Group Title: analysis of vocalizations of three species of east African cercopithecidae
Title: An Analysis of vocalizations of three species of east African cercopithecidae
Full Citation
Permanent Link:
 Material Information
Title: An Analysis of vocalizations of three species of east African cercopithecidae
Physical Description: xii, 109 leaves. : illus. ; 28 cm.
Language: English
Creator: Walek, Mary Louise, 1946- ( Dissertant )
Maples, William R. ( Thesis advisor )
Fairbanks, Charles H. ( Reviewer )
Markel, Norman N. ( Reviewer )
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 1972
Copyright Date: 1972
Subject: Animal communication   ( lcsh )
Sound production by animals   ( lcsh )
Anthropology thesis Ph. D   ( lcsh )
Dissertations, Academic -- Anthropology -- UF   ( lcsh )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
Thesis: Thesis -- University of Florida.
Bibliography: Bibliography: leaves 103-108.
Additional Physical Form: Also available on World Wide Web
General Note: Typescript.
General Note: Vita.
 Record Information
Bibliographic ID: UF00097647
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000577478
oclc - 13982886
notis - ADA5173


This item has the following downloads:

PDF ( 5 MBs ) ( PDF )

Full Text






Copyright by

MEary Louise ",'.alek


With Love
Mom, Dad, and Bruce


I would like to thank Dr. William R. Maples, chairman

of my supervisory committee, for his guidance and criticism,

and for his assistance in making this study possible. It was

he who made my trip to East Africa a reality and for this I

will always be grateful. Appreciation must also be extended

to the other members of the committee, Drs. Norman N. Market

and Charles H. Fairbanks, for taking time from their busy

schedules to read and evaluate this study,

I would also like to thank Dr. Arnold Paige of the

Departments of Speech and Electrical Engineering of the Uni-

versity of Florida for giving unselfishly of his time and

for his generosity in making the sonograph and other mater-

ials accessible to me. Without his kind cooperation this

study would not have been possible.

For the use of their laboratory and additional tape

recorders during the editing process, I would like to thank

the Department of Communication Sciences, also of the Univer-

sity of Florida.

To Bruce Walek I extend sincere appreciation for the

photographic reproduction of the sonograms, even though it

meant taking valuable time from his own studies to do so.

Finally, I would like to express my gratitude to the

members of my family for their support and encouragement

which contributed as much to the completion of this sTudy as

the technical assistance received elsewhere.

The study was financed by grants to William R. Maples

from Biomedical Sciences of the University of Florida and

the National Science Foundation (GS 2610).




LIST OF TABLES . . . . . . .

LIST OF FIGURES . . . . . .



. * * . . * . 0

. . . .


* . t

* . .


The Study of Primate Vocalizations
Objectives of the Present Study
The Study Sites . . .
Methods and Materials of Research
Definition of Terms . . .


Classification of Calls ...
Discrete and Graded Signals . .
A Structural-Functional Analysis
Taxonomic Implications . . .


Classification of Calls .


Vocalizations and Crop-Raiding .
Classification of Calls . . .


The Sykes Study . . . . .
The Colobus Study . . . .
The Baboon Study . . . . .

. . . 1

. . . 2

. . . 6
. . . 7

. . . 7
. .

. . . .

* 61

. . . 70

. . . 70
. . . 72

. . 77


SYKES MONKEY . . . . . .

e .

e f
t e
* .



COLOBUS MONKEY . . . . . . . 95

BABOON . . . . . . . . . . 100

REFERENCES . . . . . . . . . 103

BIOGRAPHICAL SKETCH . . . . . . . . .. 109








3 Nyah

4 Short N

5 Boom

6 Squeal

7 Shriek

8 Chut

9 Chutter

10 Low, Sh

11 Chutter

12 Growl

13 Err

14 Uh .

15 Snort

16 Snort-C

e . 0 .

e Chirps

yah ..

* * .

ort Chutter

-Squeal .

roak . .

17 Gecker White



Err White Infant

Gecker Juvenile

Two-Phase Bark .

Bark . . . .

Yak . . . .


* C C
* . C

* C *

* C C

. .C.
* * *
* . .

. . .

. . .

* * *

nt .

* . *

* *

* . .
















































Chirp . . .

Chir . . .

Two-Unit Chirp .

Two-Unit Chirp .

Three-Unit Chirp .

Nyah . . .

Short Nyah . ..

Short Nyah . .

Boo . . . .

Scueal . . .

Shriek . . .

Chut . . . .

Chut . . . .

Chutter . . .

Low, Short Chutter

Chutter-Squeal .

Chutter-Squeal (coi

Chutter-Squeal (co

Growl . . .

Growl . . .

Err . . .

Uh . . . .

* *



* .

























Figure Page

23 Snort .. .... .. ... 96

2 Snort . . . . . . . . . . 96

25 Snort-Croak . . . . . . 97

26 Snort-Croak . . . .. . . . 97

27 Snort-Croak ... . . . . . . 97

28 Gecker White Infant . . . . . 98

29 Err White Infant . . . . . . 98

30 Gecker Juvenile . . . . . . 99

31 Gecker Juvenile ... . . . . . 99

32 Two-Phase Bark . . . . . . . 101

33 Two-Phase Bark . . . . . . . 101

34 Bark I . . . . . . . . 102

35 Bark I, Yak, Bark II . . . . . . 102

Abstract of Dissertation Presented tc the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



Mary Louise Walek

August, 1972

Chairman: William R. Maples
Major Department: Anthropology

Vocalizations of three species of Old World monkeys

(Sykes monkey, Cercopithecus mitis; colobus monkey, Colobus

oolvkomos; and baboon, Papio cynocephalus) were collected in

Kenya, East Africa. Sonograms of the vocalizations were made

and analyzed in terms of their physical characteristics (fre-

quency, duration, and tonality). Representative sonograms

are illustrated and tables of all measurements made are pre-

sented. Accompanying behavioral information is also included

whenever available.

Fourteen Sykes vocalizations are catalogued and de-

scribed. An analysis of the calls in terms of discreteness

and gradedness revealed that most of the vocalizations were

of the discrete type.

The Sykes vocalizations were also examined to see if

physical properties, or structure, could be related to func-

tion. It was found that those calls transmitted over dis-

tances which were important for group cohesion and alarm or


threat were structured so as to facilitate locating the caller.

Calls operating at close-range did not have such localizing


The question of whether Cerconithecus mitis is con-

specific with C. nictitans has been raised in the literature.

This study evaluated the taxonomic status of these two species.

Based on similarities and differences in the vocalizations it

was concluded that separate specific identity should be main-

tained. However, C. mitis was found to be more closely re-

lated to C. nictitans and C. erythrotis-ceDhus than to the

mona superspecies.

Five vocalizations of the colobus monkey are described:

two for infants, one for juveniles, and two for adults. A

nonvocal behavior pattern, the jumping display, was found to

be of communicative significance in alerting a troop to dan-

ger and in indicating the direction of troop movement. It

also functioned as an intimidation display directed at the


The role which vocalizations play in the crop-raiding

activities of baboons was investigated and the apparent

"decoy" effect of the calls in distracting the attention of

Africans protecting the fields was examined. It was concluded

that the vocalizations did not represent a deliberate and

planned decoy tactic, but instead were a fortuitously diver-

sionary behavior which contributed incidently 7o the success

of the raids. Four types of calls given during the raids

are described.




The study of primate communication is assuming an

increasingly important role in the field of primatology. As

more is learned of the behavior of man's closest relatives,

the nonhuman primates, the more apparent it has become that

understanding how behavior patterns are transmitted and in-

terpreted will contribute considerably to understanding the

adaptive mechanisms of social life. Many of the earlier

studies on primate behavior, and even some of the more recent,

either failed to include communication as an aspect of behav-

ior to be investigated or else dealt with it in such a cursory

and unscientific manner as to make it almost worthless for

further investigation and comparison. Part of the problem

is no doubt related to the all-pervasive nature of communica-

tion into all aspects of social life, a fact which is well

illustrated by the following definition: " commun-

ication is the process by which the behavior of an individual

affects the behavior of others" (Altmann, 1967:326).

Communication as it has been defined can include many

different forms of behavior occurring in concert and involving

several individuals. It becomes apparent, therefore, that

the study of communication can be a complex undertaking.


Furthe- complications become apparent when one considers

that communication can take place via several different chan-

nels. Signals or messages may be olfactory, tactile, visual,

or auditory in nature and for any specific signal transmission

several of these channels of communication may be used simul-

taneously. Each of these channels has been investigated in

studies of communication among the primates (see Andrew,

1963a, 1964; Bolwig, 1964; Hinde and Rowell, 1962; Jolly,

1966; Moynihan, 1967; van Hooff, 1962).

The Study of Primate Vocalizations

The primary concern of the present study, however,

will be with primate vocal communication, the study of which

has taken various forms.

One approach consisted of transcribing the various

calls by using letters of the alphabet in order to give as

close an approximation as possible of the sound of the call

(for an example see Ullrich, 1961). Not only was this sys-

tem largely subjective in nature, but the alphabetical nota-

tion used was generally that of the language of the observer

and hence was not accessible to, or reproducible by, researchers

of all nationalities. In order to overcome some of the sub-

jectivity of such an approach and to insure a certain amount

of uniformity, there was an attempt to use the standard

International Phonetic Alphabet (for an example see Hill and

Booth, 1)57). Howv-er, it soon became apparent that non-

human priates were capable of producing sounds which

lacked counterparts in human speech. Ph:retic and alphabet-

ical transcriptions, therefore, could not render an accurate

description of many vocalizations and often gave a false,

human quality to the vocal patterns of primates other than


The most common approach to the study of primate

vocal communication until recently has been the assignment

of a descriptive name to a vocalization, one which depicts

as closely as possible the type of call given. In this way

such terms as 'squeal', 'click', and 'grunt' were used but

were really only of value when as complete an acoustic de-

scription as possible accompanied the assignment of such a

term. Moreover, adjectives such as 'high-pitched', 'shrill',

'harsh', etc., used in such descriptions, were still ambig-

uous and subjective in nature. Hence, comparisons of vocal-

izations reported in different studies remained difficult,

if not impossible, to carry out.

The problem of subjectivity was alleviated to a con-

siderable extent with the development within the last two

decades of sophisticated, sensitive recording devices and

the invention of the sound spectrograph. The spectrogram,

or sonogram, supplies a permanent, visual record of a sound

and serves both as a form of notation and as a method of

precise measurement. Although descriptive terms remain prev-

alent as names or labels for the various calls, they are now

supported by a more concrete, objective, physical description

than was possible before.

Receiving its impetus largely fom studies on insects,

anurans, and birds, the use of sonograms in the study of pri-

mate vocalizations has spread rapidly within the last several

years. Most of the work on primate communication has been

with species of Old World monkeys, especially the macaques,

baboons, and vervets. Rowell (1962) used sonograms and de-

scribed nine agonistic noises of the rhesus macaque, relating

each to the various postures and expressions which they ac-

company. Struhsaker (1967) has done an excellent spectro-

graphic analysis of vervet vocalizations combining a detailed

physical description of the calls with behavioral data when-

ever possible. Although numerous studies on baboon communi-

cation have been reported, sonograms for this species are

apparently lacking in the published literature. Communica-

tion studies involving spectrographic analysis have also been

reported on several other species, such as those of Reynolds

(in Marler, 1965) on the chimpanzee, Schaller (1963) on the

gorilla, Moynihan (1964) on the night monkey, Altmann (1967)

on howlers, Ploog (1966) on squirrel monkeys, Petter and

Petter (1966) on the aye-aye, and Petter (1962) on lemurs.

For a more complete listing of studies on primate communica-

tion, spectrographic or otherwise, the reader is referred to

Altmann's (1968) comprehensive review.

Although Old World monkeys have been extensively

studied in terms of behavior and communication, several

genera within the family Cercopithecidae have not been ade-

quately investigated. Among these are the species of the

genera Cercopithecus and Colobus.

Other than Struhsaker's (1967) study on the vervet,

Cercopithecus aethiops, and most recently his investigations

of the vocalizations of several Cercoithecus species (1970),

very little information on vocalizations is available for the

Cercopithecus group. Most of what is known of the auditory

communication of these species has been drawn out of more gen-

eral behavioral studies or from anecdotal accounts. Other

researchers on the vervet include Booth (1962), Gartlan and

Brain (1968), and Hall and Gartlan (1965). None of these

studies presents spectrographic descriptions. The redtail

monkey, C. ascanius, is the only other species investigated

to any extent and the reader is referred to the works of

Buxton (1952) and Haddow (1952). Again, spectrographic data

is not available for this species in these studies.

