls
produced by the males were usually given as part of structurally
unique sequences and in context-specific ways. Context could be
described in terms of event type, degree of threat, spatial
relations within the group, and group movements. Some of these
functions have already been shown for other guenons, particularly
Diana monkeys (54) and putty-nosed monkeys (38, 39), but
no previous study has revealed such a rich portfolio of contextspecific
calling behavior. In earlier playback experiments, series
of H and K calls were meaningful to Diana monkeys, as these
calls elicited their own corresponding eagle or leopard alarm
calls (49). If artificially combined with pairs of B, however, Diana
monkeys no longer responded to the H or K sequences, suggesting
that this was due to the boom's indication of a nonpredatory
event. Although we have not found any natural B-H or B-K
sequences in this study, we have recorded B-K sequences,
suggesting that the Diana monkeys transferred the principle ''B
cancels predation meaning'' to different combinations. An obvious
next step would be to conduct systematic playback experiments
in which the communicative functions of various call
combinations are tested experimentally.
When comparing human language to primate communication
systems, such as this one, a number of interesting similarities and
fundamental differences emerge. First, male Campbell's monkeys
are limited to a relatively small range of messages that they
can convey to their audience. This is partly because callers do not
take full advantage of the potential of their communicative
system. For example, they do not inverse the order of calls (e.g.,
AB to BA) to generate differences in meaning, and a large
number of other possible call combinations are not realized.
Second, human language is symbolic in the sense that signalers
can inform listeners about referents that are not physically
present (55). In Campbell's monkeys, we only observed calling in
response to real life experiences, and some observations suggested
that callers did not attempt to inform others. Further
research will have to test whether some of the observed contingencies
between acoustic morphology and external events were
intentionally produced or biproducts of other processes. Whatever
the outcome, our results demonstrate that the evolution of
complex morphology has begun early in primate evolution, long
before the emergence of hominids, and hence preceded the
evolution of intentional communication.
Are these results relevant for understanding the evolution of
syntax? As outlined before, this system pales in contrast to the
communicative power of grammar; using calls in different
sequences, even highly predictable sequences, is not the same as
grammar. Nevertheless, the call sequences observed in this
primate were not the results of individuals responding to serial
or compound stimuli (such as ''water'' and ''bird'' combined to
''waterbird''). In Campbell's monkeys, the different sequences
were given to highly discrete external events with very little
conceptual similarities.
The degree to which the communication system found in
Campbell's monkey is unique or a general feature of primate
communication is currently unknown. We suspect the latter,
especially for forest primates whose vocal skills must have been
under considerable pressure by natural selection due to high
levels of predation and low levels of visibility in their dim
habitats. The lack of articulatory control, which characterizes
nonhuman primates, may have favored the evolution of combinatorial
signaling.
Materials and Methods
Study Site and Subjects. The study site, a system of trails covering an area of
25 km2, is located in the Taï National Park (5 °50
N, 7 °21W), Ivory Coast, the
largest protected block of rainforest in West Africa. The study lasted from
January 2006 to September 2007 and has been approved by the relevant
government institution of Ivory Coast (Office Ivoirien des Parcs et Re´ serves).
All data were collected from two habituated and four semihabituated groups
of Campbell's monkeys (Cercopithecus campbelli campbelli). These forest
guenons form one-male groups with three to seven adult females and five to
seven juveniles and infants. We observed two male takeovers in both habituated
groups during the study, which increased the number of focal animals
to four habituated males (group 1, male 1 was replaced by male 2 in December
2006; group 2, male 3 was replaced by male 4 in April 2007). It is possible that
such takeovers also took place in the semihabituated groups, but because we
were unable to individually identify any group members, we assumed that
each group had the same male throughout the study period.
Data Collection. Observations were conducted in 15-min blocks under a focal
animal sampling regime (56) between 08:00 and 17:00 GMT. During each
observation period, the observer (K.O.) recorded all loud calls produced by the
focal male. The single adult male is the only one to produce loud calls in
the group, so there was no risk of confusion in terms of caller identification.