Very little is kno'n of the behavior of the Colobus
species. While considerable research has been carried out

on the Asiatic counterpart of the Colobinae, the langurs, the

African members of this subfamily have been largely ignored.

Available reports include those of Booth (1957) on the olive

colobus and Hill and Booth (1957) and Ullrich (1961) on the

black and white colobus. However, information as to social

communication is rather sparse and incomplete in these

studies, A more informative sTudy is Marler's (1970) spec-

trographic analysis of the vocalizations of the red colobus.

Objectives of the Present Study

In the present study the vocalizations of two species

of these relatively little-studied genera, the Sykes monkey,

Cercooithecus mitis kibonotensis, and the black and white

colobus monkey, Colobus Dolykomos Dalliatus, will be described

and analyzed. The basic aim of this study is to present a

catalogue of the vocal repertoire of these species. Accom-

panying behavioral information will also be presented when-

ever available.

The vocalizations of the baboon, Papio cynocechalus,

will also be described. This investigation formed part of

a larger study being conducted at the time by William R.

Maples, William F. Greenhood, and myself on crop-raiding be-

havior of baboons. The aim of this portion of that study is

to describe those vocalizations given while crop-raiding and

to investigate the role they play in this activity.

All descriptions and classifications of the calls are

based on spectrographic analysis.

The Study Sites

The field observations upon which the following study

is based were conducted in Kenya, East Africa, from September

to December, 1969. The study site for the Sykes and colobus

monkey material was at Diani Beach on the Kenya coast, approx-

imately eighteen miles south of rPiombassa. The site itself

was approximately one-quarter to one-half mile inland. The

material on baboon vocalizations was collected at farm sites

primarily at Shimoni, a small fishing village near the

southern tip of the Kenya coast. Some recordings were also

made near Roka along the Mombassa-::Ialindi road, approximately

twenty-four miles south of Malindi.

In general, the climax vegetation community for the

Diani Beach/Shimoni area has been described as lowland dry

forest on coral rag (Combretum schumanii-Cassipourea species)

(Moomaw, 1960). The forests at Diani Beach are broken into

smaller segments of varying size by the interspersal of

lantana thicket and cultivated land. The forest essentially

consists of two layers, an upper canopy composed of tall,

branching trees and a low, dense understory where bushes and

vines predominate. The colobus were observed to favor the

upper canopy but would readily take refuge in the lower layer

when necessary. The Sykes seemed equally at home in both

layers but it was this observer's impression that more time

was spent in the understory. At Roka, the baboons inhabited

a protected forest which has been classified as lowland

woodland (Brachystegia-Afzelia species) by Moomaw (1960).

For a more detailed discussion of the ecology of these areas

and the species of vegetation which predominate, the reader

is referred to Moomaw (1960) and Greenhood (1971).

Methods and Materials of Research

Recordings of the vocalizations ;were made with a

Nagra IV L portable tape recorder at 15 ips with a frequency

response of 30 to 20,000 cps, and a Sennheiser directional

microphone, Miodel MKH 804. Sonograms for analysis were made

on a Kay Electric Sona-Graph, .iIodel 6061-B. A transparent

plastic grid overlay was made showing time on the horizontal

axis and frequency on the vertical axis. The grid vwas placed

over the sonograms and approximate frequency and temporal

measurements were taken.

Field recording techniques at Diani Beach consisted

of approaching as closely and quietly as possible on foot a

Sykes or colobus troop. The thick, tangled vegetation of

this area precluded any possibility of following a troop on

its daily movements; therefore, recordings were made along

roads and footpaths vhich troops regularly frequented and

traversed. At Shimoni, a stationary recording site was set

up within the farm and baboon vocalizations were recorded as

the troops approached and entered the field to raid. The

recording station at the Roka site was within an observation

vehicle which would follow and remain adjacent to the move-

ments of a troop of crop-raiding baboons as they emerged from

the forest and crossed the Mi'alindi-Mombassa road to raid the

farms on the other side. In all cases, accompanying behavioral

observations were recorded in a notebook.

In areas of limited visibility it was the observation

of behavioral context which proved to be the most difficult.

A single individual attempting to operate a tape recorder,

direct a microphone, maintain visual contact with a troop

moving in dense underbrush, and take notes is limited as co

the amount of behavior which can be observed. This linita-

tion was particularly felt in the Diani Beach area. At Shi-

moni and Roka the problem was not as acute as other workers

(notably William R. Maples) were simultaneously observing

crop-raiding behavior and so correlation of the vocalizations

with their observations was possible. It would seem, there-

fore, that perhaps the most fruitful means of studying com-

munication with all its behavioral ramifications within a

somewhat hostile environment would be through a team approach.

Such a system would minimize the physical limitations and

would facilitate the recording of observations. In addition,

it would provide a system of checks and counter-checks on

what was observed. Perhaps in savanna areas, with visibility

relatively unimpaired, these difficulties do not pose as seri-

ous a hinderance as when dealing with forest populations.

In the latter case, the physical limitations imposed by the

very nature of the study itself are only too apparent.

Definition of Terms

So that there may be some form of common reference in

discussions of primate vocalizations, use will be made here

of the terms defined by Thomas T. Struhsaker in his study on

auditory communication of vervet monkeys. These are as


Unit: The unit is the basic element of a call, and
is represented as a continuous tracing along the
temporal (horizontal) axis of the sonogram.

Phrase: The phrase is a group of units that is
separated from other similar groups by a time in-
terval greater than any time interval separating
the units within a phrase.

Bout: A bout is a grouping of one or more phrases
separated from other similar groupings by a time
interval greater than that separating any of the
phrases within a bout.

Nontonal unit: A nontonal unit is composed of
sound that is more or less continuously developed
over a wide range of frequencies. This has also
been called "noise" by Andrew (1963b) and "harsh
noises" by Rowell and Hinde (1962).

Tonal unit: A tonal unit is composed of sound
characterized by one or more relatively narrow fre-
quency bands and has been referred to as "clear calls"
by Rowell and Hinde (1962) and "sound" by Andrew
(1963b). Units with a harmonic structure are in-
cluded in this category.

Compound unit: A compound unit is composed of both
nontonal and tonal sounds that appear as a continu-
ous tracing on the sonogram.

Mixed unit: Units composed of both tonal and non-
tonal sounds that are rather superimposed on one
another are called mixed units. The tonal and non-
tonal aspects are more or less separated by differ-
ences in frequency.

Distribution of major energy of nontonal sounds:
The distribution of the major energy of a nontonal
sound is represented and thus determined by the
darkest portion of the tracing on the sonogram.
This distribution is generally over a smaller range
than the frequency range of nontonal sounds. The
distribution of major energy of nontonal sounds is
described by an upper frequency (highest pitch of
major energy) and a lower frequency (lowest pitch
of major energy). (Struhsaker, 1967:282-283)

For a brief demonstration of how these terms will be

applied in the present study, the reader is referred to

Figure 26 and Figure 6.

Figure 26 is a spectrographic representation of a

colobus vocalization consisting of three different phrases.

The first is a nontonal, single-unit phrase and is given

only once at the beginning of the call. The second phrase

consists of several tonal units. The third is a nontonal,

single-unit phrase very similar to the first. The second

and third phrases are repeated in alternation. All three

phrases are said to form a bout, part of which is depicted in

the sonogram. The entire bout can be referred to as a "call"

or "vocalization".

Figure 6 represents a call of Cercopithecus mitis

and can be described as a single-unit phrase. It is also a

compound unit since a brief tonal component consisting of

gently arching frequency bands is followed by a noise or

nontonal component with a somewhat irregular appearance on

the soncgram.

For a further demonstration of how these terms are

used, the reader is referred to the descriptions of the vari-

ous calls presented in the text which can be compared with

the appropriate sonogram.



The Sykes monkeys (Cercopithecus mitis kibonotensis)

live in social groups of substantial size. Various troop

counts, made while the monkeys passed over an "arboreal

bridge" spanning a roadway, revealed totals of 13, 18, 26,

and 31. However, it never was clear how many troops were in-

volved or if an entire troop had been counted. Sykes are

primarily an arboreal species but, as mentioned previously,

seem to be equally at home in the tall trees as in the lower

brush. They seem not the least bit reluctant to move along

on the ground, especially in those areas where the vines are

quite thick and tangled.

The Sykes would vocalize quite readily. The chirp

was the most frequently heard call and would be elicited by

any sudden movement or noise, or by the approach of humans

or dogs. In the initial stages of the study, the observer

was constantly the object of much threat behavior and chirping.

However, as the Sykes became habituated to the observer's

presence and with the thickening of the foliage due to rain,

the amount of vocalizations given was considerably reduced.

Classification of Calls


The chirp (Figs. 1&2 and Table 1) was given by adult

females and immature monkeys and possibly by adult males,

although positive identification of the latter was not possible.

That adult male Sykes might give the call is doubtful in light

of Struhsaker's (1967, 1970) reports for the vervet and other

Cercopithecus species in which he noted that chirps were not

given by adult males.

The chirp was an abrupt, sharp-sounding call of high

frequency and short duration. The frequency range was 0.10 -

16.0+ kHz but the average distribution of the major concentra-

tion of energy was 2.91 6.14 kHz (N=29). Sonograms re-

vealed a single-unit phrase with an abrupt beginning consist-

ing of a narrow, tonal, chevron-shaped component followed by

a noise component of longer duration. The noise component of

approximately half of the chirps was distributed in a two-

band arrangement, while the other half consisted of a single,

broader band (Figs. 1&2). The mean duration of the chirp was

0.18 seconds (N=29).

The chirp apparently functions as a threat-alarm call.

IT is usually given in conjunction with a threat display in

which the monkey either crouches forward with legs stiff and

bobs its head and shoulders up and down, or else lunges for-

wvrd on its forelimbs. In both cases the monkey stares in-

Sently at the intruder. Thic behavior was directed toward

do-s n. lunans but never was observed between one Sykes and

another. A similar observation was made by Haddow (1952) in

his study on the redtail monkey, Cercopithecus ascanius.



Number (N) Mean (x) Range (R)

Units/Phrase 29/29 1.00 1.00-1.00

Phrase Duration 29 0.18 0.10-0.25

Lower Limit of
Major Energy 29 2.91 1.50-4.00

Upper Limit of
Major Energy 29 6.14 3.00-8.00

Chirps were also given by a feeding animal, the

feeding in some instances being an apparent nervous displace-

ment activity but on other occasions a true feeding activity.

The frequency of repetition of the chirps was vari-

able as was the intensity of the call. Some chirps were

given loudly and in rapid succession; for example in re-

sponse to a loud, sudden noise or to a dog passing nearby.

On other less tense occasions chirps were given infrequently

and softly, almost whisper-like.

Other Sykes responded to the chirping animal by look-

ing in its direction and in the direction of its threat and

by chirping also. Occasionally the other animals would move

deeper into the brush but this was not necessarily a hurried

retreat. In some instances the only Sykes remaining visible

was the chirping individual who would soon move off and join

the others.

Multiple Chirps

Included in the Sykes' repertoire in addition to the

single-unit chirps are two- and three-unit chirps (Figs. 3-5

and Table 2). Such multiple chirps were also reported by

Struhsaker (1970) for Cercopithecus cephus, C. erythrotis,

and C. nictitans. As is the case for single chirps, multiple

chirps are apparently restricted to adult females and imma-

ture monkeys.

In the two-unit chirp, the distribution of major

energy in the second unit is over a greater range (2.84 -

5.96 kHz) than in the first (3.80 5.96 kHz). This distinc-

tion was also made by Struhsaker (1970) for the three Cerco-

pithecus species listed above. Struhsaker found that this

shorter frequency range of the first unit was due both to a

higher lowest frequency and a lower upper frequency. However,

in the case of C. mitis only the lowest frequencies were

found to differ considerably. That is, in 88.9 percent of

the double chirps the lowest frequency of Unit I was higher

than the lowest frequency of Unit II, whereas the upper fre-

quencies of the two units were the same in 66.7 percent of

the calls (N=9).

Generalizations concerning energy distribution for

the three units of the triple chirp can not be made since

only one such call was recorded. However, in this one case

energy distribution increased with each subsequent unit.



Number (N) Mean (R) Range (R)

Double 18/9 2.00 2.00-2.00
Triple 3/1 3.00 3.00-3.00

Unit Duration
Double Unit I 9 0.06 0.05-0.08
Unit II 9 0.20 0.12-0.28
Triple Unit I 1 0.05 0.05-0.05
Unit II 1 0.08 0.08-0.08
Unit III 1 0.25 0.25-0.25
Interval Between
Units secss.)
Double 9 0.00 0.00-0.00
Triple 2 0.03 0.05-0.01

Phrase Duration
Double 9 0.25 0.18-0.32
Triple 1 0.44 0.44-0.44

Lower Limit of
Major Energy
Double Unit I 9 3.80 3.00-5.00
Unit II 9 2.84 2.00-3.60

Triple Unit I 1 2.00 2.00-2.00
Unit II 1 2.00 2.00-2.00
Unit III 1 2.00 2.00-2.00

Upper Limit of
Major Energy
Double Unit I 9 5.96 5.50-7.00
Unit II 9 5.96 5.00-7.00

Triple Unit I 1 3.00 3.00-3.00
Unit II 1 3.50 3.50-3.50
Unit III 1 5.00 5.00-5.00

In addition to frequency differences, the durations

of the two units of the double chirp also differ, with the

first unit being shorter than the second in all cases. The

durations of the units of the three-unit chirp also increase

from first to third, but again any generalization must await

a larger sample.

It is interesting to note that in both energy dis-

tribution and duration the final units of both the two- and

three-unit chirps fall within the range of values for the

single-unit chirp.

Nyah and Short Nyah

The nyah (Fig. 6 and Table 3) is a loud call consist-

ing of tonal and nontonal components. Although the call

itself consists of a single-unit phrase, it always occurred

in bouts in which it was repeated at intervals of approxi-

mately 0.50 seconds.

On the sonogram the nyah appears as a harmonic series

of bands beginning abruptly and arching into a mixed band of

tonality and noise. In a variant of the nyah, referred to

here as the short nyah (Figs. 7&8 and Table 4), this harmonic

band pattern at the beginning is absent, the call, therefore,

sounding like a sharp "ah". Both the nyah and the short nyah

can be given in a bout, but it seems the latter was usually

given under tense or excitable conditions, such as a pack of

stray dogs passing nearby.

Mean duration for the nyah was 0<31 seconds (N=12)

and 0.19 seconds for the short nyah (N=14). The average

energy distribution for both forms quite similar, with

0.82 5.46 kHz for the nyah (N=15) and 0.57 5.31 kHz for

the short form (N=l4). Total energy distribution ranged

from 0.10 16.0+ kHz for both calls. Since it was never

known exactly how many Sykes groups were in the area and since

individual groups were not distinguishable, it is possible

that all the nyahs of the sample were made by the same indi-

vidual. This possibility is based on the assumption that there

is but one adult male per group, a common occurrence among

Cercopithecus monkeys according to Struhsaker (1969). The

short nyahs of the sample, however, were given by one indi-

vidual only since they were all recorded on a single occa-

sion with only one troop present.



Number (N) Mean (x) Range (R)

Units/Phrase 15/15 1.00 1.00-1.00

Phrase Duration 12 0.31 0.20-0.38

Lower Limit of
Major Energy 15 0.82 0.25-1.80

Upper Limit of
Major Energy 15 5.46 4.00-7.00



Number (N) Mean (x) Range (R)

Units/Phrase 14/14 1.00 1.00-1.00

Phrase Duration 14 0.19 0.05-0.28

Lower Limit of
Major Energy 14 0.57 0.20-2.30

Upper Limit of
Major Energy 14 5.31 4.00-6.50

The animal giving the call was very difficult to
observe; in fact, this was done on only one occasion and quite

by accident. A single Sykes was observed sitting alone in

the topmost branches of a tall tree giving the call at about

18:00. No other Sykes were visible in the immediate area.

The call was made by exhaling forcibly with the mouth open.

This individual was too far away to permit accurate sexual

identification but its large size would indicate it to be

most likely a male. This assumption is further supported by

Haddow's (1952) observation of the redtail monkey, C. ascanius,

in which only the adult male gives a call Haddow described as

"Kyuh". In addition, the nyah and short nyah of the Sykes

are probably equivalent to the "loud calls" reported by

Struhsaker (1970) and Bourliere, Hunkeler, and Bertrand (1970)

for other Cercopithecus species which were also only given by

adult males.

Struhsaker's (1970) "loud calls" are of two types,

the Dow and the hack. Since he does not make explicit the

basis for distinguishing them spectrographically, and since

he does not do so in his table of measurements, it is impos-

sible to compare or correlate one or the other with the nyah

or short nyah.

The nyah call was frequently given prior to group

progression. It may have signaled group withdrawal, for soon

after the call was given no Sykes would be visible in the area,

having moved in the direction from which the call came. In

addition, such calls seem to be evoked by a potentially dis-

ruptive or threatening stimulus as was the case with the pack

of dogs. Bourliere, Hunkeler, and Bertrand report that loud

calls are given by the alpha male in the evening at the

sleeping trees and also "in response to any situation that

was potentially dangerous for the group members, or potenti-

ally disruptive to the cohesion of the group" (1970:313).

Haddow (1952) and Struhsaker (1969) concluded that the loud

calls may function as a central focal point for maintaining

intragroup cohesion on occasions when the group might be dis-

persed and may also function in intergroup spacing.


On several occasions a low frequency call was heard

which sounded very much like "whoo" but which closely resem-

bles in frequency and duration the booms of the mona super-

species as described by Struhsaker (1970). The boom (Fig. 9

and Table 5) is completely tonal and appears as a very

black band situated at the bottom of the sonogram. The call

is given as a single-unit phrase, is seldom repeated, and

appears to be strikingly uniform.

The energy distribution is restricted to such a low

frequency that the lowest distribution of energy on the sono-

gram is difficult to read, being so close to the bottom mar-

gin of the sonogram itself. This lower limit of the energy

range has, therefore, been approximated at 0.01 kHz. The

upper limit of energy distribution is 0.40 kHz (N=2).



Number (N) Mean (x) Range (R)

Units/Phrase 2/2 1.00 1.00-1.00

Phrase Duration 2 0.20 0.20-0.21

Lower Limit of
Energy (kHz) 2 0.01 0.01-0.01

Upper Limit of
Energy (kHz) 2 0.40 0.40-0.40

The individual giving the call was never seen, nor

did the animal sound as if it were very close to the observer.

However, among those species (Cercopithecus mona, C. campbelli,

and C. Dogonias) which Struhsaker (1970) studied that had

boom calls, only adult males gave the call.

Apparently, the boom was involved in some way with

coordinating group movement as it was usually given while the

group was moving, or else the group would move soon there-

after. On several occasions the boom w''as heard following

agonistic encounters within the group and so may serve to

facilitate group cohesion by overriding the disruptive and

dispersive effects of antagonism among its members.


The squeal (Fig. 10 and Table 6) was a high-pitched,

uniform call. It is entirely tonal and occurs as a single-

unit phrase. On the sonogram the squeal appears as two

slightly undulating bands, one at approximately 3.00 4.00

kHz and the other at 6.00 7.50 kHz. The end of the call

lies on a slightly lower frequency level than the beginning.



Number (N) Mean (x) Range (R)

Units/Phrase 7/7 1.00 1.00-1.00

Phrase Duration 7 0.63 0.39-1.20

Lower Limit of
Energy (kHz) 7 3.20 2.70-3.50

Upper Limit of
Energy (kHz) 7 7.87 7.00-9.00

The mean duration of the squeal is 0.63 seconds (N=7).

The mean lower frequency is 3.20 ikz and the mean upper fre-

quency is 7.87 kHz (N=7).


The shriek (Fig. 11 and Table 7) was a high-pitched,

shrill-sounding call of variable duration. Certain sections

of the call vary in degree of tonality and duration. Since

it appears as a continuous tracing on the sonogram it must be

classed as a compound, single-unit phrase.

The first part of the shriek resembles the squeal and

is completely tonal with a two-band appearance on the sonogram.

However, it is somewhat higher in pitch than is the squeal,

with a mean frequency of 3.46 8.74 kHz (N=5). This first

part continues into, and is superimposed by, a mixed tonal-

nontonal portion which appears as a noisier, broader, double

band with a mean frequency of 3.20 8.64 kHz (N=5). As the

call progresses, this portion may become noisier and its

appearance will become more distorted on the sonogram.

The mean duration of the tonal component is 0.36
seconds (N=5), while that of the mixed portion is 0.65 seconds

(N=5) for a mean total phrase length of 1.01 seconds.


On one occasion a call was given which can perhaps

best be described as sounding like "chut" (Figs. 12&13 and

Table 8). There were three such chuts or units grouped to-

gether to form a phrase and two phrases were given. There is

some tonal quality to the call but this is overlapped for the

most part by a nontonal component (thus forming a mixed unit)

giving the call a noisy appearance on the sonogram.



Number (N) Mean (x) Range (R)

Units/Phrase 5/5 1.00 1.00-1.00

Tonal Portion

Duration (sees.) 5 0.36 0.13-0.55

Lower Limit of
Energy (kHz) 5 3.46 3.10-4.00

Upper Limit of
Energy (kHz) 5 8.47 7.50-9.50

Mixed Portion

Duration secss.) 5 0.65 0.27-0.99

Lower Limit of
Major Energy 5 3.20 3.00-3.50

Upper Limit of
Major Energy 5 8.64 8.00-9.00
(kHi z)

The mean unit duration was 0.08 seconds (N=6) and the

average distribution of major energy was 2.95 4.53 hHz (N=6).

However, of the three units the third was the longest (0.12

seconds) and had a greater frequency range. The upper limit

of the total energy range of this third unit was approximately

13.0 kHz. On the sonogram the first unit appears as a single

band while the second and third units consist of two bands.

However, the second of these bands is difficult to discern

and measure since it is considerably obscured by the noise

component. Although there are some spectrographic similari-

ties with the chutter, the calls differ enough acoustically

to warrant their separate classification at this time.



Number (N) M[ean (x) Range (R)

Units/Phrase 6/2 3.00 3.00-3.00

Unit Duration 6 0.08 0.03-0.13

Interval Between
Units secss.) 4 0.12 0.05-0.20

Phrase Duration 2 0.50 0.49-0.50

Lower Limit of
Major Energy 6 2.95 2.50-3.50

Upper Limit of
Major Energy 6 4.53 3.20-5.50

The chut was given by either a subadult oc small

adult, possibly female. Two such animals were grooming high

up in a tree. Upon seeing the observer, the one being groomed

gave a squeal which was followed by the chut phrases. It then

moved away from the groomer who chirped when it noticed the

observer. Taking into consideration the context in which it

was given and the call with which it was associated, the chut

could be tentatively identified as a fear-alarm call.


The chutter (Fig. 14 and Table 9) is a completely

nontonal phrase composed of several units of short duration.

Each unit on the sonogram resembles a column of noise with

some broader than others. There is some variation in chutter

phrase duration, depending upon the number of units included

in the call. One of the three calls sampled contained twelve

units, the others each had four. The significance of phrase

length differences is not yet understood, but perhaps there

is some relationship between the intensity of stimulus

arousal and the number of units per phrase. It is also pos-

sible that the variation reflects individual differences.

Perhaps more than one type of chutter is represented in the

sample. Struhsaker (1967) was able to discern several dif-

ferent kinds of chutters for the vervet monkey. Without a

larger sample size and more information as to the function

cf the cutter calls and the circumstances which evoke them,

any statement as to a possible typology of cutter phrases

woulc be premature.

The unit duration and the interval between units was

similar among the three phrases. However, it was noticed

that the units of a phrase tend to increase in duration and

intensity such that those units late in the phrase are the

longest and loudest. The major energy distribution ranged

from 2.00 11.00 kHz with the mean distribution as 2.67 -

7,83 kHz (N=18).



Number (N) Mean (x) Range (R)

Units/Phrase 20/3 6.33 4.00-12.00

Unit Duration 18 0.06 0.01-0.11

Interval Between
Units secss.) 15 0.04 0.01-0.08

Phrase Duration 3 0.58 0.39-0.95

Lower Limit of
Major Energy 18 2.67 2.00-4.00

Upper Limit of
Major Energy 18 7.83 4.00-11.00

The individual giving the chutter call was never seen
and the situations in which they were given were seldom ap-

parent. They were sometimes, but not always, heard during

intragroup fights so apparently the cutters play some part

in agonistic encounters. Struhsaker (1969) likewise retorts

having heard cutters during the fight episodes of other

Cerconithecus species.