If the male produced a loud call, the observer determined its likely cause
among the following: (i) predator, presence of a leopard (Panthera pardus) or
crowned eagle (Stephanoaetus coronatus); (ii) intergroup, presence of neighboring
group (usually inferred by loud calls of their adult male); (iii) tree/
branch, crashing sound of falling tree or large branch; (iv) monkey alarm,
Diana monkey eagle or leopard alarm calls (54); and (v) cohesion and travel,
male spatially separated far from the group (30-50m)or beginning of group
travel after spatial separation.
The four habituated males were directly observed in dense forest vegetation
for a total focal duration of 43 h (male 1, 11 months; male 2, 6 months;
male 3, 15 months; male 4, 2 months). Scan animal samples (n  4,425) were
collected every 30 min, during which we recorded (i) the presence of other
monkey species and (ii) the location within the group's home range. We also
collected ad libitum samples from all habituated and semihabituated groups
during 2,000 h total contact time. This enabled us notably to report on the
response of males to their main predators, leopards (n  3) and crowned
eagles (n  11).
Predator Experiments. As encounters with real predators were rare, we conducted
a series of field experiments during which we presented predator
models to the different males, either by positioning a visual replica of the
predator or by broadcasting recordings of typical predator vocalizations
(leopard growls, eagle shrieks). A custom-made eagle model and a custommade
leopard model were used for all experiments. Both manipulations have
been used successfully before in other studies to simulate predator presence
(e.g., 47, 54).Wetested each of the four stimuli 10 times; that is, 40 trials total,
on seven different males. The four semihabituated males were tested only
once per stimulus category. Some of the habituated males were tested more
than once, with subsequent trials spaced by at least 2 months (N leopard
model;Neagle model;Neagle playback;Nleopard playback: male 13/2/3/2;
male 21/2/1/1; male 32/2/2/3; male 4 was never tested as he arrived toward
the end of the study).
Before each experiment, the following conditions had to be met: (i) the
study group was aware of the presence of human observers for at least 30 min;
(ii) no alarm calls were produced for at least 30 min; (iii) the predator model
(or playback speaker) was positioned ahead of the group's estimated travel
direction ensuring that no associated monkey species could detect it first. One
observer (K.O.) and one field assistant were necessary for these experiments.
The observer walked with the group and recorded the male's behavior, while
the assistant operated the predator model. For eagle trials, the model or
loudspeaker was positioned at an elevation of 2-3 m off the ground. For
leopard trials, the model or loudspeaker was positioned on the ground. Eagle
shrieks were recorded in the study area; leopard growls were purchased from
the National Sound Archives, London. All acoustic stimuli were broadcast with
a Sony WMD6C Professional Walkman connected to NAGRA DSM speakeramplifier
with the amplitude level adjusted so that the calls sounded natural
and could be clearly heard by the group.
Within the predator context, we discriminated between degrees of urgency
(i.e., likelihood of an attack) and compared the composition of sequences.
Low-urgency situations were when the presence of a predator was
revealed by acoustic cues (playback of predator calls, alarm calls by Diana
monkeys); high-urgency situations were when the predator was visible (predator
models; real predator encounters). This distinction was based on the
results of another experimental study with forest African primates, which
showed that call intensity was higher when a predator was detected in the
visual than the acoustic domain (57).