Low, Short Chutter

On one occasion a low chutter of short duration

(Fig. 15 and Table 10) was given by an adult Sykes. Two

phrases were recorded, each consisting of two units. This

chutter is lower in frequency than the chutter described

above, with a mean major energy distribution of 0.75 2.88

kHz (N=4). However, in terms of total energy distribution,

there is some indication on the sonograms of energy in the

vicinity of 7.50 kHz.



Number (N) Mean (x) Range (R)






Unit Duration

Interval Between
Units secss.)

Phrase Duration

Lower Limit of
Major Energy

Upper Limrit of
Major Energy
(kH 2)














Energy distribution and duration of each of the units

are auite similar, with the units appearing on the sonograms

as thin columns of noise. Mean unit duration is 0.04 seconds

(N=4) and phrase duration averages 0.14 seconds (N=2). This

chutter, like the previous one, is entirely nontonal.


The chutter-squeal (Figs. 16-18 and Table 11) can

best be described as a multi-phrase bout of long duration.

The following discussion is based on measurements of two

such calls which were both recorded during the same session.

Since the vocalizing individual was not seen, it is possible

that both these calls were given by the same animal.

The first phrase consists of a series of short units

resembling in some ways both the chutter and the chirp.

These units are basically nontonal like the chutter but yet

there is a hint of a chevron-shaped tonal component (as in

the chirp) superimposed over the nontonal element. The

first bout had a phrase consisting of nine chutter units and

the second bout had four chutter units. With a mean unit

duration of 0.07 seconds (N=13), the units are as short as

those of the chutter but do not have as wide a frequency


Shortly following this first phrase are single-unit

phrases which appear on the sonogram as nontonal-tonal undu-

lating bands. There were five of these phrases in the first

bout and two in the second. These phrases are of long dura-

tion; two sampled had a mean duration of 3.93 seconds. The

mean of the lower frequency of major energy distribution is

3.75 kHz while the mean upper frequency is 7.00 kHz.

The final phrase of the bout is also a single, mixed,

tonal-nontonal unit but the nontonality or noise is not as

pronounced as in the preceding phrases. The average dura-

tion of this unit is 0.62 seconds (N=2) and the mean fre-

quency range is 3.25 5.65 kHz (N=2).



Number (N) Mean (x) Range (R)

Chutter Portion


Unit Duration

Interval Between
Units secss.)

Phrase Duration

Lower Limit of
Major Energy

Upper Limit of
Major Energy

Mixed Portion I

Interval Between
Chutter & Mixed I


Phrase Duration



























TABLE 11 (Continued)

Number (N) Mean (R) Range (R)

Lower Limit of
Major Energy

Upper Limit of
Major Energy

Mixed Portion II

Interval Between
Mixed I & Mixed II


Phrase Duration

Lower Limit of
Major Energy

Upper Limit of
Major Energy

Total Duration

























Neither the final nor preceding phrases show any

great similarity, audibly or spectrographically, to the shriek

or squeal phrases reported earlier.

The sum of the duration of each of the phrases as

well as the intervals between them results in a lengthy call

averaging approximately 5.35 seconds (N=2).


The growl (Figs. 19&20 and Table 12) is predcminately

a tonal call although there may be some nontonality present,

particularly at the end of the vocalization. The growl con-

sists of a multi-unit phrase which may occur singly or may be




Number (N) Mean (x) Range (R)

Units/Phrase 135/17 7.94 3.00-20.00

Unit Duration 135 0.02 0.01-0.03

Interval Between
Units secss.) 118 0.02 0.01-0.03

Phrase Duration 17 0.20 0.05-0.59

Interval Between
Phrases secss,) 9 0.L9 0.08-0.67

Lower Limit of
Major Energy 17 0.52 0.25-0.75

Upper Limit of
Major Energy 17 1.15 1.00-2.00

The number of units per phrase was variable but aver-

aged 7.94 units per call (N=17). The unit duration and the

interval between units was similar in all the phrases sampled,

so phrase duration was largely dependent on the number of

units present. 'Since the number of units varied from three

to twenty, phrase duration varied considerably also. However,

the mean phrase duration was 0.20 seconds (N=17). Despite

their temporal variation, the calls are grouped into a single

category at this time on the basis of their acoustic and

spectrographic similarities.

Energy was distributed on approximately half the sono-

grams in a single band formation (N=8) and in a double (N=5)

or triple (N=4) band arrangement on the remainder. In the

last two cases, however, most of the energy would be concen-

trated in the lowermost band. The average distribution of

the heaviest energy concentration, therefore, was 0.52 1.15

kHz (N=17) but the total energy distribution might extend to

3.50 kHz in some cases.

Identification of the vocalizing individual was dif-

ficult but the call was apparently given by both juveniles

and adults. Examples of instances when the growl was heard

include: during chirping sequences by a Sykes threatening

the observer, as Sykes fed while looking at the observer,

during fight episodes, and while a troop was moving in the



A call somewhat similar acoustically to the growl

is the "err" (Fig. 21 and Table 13). On the sonogram, how-

ever, each call is broken into four or five short, discrete

units which extend to higher frequency levels than the growl.

The "err" is completely tonal.

The mean phrase duration of the "err" is 0.17 seconds

(N=3), with a mean energy range of 0.81 5.25 kHz for the

units (N=14).



Number (N) Mean (x) Range (R)

Units/Phrase 14/3 4.67 4.00-5.00

Unit Duration 14 0.02 0.01-0.05

Interval Between
Units secss.) 11 0.02 0.01-0.05

Phrase Duration 3 0.17 0.12-0.28

Lower Limit of
Energy (kHz) 14 0.81 0.50-1.20

Upper Limit of
Energy (kHz) 14 5.25 4.00-7.50

The individuals vocalizing were not identified and

the function of the "err" call is not understood. However,

the call was heard immediately following a shriek on two of

three occasions.


A short call best described as sounding like "uh"

(Fig. 22 and Table 14) was given before a shriek three times

out of a sample of five. The call was given in pairs or in

threes with a mean interval between the units of 0.12 seconds

(N=7). The mean unit duration was very short at 0.04 seconds




Number (N) Mean (x)


Unit Duration

Interval Between
Units secss.)

Phrase Duration

Lower Limit of
Major Energy

Upper Limit of
Major Energy













Range (R)







The "uh" is lower in frequency than the shriek which

sometimes followed (three out of five times). The mean lower

frequency of major energy is 1.78 kHz and the mean upper fre-

quency is 5.56 kHz (N=12). The call is entirely nontonal and

appears as a narrow column of noise on the sonogram.

Discrete and Graded Signals

In a study such as this in which the vocal reper-

toire of a primate species is being classified and described,

it is customary to examine the nature of the signals as to

whether they are discrete or graded. Discrete signals are

separate and distinguishable whereas graded signals are more

variable and can be arranged to form a continuum with other

such signals.

Marler has pointed out that continuous or graded

sounds are more common in rhesus monkeys, baboons, chimpan-

zees, and gorillas, resulting in sound systems which are

"much more complex in the development of these subtle contin-

uous and multidimensional variations of the sound signals"

(1965:566). The graded nature of the vocalizations of these

terrestrial or semi-terrestrial species which are relatively

isolated from other primate species of similar size and struc-

ture contrasts significantly, says Marler, with the relatively

discrete nature of the sound systems of forest-living primates

in frequent contact with other primate species, such as is

the case with Cercooithecus and Colobus.

In such circumstances the possibility of inter-
specific confusion is greatly increased and, cor-
respondingly, the limited evidence available on
the sound signals of such forms suggests that they
are more highly structured and that this structuring
is used at least in part to maintain species speci-
ficity in a major proportion of the vocal reper-
toire. (Marler, 1965:565)

The vocal repertoire of the Sykes monkeys seems to

comply with Marler's thesis. In the study area there were

three other diurnal species of nonhuman primates observed,

the baboon (Papio cynocephalus), the colobus (Colobus poly-

komos), and the vervet (Cercooithecus aethiops). The Sykes

were at times observed in close proximity to all three

species. Of these species, the vervet is the most closely

related to the Sykes, morphologically and behaviorally.

According to Struhsaker (1967), the majority of the vervet

vocalizations which he studied were of the discrete type.

Similarly, most of the Sykes' calls, w-hile showing some de-

gree of variability within categories, are also of the dis-

crete type. Signals which may prove to be of the continuous

type would include the shriek and the squeal, and the calls

of the chutter group.

For example, the squeal was described as a tonal call

consisting of two parallel bands with a mean energy distribu-

tion of 3.20 7.87 kHz (N=7). The shriek, on the other hand,

was composed of two components, a two-band tonal section fol-

lowed by a mixed portion of tonal and nontonal sound. The

initial tonal component strongly resembles the squeal, but

the mean energy distribution is slightly higher at 3.46 -

8.74 kHz (N=5). However, the ranges for the energy values

for the two calls do overlap (Table 6 and Table 7). It would

seem likely, therefore, that there may be a graded continuum

between squeals and shrieks.

There may also be a continuum within the chutter

group, although the sample is small and hence not extremely

reliable. Perhaps the chut, the chutter, and the short chut-

ter are but points along a graded continuum of such calls.

Also, there may be some functional significance in the number

of units per chutter phrase, creating a gradient in this re-

gard as well.

Future fieldwork and further study may point out

more continuous signals for the Cercoithecus mitis group as

well as modify those tentatively identified here.

This investigation has demonstrated the possible

existence of a graded series of signals or messages. That

this is not sufficient evidence to indicate a graded commun-

ication system has been aptly pointed out by Altmann (1967).

He stresses that signals used in graded or analogical com-

munication must be differentiated from those signals which

may appear graded but merely reflect the normal range of vari-

ation which a sample of calls may be expected to show. Such

variation would be due to the morphological and physiological

variations of the structures involved in sound production.

It is necessary, says Altmann, to demonstrate functional con-

tinuity: "What is needed is (1) a continuous array of mes-

sages, (2) a continuous array of responses to these messages,

and (3) a one-to-one mapping between messages and responses"


The failure to make the distinction between graded

signals which are associated with graded responses and those

which are not has resulted in some confusion and misunder-

standing concerning the use of the terms 'discrete' and

'graded'. The terms were introduced into the study of pri-

mate communication largely through the work of Hockett (1960)

who proposed thirteen design features, of which discreteness

was one, which he felt were characteristic of human lang-

uages. Apparently in their haste to look for Hockett's de-

sign features among the communication systems of nonhuman

primates, many researchers failed to recognize that discrete-

ness and gradedness among the response patterns as well as

discreteness and gradedness among the signals themselves must

be demonstrated in order for a communication system to be

classified as one or the other. Consequently, although these

terms occur frequently in the literature, Altmann (1967)

points out that a graded communication system has never been

demonstrated. Even if such a system were described, Altmann

further emphasizes that discontinuities between the major

categories of communication patterns themselves will make for

a certain amount of discreteness. The classification of the

system as one or the other, therefore, largely becomes a

question of semantics.

It would seem that an understanding of the nature of

communication systems among nonhuman primates lies not in its

comparison, feature for feature, with human language but

rather with the understanding of the adaptive significance of

such features as may be present. That the latter has not yet

been accomplished is due in large part to the fact that pri-

mate communication studies are still in their infancy. Hope-

fully, however, through future field studies and the amassing

of greater amounts of information, certain adaptive trends

and correlations will become apparent which will have signi-

ficance for understanding not only the nature of nonhuman

primate communication but the nature of human language as


A Structural-Functional Analysis

There are many different types of vocalizations among

the primates and these vocalizations serve various functions.

However, the exact nature of the relationship between func-

tion and structure of primate calls has not yet been deter-

mined. Although work along these lines has been carried out

for some insect species (Alexander, 1960, 1968), anuran

species (Bogert, 1960; Littlejohn, 1959), and avian species

(Marler, 1955, 1959; Marler and Hamilton, 1966), most discus-

sions of primate vocalizations in structural-functional terms

have remained generalized.

There are several reasons which may account for these

generalized studies. Firstly, the study of primate communica-

tion is still a relatively new endeavor and the emphasis,

until recently, has been on collection and description rather

than on detailed investigative analysis. Secondly, primate

communication has been shown to be an exceedingly complex pro-

cess involving several channels of communication; namely,

visual, auditory, tactile, and olfactory. The relative con-

tribution which each of these channels makes to a species'

communication system, the degree to which they are interrelated,

and the extent to which they limit and modify communication by

other channels are difficult to assess. Therefore, a discus-

sion of the structure of a particular vocalization must also

account for the possibility of simultaneous transmission of

information by some other means, such as visual signals.

Thirdly, the determination of the function of a call is often

difficult, requiring an evaluation of the behavioral and en-

vironmental contexts in which the call was given as well as

a description of the nature of the responses which it evoked.

Consequently, among those primate species which have been

studied, there remain numerous calls whose functions are un-

known. Lastly, it has not yet been determined which aspects

of a primate vocalization, if any, are most significant for

the transmission of information. That some components of a

call, such as frequency, intensity, duration, and timing,

are differentially informative has been shown to be the case

among arthropods (Alexander, 1968), anurans (Bogert, 1960),

and birds (Falls, 1963; Marler, 1955, 1959). Experimental

designs, such as play-back techniques both in the laboratory

and in the field, will probably be of considerable impor-

tance in evaluating the communicative significance of the

various components.

The present investigation will attempt to apply a

structural-functional approach to the study of Sykes vocaliza-

tions by focusing on several principles of sound localization

as outlined by Marler (1955, 1959) in his studies on bird

vocalizations. Structural comparisons will be made between

those Sykes calls used to communicate over long distances

and those which operate at close range. The discussion will

be limited to those calls whose functions are known.

As Marler (1955, 1959) points out, sound localization

in vertebrates relies on binaural comparisons of differences

in phase, intensity, and time of arrival. Intensity differ-

ences become apparent when the head acts as an obstruction,

thus hindering the sound from reaching one ear. The direc-

tion of the sound can be inferred due to the resulting dif-

ference in intensity at the two ears. Since this "shadowing

effect" is significant only when the obstruction is of the

same order of size as the sound wavelength or larger, locali-

zation by intensity differences is most useful with high fre-

quency sounds (Marler, 1959).

On the other hand, localization by phase differences

is confined to low-pitched sounds (Marler, 1959). Phase dif-

ferences are the result of pressure changes occurring at

different moments in the two ears. In high frequency sounds

where the wavelength is less than the distance between the

ears several wave oscillations may occur, resulting in ambig-

uous phase comparisons. In addition, the nerve leading from

the ear is unresponsive for about a thousandth of a second

after each oscillation, so a sound with more than a thousand

vibrations per second would not be completely represented by

nerve impulses (Marler, 1959).

Accurate comparisons of differences in time of arri-

val of sound at the two ears is dependent on the way a sound

begins and ends rather than on frequency ranges. It has been

found that sharp discontinuities are most easily perceived,

especially if frequently repeated (PMarler, 1959).

Marler concludes that the ideal sound permitting

source localization "will include a high pitch for location

by intensity difference, a low pitch for location by phase

difference, and it will be sharply broken and repetitive for

location by time difference" (1959:175). In man it has been

found that pure tones in the region of 2.0 5.0 kHz are dif-

ficult to locate by comparisons of phase and intensity differ-

ences, being too high for the former and too low for the

latter to be effective (Marler, 1955). For this frequency

range only time differences can be used to successfully locate

sounds. The lower limit of this frequency range is probably

similar in the Sykes monkey due to the refractory period of

the auditory nerve. However, the upper limit is probably

higher (exactly how much so is yet to be determined) since

the diameter of the Sykes' head is smaller than man's and

the "shadowing effect" will thus not take place until a smaller

wavelength or higher frequency is reached.

Five Sykes calls, about whose functions something is

known, were chosen for investigation. These vocalizations

will be examined in terms of whether they can or can not be

easily localized by phase, intensity, or time differences.

The five calls are the nyah, boom, chirp, squeal, and shriek.

It has been mentioned elsewhere that the nyah call

given by adult males functions to promote group cohesion and

coordinate group movements. Struhsaker (1970) emphasizes

that by doing so, such calls may also serve to maintain re-

productive isolation in areas of mixed species associations.

When the call is given the group members, who have been dis-

persed at the time, begin to move in the direction from which

the call came. The call is loud and carries over considerable

distances. Visual contact does not seem to be of importance

here so one would expect the nyah, in keeping with its func-

tion, to be a call which is easily localized. Analysis of its

structure shows 7his to be the case. The average range of

major energy distribution for the nyah was shown to be between

0.82 and 5.46 kHz (Table 3) which would apparently include a

low enough pitch for localization by phase differences. The

upper limit of this range would appear at first glance to be

too low to facilitate localization by intensity differences

since it most likely falls within the "hard-to-locate" range

discussed earlier. However, since in total energy distribu-

tion the upper frequencies of the nyah may extend to 16.0 kHz

or more, localization by means of intensity differences is

indeed possible. The call is of a short duration (0.31 sec-

onds), has an abrupt beginning, and is almost always repeated

thus enabling localization by temporal differences. In short,

the nyah call fulfills all three of Marler's requirements for

an ideal sound permitting source localization. Indeed, this

is what one would expect of a call which apparently plays

such a vital role in group maintenance and cohesion.

Another call whose function is apparently concerned

with group cohesion without requiring visual contact with the

caller is the boom vocalization. This call, however, differs

from the nyah in several respects. The boom is a very low-

pitched call with an average energy distribution of 0.01 -

0.40 kHz (Table 5). This low frequency would permit locali-

zation by phase differences but not by intensity differences.

Although the call is of a short duration (0.20 seconds), it

is not repeated, nor does it begin or end abruptly, thus

making localization by temporal differences difficult. There

are several possibilities which may account for the structure

of this call. Although the boom would not appear as ideal

for localizing a sound source as the nyah, perhaps it is given

under circumstances where an extreme degree of localization

would be disadvantageous to the caller or the group, such as

in reaction to the observer or a potential predator. Another

possible explanation is that perhaps phase difference alone

is an efficient means of sound localization as long as the

call is restricted to very low frequencies. In addition, a

low-pitched group cohesion call might be used because low

frequency sounds carry over greater distances than sounds of

higher frequencies (Moynihan, 1967). Also, there is less

chance that a low-pitched call will be refracted by trees or

other objects as is the case with high-pitched sounds (rMarler,

1955), which is obviously an important consideration for an

arboreal species. The key to explaining the nature of the

boom clearly lies with a better understanding of its function

as well as with the formulation of an experimental approach

which would test the localization properties of the call.

The chirp acts as a threat/alarm call and its phys-

ical properties would seem to facilitate localization. Al-

though the average major energy distribution of 2.91 6.14

kHz (Table 1) falls within the "hard-to-locate" zone, the

total energy range of 0.10 16.0+ kHz most likely permits

localization by means of phase and intensity differences. The

call has an abrupt beginning, is of short duration (0.18 sec-

onds), and is usually repeated. With increasing intensity or

levels of excitement, the call is repeated more frequently

and may have an even more abrupt beginning. Localization by

comparison of temporal differences is clearly possible and

may be of considerable importance. One might expect an alarm

call to be given in such a way as to prohibit localization of

its source for safety's sake, as is the case for some avian

calls (izarler, 1955). That this is not the case for the chirp

is supported by the observation that the other group members

look toward the caller upon hearing the chirp and may then

chirp also. In addition, the caller is frequently exposed

in the open when giving the call and may not move back into

the brush for some time; in fact, other Sykes may actually

move up to be near the caller as they join in chirping.

Since the chirp is easily located it aids in indicating the

source of danger to other group members and in enlisting their

support in threat.

The last two calls to be considered are the squeal

and shriek. Both seem to be associated with agonistic encoun-

ters although their exact function is undetermined. Their

energy distributions of 3.30 7.87 kHz for the squeal (Table 6)

and 3.20 8.74 kHz for the shriek (Table 7) are apparently

within the "hard-to-locate" range. The calls are of a longer

duration than the others which have been discussed, 0.63

seconds (squeal) and 1.01 seconds (shriek), and both start

and end gradually. Precise localization of the sound source

would appear to be difficult. However, the most important

difference between the shriek and saueal and the other calls

which have been described is that the functions of the laCzer

require transmission over distances, whereas the shriek and

squeal apparently function in close-range a3onistic encounters

Consequently, other channels of communication, such as facial

expressions or touch, may contribute to the message content

and indicate the location of the vocalizer.

It is difficult to explain the physical properties

of the remaining calls in the Sykes' repertoire until more

is known of their function. Since most of these calls seem

to function on a close-range basis, one would expect other

channels of communication to be of some importance in signal


It should be emphasized that a discussion of the

structural and functional components of a call in terms of

its localization properties is an oversimplification of the

nature of primate vocalizations. Other factors, such as re-

quirements for species specificity and individual identifi-

cation or the nature of the physical and social environments,

may affect the physical structure of a primate call. For

example, Marler (1965) points out that in those calls used

over distances to maintain spacing between groups, the re-

quirements of species specificity necessitates use of a

larger range of sound properties than in close-range commun-

ication. These spacing calls, says Marler, are therefore

more highly structured and purer in tone. It is possible,

therefore, that requirements for species specificity might

be partly responsible for the highly structured nyah call of

the Sykes monkey.

Oversimplified generalizations about primate vocali-

zations can be dangerous and misleading. A statement that

arboreal species living under conditions of low visibility

have developed elaborate high-pitched calls (Cartlan and

Brain, 1968) gives the misleading impression that a high

pitch alone makes such calls functionally significant and

fails to recognize and account for the existence of low-

pitched calls among such species. What is needed instead is

a structural-functional analysis to furnish clues as to why

a particular structure or form may predominate. It is hoped

that the analysis presented in this section has demonstrated

the feasibility of such an approach.

Taxonomic Imolications

The genus CercoDithecus is said to have more species

and forms than any other primate genus in Africa (Struhsaker,

1970). Its species have usually been distinguished taxonomi-

cally on the basis of pelage coloration. Unfortunately, wide

variability in pelage coloration within the genus and within

its species has resulted in some disagreement over the inter-

pretation of phylogenetic affinities. In cases where morpho-

logical characters are inadequate or equivocal, the taxono-

mist has been urged to make use of data from studies in be-

havior, ecology, physiology, and biochemistry (Lanyon, 1969).

Through the use of improved recording equipment and

the development of the sound spectrograph, one aspect of be-

havior, i.e., vocal communication, has been shown to be of

some significance in taxonomic studies of orthopteran insects

(Alexander, 1962), anuran amphibians (Bogert, 1960; Blair,

1962, 1964), and birds (Marler, 1957; Lohrl, 1963). Studies

such as that of Struhsaker (1970) have indicated that vocal-

izations, when used with the same precautions and care appli-

cable to any potential taxonomic character, can be of consid-

erable importance in primate taxonomy as well, Struhsaker

examined the taxonomic position of several Cercooithecus

species, basing his decisions largely on similarities or dif-

ferences in certain vocalizations.

One of the species studied by Struhsaker (1970) was

Cercopithecus nictitans. Several authors have pointed out

the strong physical resemblance between C. nictitans and

C. mitis (Booth, 1956; Haddow, 1952; Kingdon, 1971; and Tap-

pen, 1960). Hill (1966) placed C. nictitans in the mitis

superspecies and the 1969 Wenner Gren Conference raised the

question that perhaps C. nictitans and C. mitis might be best

considered as conspecific (Thorington and Groves, 1970).

A comparison of the vocalizations of these two species might

shed some light on their phylogenetic relationship.

Of the two species, C. nictitans is the more restricted

in distribution, being largely confined to the West African

area. Two subspecies are generally recognized. C. n. martini

is the most westerly located (Ghana, Nigeria, Cameroun).

C. n. nictitans is also on the western coast (Gabon, Cameroun)

but extends eastwards into the Central African Republic area.

C. mitis, characterized by numerous subspecies, is much more

extensive in its range, being found in central, east, and

south Africa. C. mitis is said to be the only forest guenon

having such a wide distribution in southern and eastern

Africa (Kingdon, 1971). It apparently inhabited these areas

before climatic changes isolated many of the forests, thus

closing them to those Cerconithecus species which have devel-

oped more recently (Kingdon, 1971). In fact, C. mitis is

believed to be near the stock from which the main guenon ra-

diation has derived (Kingdon, 1971). For a more detailed

discussion of the distribution of the various C. mitis sub-

species, the reader is referred to Rahm (1970).