Ouattara et al. PNAS Early Edition  5 of 6
PSYCHOLOGICAL AND
COGNITIVE SCIENCES
ANTHROPOLOGY
Call Recordings and Analyses. All vocal responses were recorded using a
Sony TCD-D100 DAT Walkman, a Sennheiser ME88 microphone, and a
Lavalier microphone for additional observer comments. All calls were
digitized at a sampling rate of 44.1 kHz with 16 bits accuracy. A total of 224
male calling events were collected. In a previous study, we were able to
reliably distinguish six different loud call types according to their acoustic
structure: Boom (B), hok (H), krak (K), krak-oo (K), hok-oo (H), and
wak-oo (W) (50) (Fig. 1). Krak-oo and krak and hok-oo and hok calls were
identical, apart from the optional -oo suffix. Krak-oo calls could be emitted
singly; all other calls were given as part sequences of two or more calls. We
determined the composition of each call sequence and related them to the
eliciting context. To this end, we (i) identified all call types in the sequence,
(ii) calculated the total number of each type produced, and (iii) determined
the positioning of each call type in the sequence. Statistical analyses were
conducted using the Statistica software package, using mainly nonparametric
procedures.
ACKNOWLEDGMENTS. We thank the Office Ivoirien des Parcs et Re´ serves for
permission to conduct research in the Taï National Park, the field assistants of
the Taï Monkey Project for all their invaluable help during data collection, and
N. Elizer, G. Emile, C. Blois-Heulin, V. Biquand, H. Bouchet, J. Fischer, M.
Hausberger, and the GIS (groupement d'inte´ re^ t scientifique) "Cerveau, Comportement,
Socie´ te´" for support and discussions. This work was supported by
the European Commission (Sixth Framework Programme, What it means to be
human), European Science Foundation Eurocore programmes (Origin of Man
Language and Languages), the Wissenschaftskolleg zu Berlin, the French
ministry of foreign affairs (Egide), and the Centre Suisse de Recherches Scientifiques
in Abidjan.
1. Snowdon CT, Hausberger M (1997) Socials Influences on Vocal Development (Univ
Press, Cambridge, UK).
2. Elowson AM, Snowdon CT, Lazaro-Perea C (1998) ''Babbling'' and social context in
infant monkeys: Parallels to human infants. Trends Cogn Sci 2:31-37.
3. Hausberger M, Richard-Yiris M-A, Henry L, Lepage L, Schmidt I (1995) Song sharing
reflects the social organization in captive group of European starling (Sturnus vulgaris).
J Comp Psychol 109:222-241.
4. Tyack PL (2000) Dolphins whistle: A signature tune. Science 289:1310-1311.
5. Lemasson A, HausbergerM(2004) Patterns of vocal sharing social dynamics in captive
group of Campbell's monkeys (Cercopithecus campbelli campbelli). J Comp Psychol
118:347-359.
6. Hausberger M (1997) in Social Influences on Vocal Development, eds Snowdon CT,
Hausberger M (Univ Press, Cambridge, UK), pp 128-156.
7. Yurk H, Barrett-Lennard L, Ford JKB, Matkin CO (2002) Cultural transmission within
maternal lineages: Vocal clans in resident killer whales in southern Alaska. Anim Behav
63:1103-1119.
8. Tanaka T, Sugiura H, Masataka N (2006) Cross-sectional and longitudinal studies of the
development of group differences in acoustic features of coo calls in two groups of
Japanese macaques. Ethology 112:7-21.
9. Hauser MD, Teixidor P, Field L, Flaherty R (1993) Food-elicited calls in chimpanzees:
Effects of food quantity and divisibility. Anim Behav 45:817-819.
10. Townsend SW, Deschner T, Zuberbu¨ hler K (2008) Female Chimpanzees use copulation
calls flexibly to prevent social competition. PLoS ONE 3:e2431.
11. Seyfarth RM, Cheney DL, Marler P (1980) Vervet monkey alarm calls: Semantic communication
in a freeranging primate. Anim Behav 28:1070-1094.
12. Rainey HJ, Zuberbu¨ hler K, Slater PJB (2004) Hornbills can distinguish between primate
alarm calls. Proc R Soc London B 271:755-759.
13. Zuberbu¨ hler K, Cheney DL, Seyfarth RM (1999) Conceptual semantics in nonhuman
primates. J Comp Psychol 113:33-42.