The subspecies which will serve as the basis for the

present discussion of vocalizations are C. n. martini and

C. n. nictitans studied by Struhsaker (1970) in Cameroun and

C. m. kibonotensis from the southeast Kenya coast whose vo-

calizations have been described in this paper.

The reader should note that the C. mitis and C. nic-

titans populations here being considered are allopatric, a

situation which can make for considerable difficulty in as-

sessing phylogenetic relationships (Mayr, 1969). In this

regard, it is the examination of sympatric species which is

most conclusive, since any mechanisms assuring species iden-

tity through reproductive isolation can be more reliably

recognized in the sympatric state. However, Struhsaker

(1970) has shown that interpopulation variability in the

basic structure of vocalizations within a species appears to

be minimal. The species he considered included C. aethiops

(populations from East and West Africa), C. mona, C. nicti-

tans (populations from both subspecies), and C. pogonias

(populations from the subspecies DoEonias and Eravi). Within

each of these species and across subspecies lines Struhsaker

fo:ond that the vocalizations snowed a considerable amount of

stability and uniformity. Therefore, it will be assumed that

the vocalizations of C. mitis are relatively uniform through-

out its range also. In this way, it will be assumed that the

vocalizations of C. mitis kibonotensis of East Africa would

be basically similar to those C. m. stuhlmanni and C. m.

maesi, two western subspecies of C. mitis which are reported

to be sympatric with C. n. nictitans at the eastern limits of

its range (Rahm, 1970; Kingdon, 1971). Ideally, a study to

determine whether C. nictitans and C. mitis are consDecific

should be carried out in this area of sympathy. Until such

studies are available, however, and until studies indicative

of the extent of variation of the vocalizations within C.

mitis are carried out, the investigation presented here will

proceed on the assumption which has been made based on Struh-

saker's analysis (1970). No attempt is being made, therefore,

to present the findings of this study as a final taxonomic


Hill (1966) apparently feels there is enough dis-

tinction between the western and eastern subspecies of the

C. mitis group to warrant their separation into two different

species, C. mitis and C. albosularis respectively. However,

other workers (Rahm, 1970; Kuhn, 1967; Kingdon, 1971) have

classified these as a single polytypic species. That hybrids

have been reported between C. m. stuhlmanni and C. m. albo-

gularis (Booth, 1968) not only tends to support the latter

approach but also adds more weight to the assumption that

isolating mechanisms, including any vocalizations which may

function as isolating mechanisms, are probably uniform

throughout the range of C. mitis. It should be pointed out

that since the eastern subspecies of C. mitis are relatively

isolated from other forest cuenons while those in the west

may be sympatric with several, such as, C. mona, C. ascanius,

C. l'hoesti, and C. neglectus, there may be selective pressures

operating in the west which may not be operating in the east.

These differences may be reflected in the vocalizations of

these sDecies.

Struhsaker (1970) believes that the loud calls given

by adult male Cercopithecus monkeys function to maintain re-

productive isolation between species by enhancing group

coherency, especially in response to alarming stimuli. In

the sympatric state, therefore, the loud calls of two species

would-be different; Struhsaker, working with several species,

found this to be the case.

A comparison of the loud calls which Struhsaker re-

ports for C. nictitans (1970:407, 427) with those reported

here for C. m. kibonotensis (Figs. 6-8 and Tables 3&4) re-

veals some differences. The nyahs of C. mitis cover a higher

frequency range than do the pows of C. nictitans and have a

more elaborate development of harmonics. The tonality of the

C. mitis calls also has an arching configuration, rising and

then falling in frequency, as opposed to the descending

quality of the tonality in C. nictitans. The short nyahs

have some spectrographic resemblance to the hacks of C. nic-

titans, but again a higher frequency is reached in C. mitis.

No calls equivalent to the pow-hacks which Struhsaker found

to be intermediate between pcws and hacks were found for

C. mitis. Struhsaker could not discern any significant dif-

ference in the loud calls of the two subspecies of C. nicti-

tans, nor could he distinguish the hacks of C. eryvhrotis

and C. cephus (1970:408, 429) which he subsequently placed

in a single species. The loud calls of C. mitis and C. nic-

titans are apparently not as similar as those of these other

conspecific populations.

In addition, booms were recorded for C. mitis but not

for C. nictitans and in this regard C. mitis resembles the

mona superspecies. The C. mitis booms (Fig. 9 and Table 5)

resemble the C. mona calls (Struhsaker, 1970:409, 431) very

closely. However, since this type of call is restricted to

such low frequencies, variability may be expected to be

limited so that similarities may not be significant. Booms

are typically given before a series of hacks in C. mona.

Booms were also frequently heard in conjunction with nyah

calls in C. mitis. However, these similarities in boom calls

should not be taken to indicate an especially close relation-

ship between C. mona and C. mitis. The hacks of C. mona

(1970:410, 423) do not appear to be any more similar to those

of C. mitis than the latter was to C. nictitans. In addition,

only hack-type units were recorded for C. mona whereas C. mitis

definitely shows two types, i.e., nyahs and short nyahs. No

calls similar to the "ooo" contact calls reported by Struh-

saker for the mona supersDecies were found in C. mitis.

There are also several differences between other calls of

C. mona and C. mitis and these will be discussed below.

The chirp (alarm) calls of C. nitis (Figs. 1-5 and

Tables I&2) bear some resemblance to the chirps of C. nicti-

tans and C. erythrotis-ceohus (Struhsaker, 1970:411, 44,.

435, 436). In terms of the spectrographic form of the chirps

the three species are very similar, but those of C. mitis

are of somewhat higher frequencies as well as of longer dura-

tion. Multi-unit chirps were present in all three species.

Struhsaker (1970) found the chirps of C. erythrotis and C.

cephus to be very similar. Some of these chirps were iden-

tical to those given by C. nictitans. Similarity in alarm

calls among sympatric species faced with common predators

may be explained as a case of parallelism. Struhsaker (1970)

points out, however, that C. mona and C. pogonias also asso-

ciate freely with C. nictitans and C. erythrotis-ceohus but

maintain very distinct alarm calls. Therefore, Struhsaker

concludes that C. nictitans and C. erythrotis-cephus may be

more closely related to each other than to C. mona.

Struhsaker (1970) reported grunts and chutters for

C. nictitans and C. erythrotis-cephus but found that no such

calls were given by members of the mona superspecies. The

grunts are thought to function in maintaining group cohesion

while the chutters were given during agonistic encounters.

C. mitis has vocalizations similar to these grunts and chut-

ters, although the samples for C. mitis are too small to be

conclusive. A low, short chutter given by C. mitis (Fig. 15

and 'able 10) resembles very closely the C. nictitans chut-

ter2 Jscri:'bed by Struhsaker ( 970:419, 1-41) both in energy

disT ribu tlon and duration. However, the relationship of this

cutter to others of higher frequencies which have been re-

ported for C. mitis in this paper is not clear. Struhsaker

does not mention any other chutter types for C. nictitans

and C. erythrotis-cephus but his 1970 study was apparently

not intended to be descriptive of the complete sound reper-

toires. Therefore, it remains to be seen whether the exis-

tence of these other chutter-types in C. mitis is indicative

of a significant species difference. The "err" reported here

for C. mitis (Fig. 21 and Table 13) may be equivalent to the

grunts; if so, some differences are apparent. While all the

C. nictitans grunts (Struhsaker, 1970:418, 441) were single-

unit phrases and the C. erythrotis grunts (Struhsaker, 1970:

418, 441) were either single- or triple-unit phrases, the

"err" of C. mitis was made up of four or five units. Although

unit and phrase duration for all three species are basically

similar, the C. mitis call extends to a much higher frequency.

Again, Struhsaker found that the grunts and chutters of C.

erythrotis and C. cephus were very similar to each other and

also to those of C. nictitans; the grunts, in fact, vere indis-


Of the vocalizations discussed above, the one which

plays the most important role in helping to maintain species

identity through reproductive isolation is the loud call.

Struhsaker points out that vocalizations having different

functions evolve at different rates: "Calls serving to alarm

or coordinate groups may evolve more slowly than those enhan-

cing reproductive isolation through the maintenance of conspe-

cific intragroup cohesion around one male" (1970:370). If

this assumption is correct, then one would expect the chirps,

grunts, and chutters to not be under the same pressures for

species specificity that influence the loud calls. Further-

more, species showing similarities in these calls may be more

closely related to each other than to species which do not

have such similar calls (Struhsaker, 1970).

It has been shown that the loud calls of C. mitis are

readily distinguishable spectrographically from those of C.

nictitans. Since these differences are greater than those

found among conspecifics, a separate specific status for C.

mitis and C. nictitans seems warranted at this time. This

conclusion is further supported by the observation that the

chirps, grunts, and chutters of C. mitis are also not as

similar to those of C. nictitans as are these calls within

the C. nictitans subspecies and within the C. erythrotis-

cephus species. According to Struhsaker (1970), the fact

that grunts, chutters, and chirps are absent in the mona su-

perspecies would seem to indicate that these calls are not

to be interpreted as conservative, widespread vocalizations

but rather may be indicative of close phylogenetic relation-

ships. Following Struhsaker's reasoning, one can conclude

that the similarities of the C. mitis calls to those of C.

nictitans and C. erythrotis-cephus, while not indicative of

conspecific status, would indicate that C. mitis, C. nicti-

tans, and C. erythrotis-ceohus are more closely related to

each other than to the mona superspecies.

It should be emphasized again that future studies

may reveal that the vocalize ions of the more westerly

distributed C. mitis populations may be considerably differ-

ent from those reported here for C. m. kibonotensis. Infor-

mation as to the amount of variation in vocalizations through-

out C. mitis' range is needed to either support or refute the

assumption made here concerning their supposed stability and

uniformity. Information is also needed as to the extent to

which C. mitis and C. nictitans may associate in their zones

of sympatry. If the two species should be found to hybridize

freely in these zones, then serious consideration of a con-

specific status would be called for.

Although the samples collected for C. mitis were often

too small to permit rigorous statistical comparison and test-

ing, it is recognized here that a true assessment of the de-

gree of relationship between two species must be based on

careful statistical measures before any great reliability in

the data may be assumed. It is also recognized that taxonomic

conclusions should not be based on one or a few characters

alone, but should depend instead on the total assessment of

morphology, physiology, behavior, karyotypes, etc. However,

if Struhsaker's observation is true that "of all the single

classes of characters so far used, vocalizations seem the most

stable and thus the single most reliable indicator of phylo-

genetic affinities within this genus" (1970:402), then fu-

ture studies of vocalizations of the Cerconithecus species

may very well prove to be of considerable importance in the

assessment of phylogenetic relationships within this highly

variable genus. The analysis of vocalizations, therefore,

should merit considerable attention by primate taxonomists.



In sharp contrast to the relatively vocal Sykes mon-

keys are the taciturn colobus monkeys (Colobus polykomos cal-

liatus). These beautiful, black and white primates sit for

hours at a time without making a sound.

Colobus live in small groups; those in the study area

ranged in size from three to eight animals per group. They

maintain a relatively small territory in comparison to the

Sykes. Each troop was observed to occupy the same general

area for the duration of the study and, as a result, they

could be located quite readily.

The taxonomic designation, C. Dolykomos, being used

here follows that of Schouteden (1947), Tappen (1960), Napier

and Napier (1967), and Kuhn (1967). However, others, i.e.,

Schwarz (1929), Dandelot (1968), and Rahm (1970), differen-

tiate C. angolensis from C. polykomos and under this system

the colobus of this study would fall within the former desig-

nation. It is generally agreed that before the taxonomic

Problems can be resolved more information is needed about the

adaptations of these monkeys and areas of distribution over-

lap need to be investigated (Tappen, 1960).

The colobus have onr habit which, while not vocal,

definitely serves as a means of communication. The colobus

take fantastic leaps, land with a resounding thump against

tree limbs and branches, and continue in this way jumping

from tree to tree. The other colobus are alerted (probably

by both the sight and the sound) and look in its direction,

watching the leaping colobus closely. These displays are

quite impressive and seemed to be given in response to the

observer's presence, probably functioning in part as an

attempt to intimidate the intruder.

This ability to take leaps of considerable distance

also serves to indicate danger and the direction of troop

movement. If a colobus is startled, it will vault itself

seemingly into midair only to come crashing down into the

thicker brush below. The other colobus are immediately alerted

and also leap dowm into the lower level. A similar observa-

tion has been made by Schenkel and Schenkel-Hulliger (1967).

In a short while, when the danger has apparently passed, the

colobus will gradually move back up into the higher branches.

Marler (1968) also emphasizes the reliance of the colobus on

visual cues for maintaining intragroup spacing during move-

ment, pointing out that the conspicuous color of the colobus

and their habit of keeping high in the trees facilitate vi-

sual coordination of their activities.

Colobus tend to remain silent unless sufficiently
disturbed. Such disturbing circumstances included the close

proximity of the observer, the milling around of a group of

four or five dogs in the woods directly below the colobus

troop, and the circling overhead of a pair of Bateleur Eagles

(Terathooius ecaudatus).

On the latter occasion, a group of eight colobus were

sitting relaxed in the tall trees. In this troop was a female

with a small white infant. The female had been sitting unper-

turbed with the infant squirming about in her lap but as a pair

of Bateleur Eagles slowly circled closer she became quite tense

and attentive, craning her neck to watch the birds. After a

few moments she clutched the infant to her and moved over to

a more thickly leaved tree. As she did so, a male colobus

gave a croaking, snorting sound and bounded over to the fe-

male's side, also watching the eagles intently. Moments later,

the male was observed holding the infant so he must have ta-

ken it from the female when he moved near her. However, inci-

dents such as this were infrequent and other than the vocal-

izations of infants or juveniles, the troop was generally


The white infant was a focal point of interest for

the entire group. In fact, all the adult females were so

interested in the infant that at times it was very difficult

to determine who the actual mother was. This was further

complicated by the frequent practice of passing the infant

from one adult to another when it would begin to squirm and


The colobus were frequently found in the same immed-

iate vicinity with the Sykes monkeys. The two species could

of-cen be seen feeding in the same trees or even on the same

branch just a few feet apart.

Classification of Calls


A loud, snort-like sound (Figs. 23&2L and Table 15)
was given by adult colobus on several occasions. The call is

most likely an alarm and/or threat call as it was evoked by

the sudden appearance of the observer, by a group of dogs,

and by the Bateleur Eagles. A similar call was reported for

C. polykomos by Hill and Booth who described it as an "explo-

sive snort, uttered partly through the nose" which was given

by adults of both sexes (1957:311).



Number (N) Mean (x) Range (R)

Units/Phrase 4/3 1.33 1.00-2.00

Unit Duration 4 0.13 0.06-0.15

Lower Limit of
Major Energy 4 0.25 0.25-0.25

Upper Limit of
Major Energy 4 10.45 3.30-13.00

The snort begins abruptly and appears primarily as a
nontonal vocalization on the sonogram. The call averages

approximately 0.13 seconds in length (N=4). The major energy

distribution is about 0.25 13.00 kHz with the greatest con-

centration of noise showing on the sonogram in the vicinity of

4.0 8.0 kHz. The snort usually consisted of a single-unit

phrase, although on one occasion a two-unit snort phrase was

given, the two units being separated by an interval of 0.L4

seconds. The second snort was approximately only half as long

as the first and its energy distribution was confined to a

lower frequency range (0.25 3.30 kHz).


The snort may stand by itself as a call or it may pre-

cede another call identified here as the snort-croak (Figs.

25-27 and Table 16). Marler (1970) mentions that C. gueraza

has a distinctive roaring call which sometimes may be preceded

by a snort. This may be equivalent to the snort-croak de-

scribed here, but this writer hesitates to use the term 'roar'

because such a term would seem to imply a single, long, loud

call which definitely wvas not the case in the snort-croak.

The snort-croak is probably equivalent to the "full call" re-

ported by Hill and Booth (1957) as given by adult male C.

polykomos. These authors also noted that the snort-like alarm

call may precede the full call.

A jumping-roaring display said to be highly ritualized

and which may function as a mechanism of intergroup spacing

has been described by Ullrich (1961) and Schenkel and Schenkel-

Hulliger (1967). The roaring is reported to spread from group

to grouD so that a chorus develops. Although the snort-croak

was never heard under these circumstances while this observer

was in the field, it is possible that it is equivalent to this

roaring display. Likewise, whether the "full call" described

oy Hill and Booth (1957) is the same as the roaring display

can not be determined from their report. Based on the ver-

bal descriptions of .he calls, they all seem to be similar;

confirmation of this must await spectrographic comparison.

The snort-croak is actually a long bout consisting

of three different phrase-types. Besides the initial single-

unit snort phrase, the snort-croak includes a multi-unit

phrase of a croaking, gutteral sound given in alternation

with a faint, short, inhalation, single-unit phrase.

The initial snort of the snort-croak is very similar

acoustically and spectrographically to the snort phrase which

has already been described. The phrase duration of the snorts

of the snort-croak averages 0.11 seconds and the major energy

is distributed over a range of 0.25 12.50 kHz (N=4).

Each croak phrase is composed of several short tonal

units. Those measured had anywhere from eight to sixteen

units (mean of 10.21 units; N=24). Each unit is of a very

short duration (0.01 or 0.02 seconds) but together as a phrase

the mean duration is 0.24 seconds (N=24). The croak is low

in frequency and has a deep, gutteral sound. The average dis-

tribution of the heaviest concentration of energy is from

0.26 0.97 kHz. The number of croak phrases per bout is

variable, averaging 5.38 phrases per bout (N=13) but ranging

from two to seventeen. It should be pointed out that the

seventeen croak phrases represent an extreme; the next highest

number of phrases is nine. The bout with seventeen phrases

was given during the encounter with the eagles and was given

more rapidly than the others. In addition, the croak phrases

were shorter, averaging 0.16 seconds as compared to an average

of 0.34 seconds (N=11) for the other croak phrases. A cor-

tion of this seventeen-phrase bout is pictured in Figure 27.

In between the croak phrases is a faint sound made as

the colobus inhales. This unit is nontonal and averages 0.05

seconds in duration (N=22). The average major energy distri-

bution of this single-unit phrase is 0.27 1.25 kHz (N=22).



Number (N) Mean (x) Range (R)

Snort Phrase


Phrase Duration

Lower Limit of
Major Energy

Upper Limit of
Major Energy

Croak Phrase


Unit Duration

Interval Between
Units secss.)

Phrase Duration

Lower Limit of
Energy (kHz)

Uncer Limit of
Energy (kHz)



























0.97 0.75-1.00

TABLE 16 (Continued)

Number (N) Mean (x) Range (R)

Inhalation Phrase

Units/Phrase 22/22 1.00 1.00-1.00

Phrase Duration 22 0.05 0.04-0.07

Lower Limit of
Major Energy 22 0.27 0.25-0.40

Upper Limit of
Major Energy 22 1.25 0.70-2.00

Gecker White Infant

The gecker (Fig. 28 and Table 17) is a high-pitched,

squeaky sound and was given quite frequently by a white in-

fant colobus. Usually the infant would be squirming and

wriggling around while giving the call but it was impossible

to tell if it was trying to nurse. The adult holding the

infant would either shift it in its arms or else let another

colobus take it.

The frequent, at times almost constant, geckering by

the infant was in sharp contrast to the usual nonvocal atti-

tude of the rest of the troop. As a result, the somewhat

noisy infant seemed to be able to solicit attention and be an

important focal point for the rest of the group.

Energy distribution of the gecker ranged from 0.25

to 13.00 kHz, but on the sonogram the darkest portion averaged

3.64 6.69 kHz (N=11). The mean phrase duration was 0.32

seconds (N=11) and several such single-unit phrases were com-

bined to give bouts of variable length. It is difficult to

assess the duration of a bout or the number of phrases per

bout since the call varies throughout in loudness and gives

an impression of randomness and irregularity rather than of

a fixed pattern of phrase sequences.



Number (N) Mean (x) Range (R)

Units/Phrase 11/11 1.00 1.00-1.00

Phrase Duration 11 0.32 0.27-0.40

Interval Between
Phrases secss.) 0 0.13 0.05-0.28

Lower Limit of
Major Energy 11 3.64 2.00-4.50

Upper Limit of
Major Energy 11 6.69 6.10-7.70

The gecker is both tonal and nontonal, with the tonal

section first forming an arch-like pattern on the sonogram

which then levels out into the noise component. As mentioned

above, one of these "arches" is considerably darker than the

rest; the others occur above and below it in a harmonic


Err White Infant

On one occasion a phrase consisting of three "err"
sounds was heard at the end of a geckering sequence (Fig. 29

and Table 18). The exact relationship of this call to the

gecker is not understood; it may be a part of the gecker or

it may be an individual vocalization. However, on the sono-

gram the call is different in appearance from the gecker.



Number (N) Mean (x) Range (R)

Units/Phrase 3/1 3.00 3.00-3.00

Unit Duration 3 0.11 0.03-0.15

Interval Between
Units secss.) 2 0.05 0.05-0.05

Phrase Duration 1 0.35 0.35-0.35

Lower Limit of
Energy (kHz) 3 0.63 0.50-0.90

Upper Limit of
Energy (kHz) 3 2.13 1.80-2.50

The mean duration of the three tonal units was short
at 0.11 seconds (N=3). The "err" was lower in pitch than the

gecker, with a mean energy range from 0.63 to 2.13 kHz (N=3).

The duration of the phrase as a whole was 0.35 seconds.

Gecker Juvenile

A call similar acoustically and spectrographically to

the gecker of the white infant was given by a juvenile (Fig.

30 and Table 19). This individual had already attained its

black and white coloration and, except for its smaller size,

it was indistinguishable from the adult colobus. The juven-

ile no longer seemed to be very dependent on its mother al-

though it was carried occasionally, apparently when frightened

by the observer.

The juvenile gecker is also composed of tonal arch-

like bands passing into nontonal components. Harmonics are

present but the bands are closer together and more numerous

than in the infant gecker.



Number (N) Mean (x) Range (R)

Units/Phrase 13/13 1.00 1.00-1.00

Phrase Duration 13 0.11 0.04-0.14

Interval Between
Phrases secss.) 11 0.02 0.01-0.05

Lower Limit of
Major Energy 13 2.60 1.90-3.80
(kH z)

Upper Limit of
Major Energy 13 8.05 6.00-10.10

The individual single-unit phrases were of shorter

duration than those given by the white infant, with an aver-

age length of 0.11 seconds (N=13). The average energy dis-

tribution ranged from 2.60 to 8.05 kHz (N=13), but again one

band was somewhat darker than the rest at approximately 4.0 -

5.0 kHz. As in the infant gecker, it is very difficult to

judge the duration of a bout or to determine the number of

phrases per bout.

Occurring in the geckering bout on one occasion was

a phrase which appeared on the sonogram as a single, wavy,

linear unit, 0.54 seconds in duration (Fig. 31). The unit

is basically tonal and occupies a frequency range of approx-

imately 2.90 6.00 kHz. The darkest portion was again in the

vicinity of 4.0 5.0 kHz so perhaps this phrase is but a

series of geckering phrases given close together and lacking

harmonic development.

The gecker was given by the juvenile as it was startled

by the observer and ran to an adult female. Upon being em-

braced by the female, the juvenile stopped vocalizing.



Vocalizations and Crop-Raiding

The information collected on the baboon vocalizations

formed part of a larger study on baboon crop-raiding behavior.

Unfortunately, conditions were somewhat less than ideal at the

Shimoni site during the time this observer was in the field.

This site promised to be the best suited for making recordings

and observing the baboons' behavior. However, due to seasonal

conditions and a long drought in the area, no significant

crops were being grown other than a few remnants gone to seed

from the last growing season. In addition, the Africans who

usually sit up in observation platforms in order to spot ba-

boons and chase them from the field were not present.

When in the area a few years earlier, Maples (per-

sonal communication) noted what, superficially at least,

appeared to be diversionary behavior. It seemed that one

group of baboons would vocalize and thus attract and keep

the attention of the shamba, or farm, guards while the rest

of the troop would silently move into the farm a short dis-

tance away and raid the crops growing there. If this behav-

ior were indeed coordinated and intentional, it would shed

an interesting and new light on the nature of communication


systems among nonhuman primates. It wa's the purpose of this

part of the study to investigate and record those vocaliza-

tions given in conjunction with this behavior.