14. Slocombe KE, Zuberbu¨ hler K (2005) Agonistic screams in wild chimpanzees (Pan
troglodytes schweinfurthii) vary as a function of social role. J Comp Psychol 119:67-77.
15. Basile M, Boivin S, Boutin A, Blois-Heulin C, Hausberger M, Lemasson A (2009) Socially
dependent auditory laterality in domestic horses (Equus caballus). Anim Cogn 12:611-
619.
16. Basile M, Lemasson A, Blois-Heulin C (2009) Social and emotional values of sounds
influence human (Homo sapiens) and non-human primate (Cercopithecus campbelli)
auditory laterality. Plos One 4:e6295.
17. Hauser MD, Anderson K (1994) Left hemisphere dominance for processing vocalizations
in adult, but not infant, rhesus monkeys: Field experiments. Proc Natl Acad Sci
USA 91:3946-3948.
18. George I, Cousillas H, Richard J-P, HausbergerM(2008)Apotential neural substrate for
processing functional classes of complex acoustic signals. PLoS ONE 3:e2203.
19. Bo¨ye M, Gu¨ ntu¨ rku¨ n O, Vauclair J (2005) Right ear advantage for conspecific calls in
adults and subadults, but not infants, California sea lions (Zalophuscalifornianus):
Hemispheric specialization for communication? J Neurosci 21:1727-1732.
20. Chomsky N (1981) Knowledge of language: Its elements and origins. Print Trans R Soc
London B 295:223-234.
21. Pinker S (1994) The Language Instinct: How the Mind Creates Language (William
Morrow, New York, NY).
22. Hauser MD, Chomsky N, Fitch WT (2002) The faculty of language: What is it, who has
it, and how did it evolve? Science 298:1569-1579.
23. Yip M (2006) The search for phonology in other species. Trends Cogn Sci 10:442-446.
24. Aitchison J (2000) The Seeds of Speech: Language Origin and Evolution (Univ Press,
Cambridge, UK).
25. Geissmann T (2002) Duet-splitting and the evolution of gibbon songs. Biol Rev Camb
Philos Soc 77:57-76.
26. Suzuki R, Buck JR, Tyack PL (2006) Information entropy of humpback whale songs.
J Acous Soc A. 119:1849-1866.
27. Clarke E, Reichard UH, Zuberbu¨ hler K (2006) The syntax and meaning of wild Gibbon
song. PLoS ONE 1:e73.
28. Vallet E, Beme I, KreutzerM(1998) Two-note syllables in canary songs elicit high levels
of sexual display. Anim Behav 55:291-297.
29. Holland J, Dabelsteen T, Paris AL (2000) Coding in the song of the wren: Importance of
rhythmicity, syntax and element structure. Anim Behav 60:463-470.
30. Clucas BA, Freeberg TM, Lucas JR (2004) Chick-a-dee call syntax, social context, and
season affect vocal responses of Carolina chickadees (Poecile carolinensis). Behav Ecol
Sociobiol 57:187-196.
31. Hauser MD, Newport EL, Aslin RN (2001) Segmentation of the speech stream in a
nonhuman primate: Statistical learning in cotton-top tamarins. Cognition 78:B53-B64.
32. Gardner RA, Gardner B (1969) Teaching sign language to a chimpanzee. Science
165:664-672.
33. Premack D (1971) Language in a chimpanzee? Science 172:808-822.
34. Patterson F (1981) Can an ape create a sentence? Some affirmative evidence. Science
211:86-87.
35. Savage-Rumbaugh ES, Roger L (1996) Kanzi: The Ape at the Brink of the Human Mind
(Wiley & Sons, New York, NY).
36. Crockford C, Boesch C (2005) Call combinations in wild chimpanzees. Behaviour
142:397-421.
37. Clay Z, Zuberbu¨ hler K (2009) Food-associated calling sequences in bonobos. Animal
Behav 77:1387-1396.
38. Arnold K, Zuberbu¨ hler K (2006) Language evolution: Semantic combinations in primate
calls. Nature 441:303.
39. Arnold K, Zuberbu¨ hlerK(2008) Meaningful call combinations in anon-humanprimate.
Curr Biol 18:202-203.
40. Freeberg TM, Lucas JR (2002) Receivers respond differently to chick-a-dee calls varying
in note composition in Carolina chickadees, Poecile carolinensis. Anim Behav 63:837-
845.
41. Templeton CN, Greene E, Davis K (2005) Allometry of alarm calls: Black-capped chickadees
encode information about predator size. Science 308:1934-1937.
42. Bergman TJ, Beehner JC, Cheney DL, Seyfarth RM (2003) Hierarchical classification by
rank and kinship in baboons. Science 302:1234-1236.
43. Cheney DL, Seyfarth RM (2007) Baboon Metaphysics (University of Chicago Press,
Chicago, IL).
44. Lemasson A, Gautier J-P, Hausberger M (2003) Vocal similarities and social bonds in
Campbell's monkey (Cercopithecus campbelli). C R Biol 326:1185-1193.
45. Lemasson A, Richard J-P, Hausberger M (2004) A new methodological approach to
context analysis of call production. Bioacoustics 14:111-125.
46. Lemasson A, Hausberger M, Zuberbu¨ hler K (2005) Socially meaningful vocal plasticity
in Campbell's monkeys. J Comp Psychol 119:220-229.
47. Ouattara K, Zuberbu¨ hler K,N
Goran KE, Gombert J-E, LemassonA(2009) The alarm call
system of female Campbell's monkeys. Anim Behav 78:35-44.
48. Gautier J-P, Gautier-HionA(1977) inHowAnimals Communicate, ed Sebeok T (Indiana
Univ Press, Bloomington, IN), pp 890-964.
49. Zuberbu¨ hler K (2002) A syntactic rule in forest monkey comunication. Anim Behav
63:293-299.
50. Ouattara K, Lemasson A, Zuberbu¨ hler K (2009) Campbell's monkeys use affixation to
alter call meaning. PLoS ONE 4:e7808.
51. Struhsaker TT (1970) in Old World Monkeys: Evolution, Systematics and Behaviour, eds
Napier JR, Napier PH (Academic, New York, NY), pp 365-444.
52. Zuberbu¨ hler K (2001) Predator-specific alarm calls in Campbell's guenons. Behav Ecol
Sociobiol 50:414-422.
53. Lemasson A, Blois-Heulin C, Jubin R, Hausberger M (2006) Female social relationships
in a captive group of Campbell's monkeys (Cercopithecus campbelli campbelli). Am J
Primatol 68:1161-1170.
54. Zuberbu¨ hler K, Noe¨ R, SeyfarthRM(1997) Diana monkey long-distance calls: Messages
for conspecifics and predators. Anim Behav 53:589-604.
55. Deacon T (1997) The Symbolic Species (WW Norton & Company, New York, NY).
56. Altmann J (1974) Observational study of behaviour: Sampling methods. Behaviour
49:227-265.
57. Schel AM, Tranquilli S, Zuberbu¨ hler K (2009) The alarm call system of two species of
black-and-white colobus monkeys (Colobus polykomos and Colobus guereza). J Comp
Psychol 123:136-150.
6 of 6  www.pnas.orgcgidoi10.1073pnas.0908118106 Ouattara et al.

-----Original Message-----
From: Linguistic Anthropology Discussion Group [mailto:LINGANTH at LISTSERV.LINGUISTLIST.ORG] On Behalf Of Celso Alvarez Cáccamo
Sent: Sunday, December 13, 2009 7:23 PM
To: LINGANTH at LISTSERV.LINGUISTLIST.ORG
Subject: Re: Fwd: Campbell's Monkeys, in Ivory Coast, Are Seen as Using Syntax - NYTimes.com

Thank you, Hal and Eric.  Eric, I suppose the PNAS article has another
title, but I suppose it's basically the same research, right?

-celso