Although the normal raiding conditions were absent

during the course of this observer's investigations, the

baboons did come to the shamba practically every day to for-

age for whatever remains could be found in the field. As a

result, enough information was collected to allow at least a

tentative explanation as to the nature of this seemingly de-

liberate "decoy" behavior.

Although shamba guards were lacking and consequently

the raiding situation not as tense as it might have been under

normal circumstances, it was noted that the baboons, in reac-

tion to an observer moving about rapidly in the field, would

move into the forest edge and gradually begin to move silently

around the perimeter of the farm. The baboons at the rear of

this troop movement would be up in the trees giving the

threat/alarm bark. It is to these vocalizing baboons that

the observer's attention is drawn. Consequently, he remains

nearer to this group of baboons and, presumably, the shamba

guards would be inclined to do the same. The vocalizing

animals tend to minimize their own exposure to danger by

sounding the alarm when the human intruder is moving away or

looking in another direction, a practical behavioral adap-

tation since shamba guards may be armed with rocks, sticks,

or bows and arrows. With the humans' attention thus drawn

to this vocalizing group, the rest of the troop is relatively

free of human intervention and has the opportunity to enter

the shamba again to carry out additional raids.

It would seem, therefore. that even though the vocal-

izations given during the raids are diversionary in nature,

they are probably not deliberately intended to be so. Instead,

the calls given in reaction to the intruding human help to set

up a fortuitous chain of events which allows part of the troop

to continue raiding.

At Roka, similar observations were made although

the conditions were somewhat different. Instead of a farm

surrounded on three sides by forest, as was the case at Shi-

moni, the farms at Roka were separated from the forest by a

fairly well-traveled road. When raiding, the baboons would

have to come out of the forest and cross the road to enter

the farms. Unlike Shimoni, crops were being grown in most

of the fields and some farms were protected by shamba guards,

usually women or children. In general, these baboons were not

very vocal at all but would carry out their raids in silence.

But again, it was observed that vocalizations of any kind or

any movement in the trees would tend to attract the guards

to that area while further down the road other baboons would

be free to enter the farms undisturbed. Although not inten-

tional, such behavior functions very well as a decoy mechanism.

The calls described in the following sections were

recorded during encounters such as those described above.

Classification of Calls

Two-Pnase Bark

e two-ohase bark (PFis. 32&33 and Table 20) was re-

corded on one occasion at Roka. This was just after a shamba

raid had been completed and all the baboons were on the for-

est side of the road. The observation vehicle was driven

adjacent to the point of entry into the forest. It was then

that a large, fully adult male conspicuously sitting in a tall

tree at the forest edge emitted two two-phase barks. The two-

phase bark has been reported by Hall and DeVore (1965) as an

attack/threat vocalization given in reaction to humans and

large predators.



Number (N) Mean (x) Range (R)

Units/Phrase 2/2 1.00 1.00-1.00

Phrase Duration 2 0.62 0.51-0.73

Lower Limit of
Energy (kHz) 2 0.25 0.25-0.25

Upper Limit of
Energy (kHz) 2 4.10 3.20-5.00

The two-phase bark is a loud, tonal, single-unit

phrase of low frequency and long duration. The energy dis-

tribution for the first two-phase bark is from 0.25 to 5.00

kHz and the distribution for the second is 0.25 3.20 kHz.

However, the heaviest concentration of energy for both calls

is in the 0.50 1.50 kHz range. The duration of the calls

is 0.73 seconds for the first and 0.51 seconds for the second.

On the sonogram the call appears as a very dark band

in the lower frequencies with some harmonic development.


Humans moving about in the farm at Shimoni would usu-

ally provoke bark-like vocalizations (Figs. 34&35 and Table

21) from nearby baboons who would then withdraw into the


There seemed to be two types of barks distinguishable

both acoustically and spectrographically; however, this may

be due to sex and/or age differences. Basically, the differ-

ence seems to be one of pitch, with one bark type noticeably

lower and less intense. Both may be described as single-unit




Number (N) Mean (x) Range (R)

Bark I


Phrase Duration

Lower Limit of
Energy (kHz')

Upper Limit of
Energy (kHz)

Bark II


Phrase Duration

















TABLE 21 (Continued)

Number (N) Mean (x) Range (R)

Lower Limit of
Energy (kHz) 7 0.48 0.40-0.70

Upper Limit of
Energy (kHz) 7 2.88 2.50-3.00

The average energy distribution for Bark I is 0.27 -
4.13 kHz (N=12). The call is tonal and harmonics are well

developed. The mean duration is 0.32 seconds (N=12).

Bark II is lower in pitch with an average energy dis-
tribution of 0.48 2.88 kHz (N=7). The harmonic pattern

does not seem to be as well developed here as in Bark I and

the upper frequencies of the call are somewhat weaker. The

call averages 0.32 seconds in length (N=6).


The yak (Fig. 35 and Table 22) was believed to have

been given by a juvenile. It was given under the same condi-

tions as the barks and was given in conjunction with them.

Structurally the yak is very similar to the barks
(especially Bark I) on the sonogram. It too consists of a

single-unit phrase. The main difference is that the yak is

higher in pitch and has a very well-developed harmonic pat-

tern. The yak is also acoustically distinguishable from the


TA3LM 22


Number (N) Mean (x) Range (R)

Units/Phrase 3/3 1.00 1.00-1.00

Phrase Duration 2 0.30 0.30-0.31

Lower Limit of
Energy (kHz) 3 0.50 0.50-0.50

Upper Limit of
Energy (kHz) 3 6.40 5.00-8.0+

The average energy distribution of the yak is approx-

imately 0.50 6.40 kHz (N=3); however, the harmonic pattern

on one occasion extended beyond 8.0 kHz. The average dura-

tion of the call is 0.30 seconds (N=2).

On the occasion when the calls were recorded the yak

occurred in a repeated sequence of Bark I, Yak, Bark II

given while an observer moved about in the shamba at Shimoni.

The baboons giving the calls were at the forest edge. While

they were vocalizing, the rest of the troop was quietly cir-

cling the farm a short distance within the forest.



The study of primate behavior developed rapidly in

the years following World War II to become a popular disci-

pline within the fields of zoology and physical anthropology.

However, one aspect of its study, primate communication, has

only recently begun to receive the attention which it rightly

deserves. That as yet no single study has been able to deal

with social communication in all its aspects is testimony to

the complexity of the task. Instead, individual researchers

have chosen to study some specific aspect or charnel of com-

munication, usually of a single species, in the hope that

with enough information collected in this manner a larger,

more comprehensive overview of the communication process as

a whole may eventually be possible. It is with this intent

that the present study has been conducted.

By means of sonograms the vocalizations of three

species of Old World monkeys (Sykes monkey, colobus monkey,

and baboon) were illustrated, described, and analyzed. The

conclusions and findings are summarized in the following


The Sykes Study

I) Fourteen Sykes vocalizations were catalogued and

described in terms of their physical characteristics (fre-

quency, duration, tonality, etc.). The possible functional

significance of the calls and any accompanying behavior was

also presented.

2) The vocalizations were examined in terms of dis-

creteness vs. gradedness. Most of the calls were found to be

of the discrete type but calls forming a possible gradient

were also discussed. It was emphasized that demonstrating

the existence of graded signals does not imply the existence

of a graded communication system.

3) The Sykes vocalizations show considerable variety

in pitch or frequency and timing. An attempt was made to

correlate these differences in structure with differences in

functional requirements, particularly the need for localiza-

tion. It was found that the properties of calls concerned

with group cohesion and alarm or threat which were transmitted

over distances without necessarily the aid of visual contact

were such as to facilitate localization of the vocalizer.

On the other hand, the structure of calls which were trans-

mitted at close-range where visual or other cues may be oper-

able did not show those features which would facilitate local-


4) The taxonomic position of the Sykes monkey, Cer-

cocithecus mitis, in relation to C. nictitans has been a

matter of some speculation. A preliminary analysis and

comparison of the vocalizations of the two species indicated

that separate specific status should be maintained at this

time. However, analysis revealed that C. mitis was more closely

related to C. nictitans and C. erythrotis-ceohus than to the

mona superspecies.

The Colobus Study

1) Two calls were described for adult colobus as well

as one for juveniles and two for infants. When possible,

their probable functions were discussed and any accompanying

behavior was described.

2) Although vocalizations were the primary concern

in this study, a nonvocal communication pattern, the jumping

display, seemed to play an important role in the life of the

relatively nonvocal colobus. This behavior was described

and its possible functional significance was discussed.

The Baboon Study

1) The physical characteristics of four types of

calls given during crop-raiding were described.

2) The apparent "decoy" effect of the calls in

attracting the attention of farm guards protecting the crops

while other baboons silently entered the fields to raid was

described and analyzed. It was concluded that the calls

were fortuitously diversionary in nature rather than repre-

senting a deliberate and intentional plan of behavior.

There had been a tendency in studies of primate be-

havior and communication to accept as generalizations for all

species that which had been described for but a few well-

studied populations. That the behavior of all species can

not be seen in a baboon-rhesus context or that the behavior

of all baboons is not typified by the Nairobi Park population,

for example, has become increasingly apparent as more and more

primate species and populations in various ecological settings

are being studied.

Before the formation of encompassing generalizations

then, if indeed such should prove to be possible, the primary

need in communication studies is one of collection of data

on as many species as possible. Ideally, information on com-

munication systems should be accompanied by information on

behavioral context and functional significance. Through ob-

servation of the behavioral context accompanying a communi-

catory act, something can be learned regarding the functional

significance of a particular vocalization, facial expression,

etc. Analysis of the relationship between structure and

function may indicate why certain structural traits have

proved adaptive and have been maintained through the process

of natural selection. Comparison of such findings from sev-

eral studies on separate species or on different populations

of the same species may reveal ways in which the ecological

setting can influence which adaptive mechanisms will predom-

inate. In addition, an understanding of the degree of vari-

ation to be found in communication patterns within species

will aid in taxonomic interpretations and in the assessment

of vocalizations and other channels of communication as tax-

onomic characters. In this way, one can begin to understand

something of the relationship between communication, organ-

ized social life, and ecology: of man's closest relatives,

the nonhuman primates.

Finally, it should be emphasized that primate commun-

ication should not be studied solely in terms of how many fea-

tures of human language may or may not be present, but also

as an entity worthy of study in its own right. However, an

understanding of the nature of communication systems among

nonhuman primates and how these systems relate to social or-

ganization and environment may eventually shed light on the

circumstances accompanying the origin of language in man and

on the nature of this early system.

That primate communication is a behavior which has

proved adaptive and evolutionarily successful can not be dis-

puted, but the mechanisms behind that adaptation are as yet

not fully understood. It is hoped that through further study

and fieldwork more and more information may be collected and

correlated to reveal something of the adaptive significance

of communication systems in general and of how these systems

are integrated into the larger whole of primate social life.

It is this intricate weaving and integration of behavioral

systems, of which communication is but one, which makes the

coordination of individuals into a society possible, and

which in turn insures the survival of the species.





8 iZ'j.

0i i~~

,9 LYI1

i I i

"' i. L 'I

4 jt -
'illef I ll, 1151( Ift s Ieritt 1 i


ft 1
S.4sra II


Time secss.)

Figure 1.

J I l .

1 i- pi'
'I 'i l '





Time secss.)

Figure 2. Chirp



O ,l


tl; -:

~a me


12 R

0 0.20 0.40

Figure 3. Two-Unit Chirp

Time secss.)

i" I ;

i ** 5" ,'V. , ; '" d

0 0.20 0.40

Figure 4. Two-Unit Chirp

Time secss.)

16 .

12 S -


0.20 0.40

Figure 5.

Time secss.)

Three-Unit Chirp

i .1

0 0.20 0.40

Time secss.)

Figure 6. Nyah




12 -L:4MR


F i I






Figure 7. Short Nyah

(short nyah followed by chirp of another Sykes)

16 ,. .

12 ;n A:
' -8..

0 .I'r
IaX h


'il :.

S1 0-.




Time secss.)

Figure 8. Short Nyah

(short nyahs followed by chirp of another Sykes)

-' y
*I 'I




.. I i I i H-




Time (sees.)

Figure 9. Boom

(ignore background noise at 3 kHz and 8-10 kHz)


ai 2 7 -- _-....- ..... ... _-.Qr,. ,..,..

0.60 Time


Figure 10. Scueal

- I

16 rm

12 B








University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs