Vol. 111, No. 981 The Amcrican Naturalist Septcrnber-October 1977
ON 'I'H 11: OCCURR11:NC11: ANI) SIGN I FICANCTC 011'
MO'I'IVA'I'I ON-K'I'RUC'I'URAT, RULES IN SOM11: RI RII AN11 MAMMAL KOUNIIK
National Zoological Parlr, Smithsonian Institution, Washington, D.C. 20008
This paper describes a concept, not altogether new but largely neglected, t h a t sllould lead to a greater undentanding of the information contained in certain classes of vocal communication signals of birds and mammals. The concept is based on empirical data, first pointed out by Collias (1960, p. 382), showing t h a t natural selection has resulted in the structural convergence of many animal sounds used in "hostile" and "friendly" contexts. Simply stated, birds and mammals use harsh, relatively low-frequency sounds when hostile and higher- frequency, more pure tonelike sounds when frightened, appeasing, or approach- ing in a friendly manner. 'I'hus, there appears t o he a general relationship between the physical structures of sounds and the motivation underlying their 1180.
I hope to develop the idea that this relationship has had a far greater influence on the evolution of animal communication systems than has hitllerto been discussed I will discuss the idea t h a t there exist motivation-structural rules (MS) governing the physical structure of close contact sounds in animal com- munication systems. The greatest value of the MS concept is that it provides the opportunity t o compare the evolution of vocal communication in any species against an abstract concept. 'I'he adaptive nature of communication systems against varying backgrounds of environment, social system, and competition will appear in clear relief.
1 will discuss the occurrence of MK on two levels. The first level is intra- specific and shows where MS affect sound signal structure within a species' repertoire. The second is interspecific and shows t h a t the empirical evidence mentioned by Collias (1960) indeed justifies the term "rule" and t h a t discussion of its biological significance is warranted.
I n t r a s p e c i j c O c c u r r e n c e
MS rules affect primarily, but not exclusively, sounds used by animals when they are close to one another. In animals t h a t are "face to face," the possibility and consequences of attack, escape, or association are immediate, and selection
Amer. Natur. 1977. Vol. 111, pp. 855-869.
O 1977 by The University of Chicago. All rights reserved.
855
will favor sound signals that express current and rapidly changing motivational states. Proximity lessens the difficulties of communicating but the consequences are more immediate.
Signals used by animals in close proximity may consist of high frequencies that attenuate rapidly, or they may have amplitade and sound-quality changes (e.g., harmonic e~upllasis or modulation rate chailges) t h a t may be quickly
~llaslted over distance. Close proximity thus pertnits the &IS rules to exist, hut it is the immediatte consequence of close proxi~nity that produces selective pressures favoring their existc.nce.
hl contrast, sounds used in long-distance sig~lalling are molded to a lesser extent by some selection pressures affecting short-range signals. For exanlple, selection favors specific distinctiveness in long-distance sound signals because other conlmunication modes are not possihle simultant~ously (Marler 1967).
Recause tllese signals are broadcast tllrough the environment, the environment may produce selection pressures favoring certain physical properties that increase their propagation (Morton 1975).
I n t e r s p e c i j c O c c u r r e n c e
Tables 1 and 2 list some bird and ~ r ~ a n ~ i r l a l species whose close contact llostile and friendly vocal sign;xls have been described. The list is not an exhaustive survey; i t is meant to show signals from animals vilrying greatly in size, taxoilomic affinity, evolutionary age, demography, ecology, and social system.
l'he vocalizations are either parapllrased onomatopoetically or described verbally, whichever the respective authors used.
If one quickly scans the sounds listed under "aggressive," and imagines hhem occurring simultaneously, an aural picture of vocal convergence is formed. Low, harsh sounds are coilsistently associated wit11 llostile inotivation. The same relationship between higher, pure-toned sounds and friendly or appeasing motivation is true, if the sounds listed in the nonaggressive column are scanned.
We are also able to perceive the convergence in underlying motivation, because humatls use vocal intonations expressing hostility or appeasement in the same general way.
The convergent evolution exllihited in these sound structures has been suggested t o reflect Ilarwin's principle of antithesis (Collias 1963). 'I'here have been no other attempts to identify sources of selection behind the convergence.
Darwin (1873) explained only the divergence in signal structure.
Convergence in sound structures correlated with n~otivational state is badly in ncxd of theoretioal explailation. I have made sucll an attempt, first by dis- cussing the MS rules and then hy providing hypotheses on the sources of selection operating. I will incorporate examples of the MK rule operation into the arguments to illustrate how the concept rnay provide insight into the evolution of sound signal structure in other than obvious ways.
THE MOTIVATION-STRlJCTURAL RlJLE THEORY
Harsh (covering a wide-frequency band), low-frequency sounds used in hostile contexts and more pure tonelike, relatively high-frequency sounds in friendly
TABLE 1
AVIAN SOUNDS USED IN HOSTILE OR "FRIENDLY," APPEASING CONTEXTS
Species (family) Aggressive Konaggressive Source
White pelican, Pelicanus erythrorhynchus (Pelicanidae) . .
.
. . . Harsh nasal growls* Kot given Schaller (1964) Mallard, Anus platyrhynchos (Anatidae) ..
. . ..
. . . Loud harsh gaeck(9)
Soft whimpers: k n and Abraham (1974)quais
(9)
Sparrow hawk, B'alco sparverius (Falconidae) . . .
. .
..
. . . Harsh chitter Whine JIueller (1 971) Bobwhite, Colinus virginianus (Phasianidae) . .. .
. . . ..
. . .. .
Loud, rasping "cater~vauling" Tseep; squee Stokes (1967)Ring-necked pheasant, Phasianus colchicus (Phasianidae). . . . Hoarse krrrrah Squeak
( q )
Heinz and Gysel (1970) Solitary sandpiper, Tringa solitaria (Scolopacidae) . . ..
. . . . . . . Harsh, metallic sound Rising shrill whistle Oring (1968)Stilt sandpiper, itficropalama himantopus (Scolopacidae) . . . Trrrr T o i , weet Jehl (1973) Cassin auklet, Ptychoramphus aleutica (Alcidae) . .
.
. . . .. .
. . . Growled krrr krrr Kreek Thoresen (1964) Orange-chinned parakeet, Brotogeris jugularis (Psittaoidae) . . . . rrrrr Low intensity "chirp" Power (1966) Burrowing owl, Speotyto cunicularia (Strigidae) . . . rasp eeP Martin (1973) Red-headed woodpecker, Melanerpes erythrocephalus (Picidae) . Chatter, rasp Kot given Bock et al. (1971) Harlequin antbird, Rhegmatorhina berlepschi (Formicariidae). . . Growling chnuhh chee TVillis (1969)Chestnut-backed antbird, Myrmeciza exsul (Formicariidae) . . . . Snarling nasal chinngh \lusical chirps: cheup Willis and Onilti (1972) Eastern kingbird, Tyrannus tyrannus (Tyrannidae) . . . . . . . Harsh zeer High-pitched tee Smith, IT. J . (1966) Barn swallow, Hirundo rusticu (Hirundinidae) . . . Deep harsh stutter Whine call Samuel (1971) Purple martin, Progne subis (Hirundinidae) . . .
.
. . . zturack sweet Johnston and Hardy(1962) Mexican jay, Aphelocoma ultramarina (Corvidae) . . .
.
. . . . . Not given Variable weet Brown (1963)Scrub jay, A . coerulescens (Corvidae) . .
.
. . . . . . . . . . . . . . . Harsh rattle whew, scree Brown (1963)5
F Dwarf jay, A . nana (Corvidae)
.
..
..
. ..
. . .. .
. . . Harsh rasp shreeup Hardy (1971) 7 Common crow, Corvus bmchyrhynchos (Corvidae) . . . . . . . . ..
Growl Soft and plaintive Chamberlain and Corn~vell 9(1971) (= 7
Carolina chickadee, Parus carolinensis (Paridae) . . . . . . . . . . . Click-rasp Lisping tee, soft dee, Smith, S. T. (1972) M
high see Ui.
Blue-gray gnatcatcher, Polioptila caerulea (Sylviidae) . . . . . peew speee Root (1969) American redstart, Setophaga ruticilla (Parulidae)
.
. . . . . . . . . Snarl zeeep, high-pitched titi Ficken (1962) Yellow-headed blackbird, Xanthocephalus xanthocephalus(Icteridae) . . .
.
. . . . . . . . . Harsh, nasal rahh-rahh pree pree pree Nero (1963) Crimson-backed tanager, Rhamphocelus dimidiatus (Thraupidae) Rasping harsh hoarse notes Sseeeeeeeet Illoynihan (1963) Brown towhee, Pipilo fuscus (Fringillidae) . ..
. . . Snarling throat,y notes Seeep, squeal duet 3Iarsha11 (1964) Common redpoll, Acanthis j'lammea (Fringillidae) . . . . . . . . Harsh cheh cheh cheh sweeeee Dilger (1960)African village weaverbird, Ploceus cucullatus (Ploceidae). . . . . . Harsh growl look! see!; high squeal Collias (1963) CK Cil
*
Verbal or onomatopoetic (italics) renditions of sounds quoted from source author's descriptions. -1TABLE 2
JIAHHALIAS S o u s o s USED IS HOSTIT~E OR "FRIESDLY. " -~PPEASISG CONTEXTS
Species (family) Aggressive Konaggrcssire Source
. . .
Virginia opossum. Didelphis mccrsupialis (Didelphidae) Growl Screech Eisenberg c t a1 . (1976) Tasmanian devil. Sarcophilus harrisii (Dasyuridae) . . . Growl 11-hine Eisenberg et a1
.
(1975)Wombat. Vombatus lccsiorhinus (Phascolomidae) . . . Deep growl quer-qner-quer Eisenberg et a1 . (1975) Guinea pig. Cazjia porcellus (Cariidae) . . . Grunt. snort Squeak. wheet Eisenberg (1974) Lfara. Dolichotis patagonum (Caviiclae) . . . Low grunts Inflected wheet Eisenberg (1974) Curo curo. Spalacopus cynnus (Octodontidae) . . . Groxvl Short squeaks Eisenberg (1974) Degu. Octodon degus (Octodontidae) . . . Growl Inflected squeak Eisenberg (1974) Spiny rat. Proechimys semispinosics (Echimyidae) . . . Growl Twitter. w h i ~ ~ l p e r Eisenberg (1974) Agouti. Dasyprocta punctata (Dasyproctidae) . . . Groxvl. grunt Squeak. creak-syueak Smythe (1970) Pocket mouse. Heteromys (2 sp.) (Heteromyidae) . . . Low scrat'chy growl TVhining squeal Eisenberg (1963) Pocket mouse. Liomys pictus (Heteromyidae) . . . Low scratchy growl TThining squeal Eisenberg (1963) Desert pocliet mouse. Perogncrthus ( 4 sp.) (Heteromyidae) . . . Low scratchy groxvl TThining squeal Eisenberg (1963) Kangaroo rat. ~1licrodipodops pcrllidus (Heteromyidae) . . . Low scratchy growl TThining squeal Eisenberg (1963) Kangaroo rat. Dipodomys ( 6 sp.) (Heteromyidae) . . . Low scratchy growl TT7hining squeal Eisenberg (1963) Lemming. Dicrostonyr groenlandicus (Cricetidae) . . . Snarl. grind Tl'hine. peeps. squeal3 Broolis a n d Banlis (1973)
. . .
Uinta ground squirrel. Citellus armcrtus (Sciuridae) Gro\vl Squeal Balph a n d Balph (1966)
Maned ~volf. Chrysocyon bruchyurus (Canidae) . . . Growl U7hine Iileiman (1972) Bush dog. S'peothos venaticus (Canidae) . . . Buzzing growl Squeal I<leiman (1972) Coati. S n s u a ntrrica (Procyonidae) . . . Growl Squeal Iianfmann ( 1962) Large spotted genet. Oerzetta tigrirla (Viverridae) . . . Gro\%-1-hiss Whine or groan 1Vemmer (1976)
.
African elephant Loxodonta ufricclna (Elephantidae) . . . Roaring. rumbling sounds H~gh-frequency (iuntlh Tembrock (1968)
.
Indian rhinoceros Rhinoceros u n i c o r ? ~ i s (Rhinocerotidae) . . . Roaring. rumbling n-histling Tembrocli (1968) Pig. S u s scrofc~ (Suidae) . . . Growl Squeal Tembroclc (1968) Llama. L a m a gzlanacoe (Camelidae) . . . Growl Bleat (long distance only ' 1 ) Tembrocli (1968) Bluntjac. *Tfuntiacus muntjuc (Cerridae) . . . Kot given Squeali Barrette (1975) Squirrel monkey. S a i m i r i sciureus (Cebidae) . . . Shriek calls. err Peep calls. trills Schott (1976) Spider monkey. Ateles geoffroyi (Cebidae) . . . Growl. roar. cough Tee tee. chirps. twittcr Eisenberg and I<uehn
squeak (1966)
Rhesus monkey. LVncrtca mulotta (Cercopithecidae) . . . Roar. growl Screech. clear calls. Roxr-ell and Hinde (1962) squeak. nasal grunting
whine. long growl
MOTIVATION-S1'RUC'l'UIiAJd RULES 859 or appeasing contexts represent endpoints of a behavioral spectrum. Generally, animals do not conlnlit tlr~emselves inlmediately t o act out either of these endpoints when they encounter or join another conspecific unless, perhaps, when they have had previous experience with one another (i.e., they "linow"
each other). 7'11is is true even in situations wl~cre the encounter usuttlly results in fighting or clr~asing as, for example, when a male songbird enters the breeding territory of another conspecific rnale (e.g., Krnlen 1972, p. 135). An animal's comnlitnlent t o one of the diverse arrtty of possible bel~aviors when near a conspecific is under extremely strong selection pressures. 'Yo give a wcll-worn example: If the species' food is found in a certain degree of abundance and with a patchy distribution, selection inay favor responses t h a t proniote group col~esiveness because of the antiyredrttor and/or food locating value of being in a group. I n this situation, an individual whose genome (ttnd/or previous experi- ence) results in a highly aggressive phenotype anti does not join o t l ~ e r s has a lower survival potential, all else being equal. Iiowever, food availability and distribution change, anti so do the relative rnerits of a n individual's increasing or decreasing the distance bctween i t and conspecifics. 'I'lr~e point is t h a t selection c;tn never nlolrl a n aninlal's o v ~ r t rc;tct,ions on even so simple or ea.sily measured a parameter as distance relations I~etween indivithlals t o a n optinrum point;
it is always a dynamic equilibriunl. What selection does favor is the ability of the individuals in a population to respond t o one another in a, way t h a t max- imizes fitness. Coininunication is the inearls by which the aninials in a population ultimately adjust their social relationships t o various environmental and physiological fluctuations. The endpoints in the sound stnlctures used in this comnlunication are rarely adaptive for an individual t o ernploy consistently.
Rather, various grades between sounds used in fighting and appeasement are favored by selection.
I contend t h a t we may better understand evolution in signal structure if we assume t h a t the endpoint sound struct~lres permeate the cornnmnication signals t o intiicate various tlegrees of hostility or friendly appeasing behavior, even if no overt reactions call be observeti (since often neither sender nor receiver comnlits itself while uttering solliitls t h a t refiect indecision).
There are inmlrnerable examples t o show t h a t this view has merit. Marler (1956) in his study of chafiiincl~ (E'r%ng%lla coelebs) sounds lists a "social call" used to bring separated birds together (and thus not governed strictly by the close contact MS rules) and several other variants used in close contact situations.
These latter variants range from a "shrill form" used in escape and appeasement contexts t o a lower frequency, lrarsl~ form used wlr~en fighting. 'Ylnls, even though the "social callis)" can be distinguislled from other calls, its variants reflect sound qualities approaching those in the chaffi~iclles' seeee lligll-intensity-alarm and sexual-excitement call, on the one hand, and the low, buzzing xzzxz fighting call, on the other.
Susan T . Sinith (1972) describes communicative behavior in the Carolina chickadee (I'ar~cs carolinensis), a. spccies t h a t is never solitary but has periods of aggressive rivalry as well as sot:ial flot:king. Its corrlplex social interitctiorls are finely tuned by variable calls. The MS rules are evident throughout t h e
860 T H E AMElZICAN NATURAIATST
close contact calls of this species. For example, Srnith lists three "displays"
associated with a n attack probability, rendered phonetically as T-slink, slink- rasp, and cl%ck-rasp. They are all variable in str~lcture, particularly the cliclc-rasp forms. Smith found t h a t a bird giving 'I1-slink was less apt t o attack than one giving slink-rasp, a slink-rasping bird less a p t t o attack than a click-rasping one.
The MS rule predicts this motivational sequence because we find t h a t high- frequency pure tonelike somrds are reduced and harsh low-frequency sounds are ii1c:reased as the calls change from T-slink t o click-rasp. 13ut ml inlportant point to note is t h a t the chickndee employs sounds t h a t are composites of hostile and friendly motivation. 'l'hnt is, there are two signal str~lctural endpoints (harsh mid pure tonelike) t h a t are varying, perhaps independently, to produce corllpositc structures commlmicating extreirlcly slrhtle rllotivatiorial changes in the signaler. ( I suspect t h a t the syntactical order of these sounds is also important.) Both the T- t o click- and -slink t o -msp portions of the solrntls vary greatly, but the variations were not dcscrihctl in depth because of the categorization procedure used ( i . ~ . , the display conce1)t).
I n his excellent analysis of kinghird (2'ymnnus sp.) sounds, W. John Smith (1!_166) discusses variations t o the extent t h a t we niay easily identify the operation of MS rule. Smith categorizes eastern kingbird (2'. tyrannus) sounds into seven displays hut recognizes t h a t each grades considerably: "The LHV forrns iritermrtliates with nearly every other voct~lizatiori of the species" (p. 1 3 ) . Each vocalization cliangetl into a "harsher" forill whenever the calling bird beoarne aggressive. A kinghirtl being attacked by another gives high-pitched tee sounds (Si)lith3s AeV sound). "A Kingbird t h a t attacks a larger bird (like a Ttohin) and loses the initiative will usually switch from the harsh TtV t o the high AcV . . ." (p. 23).
An unusually high level of intra- and interspecific aggressiveriess during the hreetling season has heen sclectetl for in eastern kinghirtls. 'l'he evolntion of their sound signals, nrhich arc nearly all harsh, probably reflects the kinghird's aggressive nature. 111 nligration and on their tropical wintering grounds, kinghirtls aggregate in rnonospecific flocks and search for fruit (Morton 1!171).
Even though they are in flocks, almost no vocal or hostile interactions take place. Vocal conlrnu~licatio~l is adapted t o hreetling season behavior which favors aggrcssivc sound structure.
Gradation in kingbird sounds serves t o communicate rather subtle variations in motivation. The cvollrtionary need for such variation may he related t o pair bond formt~tiori and nlaintent~ncc which occur against a background of sexual rnoriornorphisn~ and high aggression, attributes which detract f r o n ~ the easy recognition of potential breeding partners.
These cxt~rnplcs illustrate ways t h a t the MS rules are incorporated into the physical structure of sounds. It is sigriifict~nt t h a t painstaking eRort was involved in inferring the. underlying motivation of the sounds because the conclusions were based on overt responses, some of which were quite subtle.
Even so, much information in the sounds, particularly iriforn~atiori contained in srnall changes of physical structure, probably escaped notice because no overt responses occurred. MS influences the physical structure of close contact
MOTIVATION-STRUCTURAI, I<UJ,RS 861 solrntls in other ways t h a t , for the sake of brevity, 1 will list. The list incl~tdcs all of the known or probable ways t h a t MS influences signal structure.
1. Harsh, low-frequency sonntls iridict~tc t h a t the sender is likely t o attack if the receiver cornes closer t o the sender or remains a t the same distance. Tn some species, the harsh endpoint is given only (luring the attack, in others when attack is imminent as well.
2. Relatively tonal, high-frequency sounds intlicatc t h a t the sender is submissive, will not be hostile if approached or if approachi~~g, or is fearful.
3. Harsh sound quality, tonal quality, and sound frequency (pitch) interact such t h a t : ( a ) the higher the frequency, the more fearful or friendly the sentler;
the lowcr the frequency, the inore hostile; ( h ) the greater t h e sound's harshness, the greater the aggressive motivation; the more pure tonelike, the more fearful or friendly, no matter what frequency range is used.
4. Sonntls rising in frequency (no n ~ : ~ t t e r wh;tt the sound's qn:~lity) indioxte :L
lower hostility or inc2re:~sing appeasen~eiit or fear, but :L lnot,ivational endpoint, is riot indicated. Sounds tlccreasing in frequency indicate an increasing hostile motivation.
5. A sourltl whose frequency rises arid 6~11s rliore or less equttlly or is frequericy coristarit but midrange in the overt~ll frequency range reflects a conflict of motivation t o approach or witlitlraw from a stinlulus. T t indicates t h a t a stimulus of "interest" has been received by the scndrr.
6. A species t h a t is generally more aggressive t o conspecifics will tend to have a harsher close contact vocal repertoire as opposed t o a species t h a t often joins or is joined in flocks, especially mixed species groups. The latter will have
a, prevalence of high-frequency pure tonal sounds in its repertoire.
7. A s p e c i e w i t h greater complexity of social interactions will evolve sound signals containing a more cornplcte range of sound qualities indicating more points along nlotivatiorial gradients arid rapid changes in motivation (as in paragraphs 1-5 above) than will a, species with less complex social interactions.
8. An individlral uttering "alarm" sounds will be more likely t o withdraw from the alarming stin~ulus, the higher pitched its sounds, if the alarm system is gradcd. 'l'hc alarrn system will tend t o be gradcd if kin are predictably (to t h e sender) near or if coordination of an escape attempt as a group will reduce t h e chances of predation on the sender.
Figure 1 depicts tliagrammatict~lly the sound structures associated with varying degrees of hostile anti appeasing contexts according t o the MS rules. Height above the haseline reflects frequency; arrows indicate t h a t the sound structure may be positionetl higher on the frequency scale if the sender if more appeasing, friendly, or fearful and lowcr in nlorc hostile animals. The slope of the calls with ascending or descending frequencies may vary, as indicated by (lotted lines.
Thus there is potential grading frorn the signals in each box to an atljacerit signal. Presumably, a sound iudict~ting ambivalence, such as occurs in mobbing behavior (e.g., Andrew 1961), may acquire a steep slope so as t o become nearly a pulse if selection pressure derived from the sound's function favors qualities t h a t enhance the sender's location by the receiver (Marler 1956).
This type of sound structure, in the center of figure 1, is predicted by MS t o
THE AMEliICAN NATURALIST
INCR. HOSTILITY
b
PIG. 1 .-Sound structures associateti with varying ciegrccs anci combinations of hostile anci fearful or appeasing n ~ o t ~ v a t i o n a l states. In each block, the frc- quency is inciicated by the figure's height above t.he baseline. A harsh sound is indicated by a wide blaclr line, a tonal sound by a thin line. Arrows inciicatc the potcntial for frequency change within each "motivation square," anci the dotted lines indicate t h a t the figure's slope rnay change (see text).
have a clrevron shape because it is iritcrrnctliate t o the two structural arid motivational endpoint sounds (fig. 1, lower left and upper riglit boxes;
paragraph 5 on the list). The chevron sliapc is, in fact, extremely oornnloii in avian vocal repertoires, usually rendered phonetically as chip. I n mammals, the intcrmetliate structure tends t o be freque~ioy constant but still short or abrupt, and the sountls are termed harks or qvunts. For both birds and mammals, this sountl type intlicates the sentler is indecisive (i.e., it may either go towartl or away from or become 111orc or less aggressive or appeasing towartl the stimulus), usually because the stimulus is too far from tlie sentler for it t o make an atlaptive response. Or tlie stimulus itself may protluce the indecision, such as when small birtls give chips while mobbing a potcntial predator.
To understand the prevalent, use of interrnediatc sound structures in the MS range, we finti t h a t it is riot possible t o separate function from communic' <I, t' ion.
That is, a n indecisive sountl structure will be atlaptive when the sentier perceives a stimulus too far away for its sensory system t o adequately protluce an atlaptive response. I n mobbing, however, the stimulus is known but the response is indecisive because function has produced selection favoring indecisiveness.
Selection favors the sender remaining indecisive for a time period because, in
this exanlplc, t o other snlall birtls the sender communicates t h a t sonlething
"of interest" is perceived. (How sorrlctliirlg beconles of iriterest is another story, hut it lnay be genetically determined, c.g., S. M. Snlith 1975, or 1c;~rnetl.) Others arc attracted to "moh" and therehy increase the foraging tirnc (this secrns the most likely henefit) of the sentier hy causing the stimulus t o move away or by reducirig the chances of the pretlator catching sornethirig anti theroby rctl~~cing the sentler's foraging time hy keeping its "interest up" (e.g., if n snako cnt,ches a kjird, the long time it takes to swallow a bird keeps it, as a stimuli~s for mobbing, more stimulating oj' nlohhirlg because of its continuous rnovenlent and other stiinlrllrs chariges
I
see Shalter 19751 that irihihit habituation [persorlal observation]). On the other h;lntl, because chips arc also given when the sender-stimulus distance may be great, selection rnay also atld specifically distinctive qualities if they function, e.g., in mate attractiorl or territorial tlcfense as well.I have thus far restricted the tiiscussion of MS rules to close-contact sountls for reasons outlinetl above. However, there appear t o be many ct~scs in which they inny affect sigrial strncture in hrondcast calls and many cases in which a sound is nsed for both close-contact and distancc coniniunication. The MS rlxlcs are especially likely t o operate on the long tiistarlee or broatlcast calls of species livirig in groups. I n these species, irltragroup hostile communication involves low, llarslr sountls antl it is reasoriahle to expect t h a t intergroup conlmurlication over distance will evolve frorrl the same sourids used in close-contact (intra- group) situations. For example, the low-frequency roar (energy is coriceritratetl a t ahout 360 H z ) of the howler nlonkey (Jlounttasp.) is often cited as ari exarrlple of selection favoring low fi.eqaencies for tlistance sol~rltl propagation in forests (Altmanri 1!167, p. 329). Ilowcver, data for sound propagation suggest t h a t the howler colrltl (lo just as well by calling a t 1500 Hz (Morton 1975). I suggest t h a t the low-frequency roar of the howler resultctl from selection favoring the use of what was originally a close-contact hostile sountl as a long-tlistance hostile sound. The tlifficulty irlvolvcd in turning a close-contact sound into a, broadcast signal is attested t o by the tlevelopnlent of the howler's laryngeal sacs and enlarged hyoid bone resonating system. The result is the ahility t o broadcast a low, harsh hostile sourltl over great tlistance.
An avian exanlple of MS rules affecting long-distance sounds is found in the stripe-backed wren (Cam,pylorhynchz~.s nuchnlis) which defends permanent group territories hy using harsh, low frequency tluets and uses shorter, softer versions of the same sourltl (luring intragronp aggression (personal observation).
However, most bird songs, because they arc long-tlistance signals, do not incorporate MS characteristics. Perhaps as a consequence many species have several t o rrlarly song types t o avoid habituation by receivers t o the aggressive message in birtl song (Petrinovich arltl Peeke 1973; Falls and Krehs 1975).
EVOLUTIONARY SIGNIFICANCE
Why should the close-contact vocal signals of animals converge t o such a great extent t h a t we may postulate a motivation-structural rule ? Why do some
864 THE AMERIC'AN NATURAJAISrY
anirnals ilot "growl" when thcy are appensing antl "whine" when they are aggressive or do somcthirlg entirely differe~lt ?
T hclieve sorne insight illto these questio~w lrlay be gained by viewing the two motivational endpoints separately, as if thcy were untler completely different sorts of selective pressures, wfrich intleetl they may hc.
Lozc?, Harsh sounds
I hypotfresizc t h a t low-frequeilcy sountls will be favored i11 hostile cnconntcrs because there is a direct relationship between low frequency ailti the size of the aninla1 protlucing the sound: Thc lnrgcr tfrc animlal, the lower the sound fre- quency it can protluoe. This is presumably a, law of physics. 80 is harshness:
A vibrating nlcirrhrarie will protluce harmonics as the result of increasing nonlinearity as tlic frequency (~riemhraric tension) falls (Greeoewnlt 1068). So if there is selectioil f;avoring tlie use of low-frequency somltls by hostile animals, these souritls will antonratically he accompanied by inoreasetl harsliiless due to harinoriic prodnotion and other OK toilal sonntls prothlcetl by n vibrating flaccitl rnembralie.
Triterestingly, Grcenewalt (1068, p. 164) ilhlstrated this phe~lomenon using blaoli-capped ohicliadee (Parus utricapillz~s) "scoltl" notes. This ca. !) g species protluced a, fi~ndarrieritxl frequency of 415 I l z (nearly as low as a howler monltey roar) in this hostile sound. But thc hird's srr~all size brought about the a~riplitnde emphasis of higher harmonics tluc t o resonailoe. Harshness derivetl frorri its physically inevitable association with low-freqne~ioy sound production has been ritllalizetl as a coiriponent of hostile com~rrunicatiori even when it occurs apart from the hostile endpoint sountl timing a selider's ]nor(, ambivalent nlotivatioiial states.
The use of low-frequency sountl represents the size of the sender t o the receiver. This communicative rclatio~lship betwccn sound frcqnency anti body size is untloubtedly an ancient coupliiig. It probably first arose in animals having intletermi~lant growth such as reptiles arid fi~nctio~letl simply t o i i l t i ~ n - itlate a smaller animal or formed part of the basis by which aniinals jndgetl each other's size before committing tlienrselves t o an action.
Botly size hems a, strong rclatiori t o the concept of MS rules in oonrmnnication antl its evoh~tion. An appreciation of this relationship necessitates an under- staridirig anti appreciation of tlie two etlgetl swortl t h a t botly size represents in evolution. A larger animal will generally win in a fight against a snraller one;
this relatio~lship has had an enorllious selective eKect on organisms. Inter- specific social tlominarice (the ability of a larger species t o exclutle a snraller fi-om some resource) is an important source of selection moltling ecological relatio~lships between species (Morse 1074). Morse (p. 826) points out t h a t species of birtls arid mammals are prone to donlinanoe-metlit~tetl iiiclie partition- ing because they are characteristically aggressive intraspecifically and use t h e same aggressive patterns interspecifically. Piglitirig with conspeoifics generates sclcctiorl pressure favoring increasing size (e.g., Johnston e t al., 1972). However, body size must ultimately be kept within energetic hountlaries. Realizing t h a t
body size may increase d r ~ e t o i~ltraspecsifio sonrccs of seleotioi~ arising froill aggressive behavior, it is liBely t h a t , duriiig cvo1utio~r;~i~y time, species t h a t tend t o fight for resonroes niay be rather prone t o extinctioii. These are f;~rtlicr froin an e~lergetically optimum size in teivis of their ability t o survive environ- m e n b l fluctuations (Van Valen 1Y73). Wv cml iiow cilvision a rlyiianiic relation I~etween botly size ant1 social behavior thnt, if the intricacy of social Izhavior is relatetl t o brain size, ~riay explain both tlrc extensive use of
MS
rnlcs in commnilication and the e~ll;~rgc,~rlent of braills over ti~ric (Jerisoii 197:3).The argunrent is t h a t genes t h a t pronlotc fighting, arlierc larger botiy size is favoretl, will tenti t o be replacetl hy gei1c.s iiivolvetl in the oomnlrlnication of motivatioil if tlie comiiinniontio~l gcnes are as effective as fighting in acqrliring the resources in question. Genes promotilig coi~inluliicatio~i will be as cffc,ctive as those promoting fighting if tlie conrnrlmication systcm replaocs sizc as the main tleterirrinant of \vlrioh gene carrier obtains the resources. This secms t o lcad inevitably t o low harsh sormtls as the adaptive solnrd strr~ctarc in lrostile contexts, since lower solmds represent larger size. But the ~ i e u r a l proocssi~ig of vocal commmlication signals ailti the ability t o recall past expcric~iccs in aggressive enconilters, t o cnahle ail atlaptive response t o close contact n,ith a oonspecific anti thus be able t o use vocal comnlunication cn'ectively, probably requires allti favors a, larger brain. The result and selective value of ~ l s i ~ i g com- munication over fighting is t h a t size will b(, illore in balance witli nlerget,ic reql~irements (i.c., the commn~licati~ig population wi!l be more likely t o survive environmental changes than tlie fighting pojmlation). 'l'hc fact thnt tlrc visual commlnlicatiori channel is also size ilidicative atitls considerable wciptrt to this hypothesis (e.g., piloerection t o increase appare~lt sizc ill aggressive e~roonnters and vice-versa).
H i g h , Pzire T o n e l i k e So7i,nds
The sanrc discussion, in reverse, col~ld apply t o the MS rule of high tonal sounds ill fearful or appeasing contexts, nanrely, tlrat a well-stretchetl vibrating membrane produces more pure tonelike sounds as well as higher frequencies.
'I'his in itself does not explain why the use of this somlti is ntlaptivc for tlie sender, because it tioes not tell us why i t might change the receiver's behavior in a manner be~lefioial t o the sendcr.
Two relatetl attributes of parent-young relationships in birds aiid ~ n a ~ m n a l s seem nrost likely t o provide tlie answer. It is well known t h a t nrariy morpho- logical features of infant animals are atlaptetl for releasing parental responses in the athllt parent(s) (e.g., lJorenz l!j4:3). Vocal sountis are probably adapted for the same finlotion in infant animals: nearly all infant vocalizations are high frequency anti pm-c torlelike sounds t h a t u~onld tentl t o attract the athllt rather than repel i t , according t o MS rules. Tn altricial birtls, young in nests are both competing for footl witli siblings and attempting t o attract tlie athllt. One wo~ltiers why they shonltl vocalize a t all in this pretiato~--vl~lnerak)le stage, hut apparently selection favors vocalizatioris t o elicit parental ~ s r s anti t o direct food toward the calling nestling. I n nestlings old enough t o see, these calls are
866 T H E AMEKlCAN NATTJKAT,TST
sometimes gradetl such t h a t , when the parent approaches closely with footl, the calls rise in pitch anti rate of utterance. The yonng bird is sending an increasingly appeasing and therefore attracting rnessage t o tlie adult, attracting the atlnlt t o p1" food into a, gaping m o l ~ t h which might be intirnidatirlg without the sound. We fintl, then, selection arising from sibling food oompctition t h a t favors nestli~lg vocal commllnication following the MS rules.
Scleotion also favors athllts t o respond appropriately t o stin~nli designetl t o elicit parental care. Because yonng anirnals are smaller thari adults, i t seems axiomatic t h a t vocalizations of infarltms wonld be rather high in the species' frequency range. Thus a n athllt vocalization tlesigned t o rednce aggression by eliciting a, parental care-hostile behavior conflict in tlie receiver and/or by a vocal indication of small size (thereby reducing a fern- response in the receiver), should be high frequency arid tonal. 'I'he probability of a signal system evolvi~lg t h a t contni~ls high-frcqnency sollritl structures in hostile contexts, the reverse of the MS rnlcs, seems remotc on these grollntls.
I have refrained from using a great number of examples of tlie llsc of MS rules so t h a t the concept itself wonld stallti clear as an nbstractior~. The utility of the concel~t as a rriearis t'o tie signal structure t o ~ilotivation may be tested by st~ltlying oommmlicatiori in particlllar species. It is hopetl t h a t by doing so ure will tlemoi~strate t h a t communication systerris are adaptive on more ge~leral levels than current literature indicates. MTe will be able t o describe aggressive- ness a t the species level, for example, as indicated in paragraph 6 in the above list.
On an intlivitlual level, we may he able t o characterize "persoi~ality" in species in which ir~dividuals in social groups tlifyer in aggressiveness according t o their use of graded sountl signals following
MS
rules and then ask how tlifferences in "personality" may he atlaptive frorri a, population genetics standpoint. The link betwecn sourid structure antl i~iotivation will bring other sources of selection on signal structm~e into clear relief. The effect of diffcring social systems and their uriderlyirig ecological bases on communication will he more obvious. We shoultl be able t o explain, e.g., t h a t nocturnal monkeys use stereotyped sountls not only beonuse they cannot afford "misuntlerstantlings"(e.g., Moynilian 1964, 11. 45; Kchott 1976, p. 246) but because the us(% of long- distance calls is morc prevalent, and thus the use of the solliltl structure- motivation relationship is not selected for.
T t is probably correct t o point out the analogy between the universal rules of grammar found in human speech (e.g., Chomsky 1975) antl MS rules of animal communication-both opert~te (luring close-co~ltact vocal comnilnliot~- tion. Perhaps a major diffcrerice in the evolution of human speech antl aniinal vocal communication is t h a t more signals function, in general, for long-tlistance communication in animals thari for man. The "display" concept of ethology, when applied t o vocal communication, emphasizes stereotypy and species distinctiveness in vocal signals. It is largely derived from studies of long-
MOTIVATION-STRUCTUFlAL RULES 867 distance vocal signals and is most useful for describing only those. Close contact permits "greater exploitation of signals t h a t are highly graded in structllre rather than stereotyped" (Marler 1967, p. 771).
The concept of MS rules in animal commllnicatiori should permit studies witli a, greater interpretation of the significance of variations in vocal signals than was provided by the display concept, which neglected the dynamic natllre of conrmuriication for the ease of descriptive categorization of signals.
STJMMARY
The convergent use of harsh, low-frequency sollrids by hostile animals and more pure tonelike, high frequency sollntls by fearful or appeasing animals is discussed in all evolutionary context. It is proposed t h a t many sounds in species' repertoires are evolved from motivation-structural rllles derived fronr selection pressures favoring the use of comnrnnication instead of, or in con- junction with, fighting t o attain resources. The use of this concept sliollld filrtlier the appreciation of the relationship between sound s t r l l c t ~ ~ r e and function.
ACKNOWLEDGMENTS
I thank E. Eisenmarin, D. Klciman, W. Dittus, and M. Shalter for their criticism of the manllscript and snggcstions. The paper was horn out of many conversations witli J . 3'. Eisenberg, whose many ideas were generously shared.
LITERATURE CITED
Abraham, R. L. 1974. Vocalizations of the mallard (ilncrs plntyrhynchos). Condor 76:401- 420.
Altmann, S. A. 1967. The structure of primate social communication. Pages 325-362 in S. A. Altmann, ed. Social communication among primates. University of Chicago Press, Chicago. 392 pp.
Andrew, R. J. 1961. The motivational organization controlling the mobbing calls of tho blackbird (Turdus merula). I , 11, 111, ancl IV. Behavior 17: 224-246; 17: 228-321 ; 1 8 : 2 5 - 4 3 ; 18:161-176.
Balph, D. M., and D. F. Balph. 1966. Sound communication in Uinta ground scjuirrels, Citellus nrmntus. J. Mammal. 47 :440-449.
Barrette, C. 1975. Social behavior of mnntjac. Ph.D. thesis. University of' Calgary. 234 pp.
Bock, C. E., H. Harlott, ancl P. Somers. 1971. Relations between Lewis' and red-headed woodpeckers in southeastern Colorado. Wilson Bull. 8 3 : 237-248.
Brooks, R. J., and E. &I. Banlts. 1973. Behavioral biology of the collared lemming (Dicro- stonyx groenlandicus (Traill)): a n analysis of acoustic communication. Anim.
Behav. Monogr. 6 ( 1 ) : 1-83.
Brown, J. L. 1963. Social organization and behavior of the Mexican jay. Condor 6 5 : 1 2 G 1 5 3 .
Chamberlain, D. R., and G. W. Cornwell. 1971. Selected vocalizations of the common crow.
Auk 8 8 : 613-634.
Chomslty, N. 1975. The logical structure of linguistic theory. Plenum, New Yorlt. 500 pp.
Collias, N. E. 1960. An ecological and functional classification of animal souncls. Pages 368- 391 in W. E. Lanyon and W. N. Tavolga, ecls. Animal sounds and communication.
Amer. Inst. Biol. Sci., Pub. No. 7. Washington, D.C. 4 4 3 pp.
868 T H E AMERICAN NATURAT,IST
. 1963. A spectrographic analysis of the vocal repertoire of the African village weaverbird. Condor 6 5 : 517-527.
L)a~,win, C. 1873. I':xprcssions of thc cmotions in man and animals. Appleton, New York.
Dilger, W. C. 1960. Agon~stic and social behavior of captive rcdpolls. Wilson Bull. 7 2 : 115-132.
Eisenbcrg, J . F. 1963. The behavior of hetcromyid rodents. University California Publ. Zool.
6 9 : 1-102.
. - ..-- . 1974. The function anti motivational basis of hystricomorph vocalizations. Symp.
Zool. Soc. Lonclon 3 4 : 211-247.
Isisenberg, J. k'., L. R. Collins and C. Wcmmer. 1975. Communicationin the Tasman~an Devil (Snrcophilus hcrrrisii) ancl a survey of auditory communication in the Marsupialia.
Z. Ticrpsychol. 3 7 : 379-399.
IiCiscnbcrg, J . F., and R. E. T<uchn. 1966. The bchavior of Ateles geoffroyi and related spccics. Smithsonian Misc. Collections 1 5 1 ( 8 ) : 1-63.
Emlcn, S. T. 197%. An experimental analysis of the parameters of bird song eliciting spccles recognition. nchavior 41 : 130-171.
Palls, ,J. B., and J . R. Iircbs. 1975. Sequcncc of songs In repertoires of western meadowlarks (Sturnelltr neglectrr). Can. J. Zool. 5 3 : 1165-1178.
Fickcn, &I. S. 1962. Agonistic behavior and territory in the American reclstart. Auk 7 9 : 607-632.
Grecnewalt, C:. 11. 1968. Birr1 song: acoustics and physiology. Smithsonian, Washington, D.C. 194 pp.
Hardy, J . W. 1071. Habitat and habits of the dwarf jay, Aphelocoma n n n a . Wilson Bull.
8 3 : 5-30.
Ileinz, G. I*., and 1,. W. Gyscl. 1970. Vocalization bchavior of the recl-necked pheasant.
Auk 8 7 : 279-295.
.rchl, J. R., J r . 1973. Breeding biology and systematic relationships of the stilt sandpiper.
Wilson Bull. 8 5 : 115-147.
Jerison, H . J . 1973. I':volution of the brain and intelligence. Academic Press, New Yorlr.
482 pp.
Johnston, R. F., and J . W. Hardy. 1962. Behavior of the purple martin. Wilson Bull. 7 4 : 243 262.
Johnston, R. F., D. M. Niles, and S. A. Rohmer. 1972. Hcrmon Bumpus ant1 natural selection in the house sparrow, Passer domestic?^^. Evolution 26 : 20-31.
Kaufmann, .T. 11. 1962. I':cology and social behavior of the coati, N a s u a nuricu, on Barro Colorado Islanti, Panama. University California Puh. Zool. 6 0 : 95-222.
lileirnan, D . G. 1972. Social behavior of the maned wolf (Chrysocyon b ~ t c c h y u n r s ) ancl bush clog (Speothos v e n a t i c u s ) : a stucly in contrast. J. Mammal. 5 3 : 701-806.
Lorenz, Ti. 1943. Die anbeborenen Formen miiglicher Erfahrung. Z. Tierpsychol. 5 : 235-409.
Marler, P . 1056. The voice of the chaffinch ancl its function as a language. Ibis 08:231-261.
. 1967. Animal communication signals. Science 1 5 7 : 760-774.
Marshall, J . T. 1964. Voice in comm~inication and relationships among brown towhees.
Condor 66 : 345-356.
Martin, D. J. 1973. A spectrographic analysis of burrowing owl vocalizations. Auk 9 0 : 564-578.
Morse, 1~). 1974. Niche breaclth as a function of social dominance. Arner. Natur. 1 0 8 : 818-830.
Morton, E. S. 1071. Food ancl migration habits of the eastern kingbird in Panama. Auk 8 8 : 925-926.
-- . 1975. l3cological sources of selection on avian sounds. Arner. Natur. 1 0 8 : 17-34.
Moynihan, M. H . 1962. Uisplay patterns of tropical American "nine-primariecl" songbirds.
11. Sorne species of Rtcmphocelus. Auk 7 9 : 655-686.
-- . 1964. Sorne behavior patterns of platyrrhine monkeys. 1. The night monkey ( A o l u s t r i t ~ i r g a t u s ) . Smithsonian Misc. Collections 1 4 6 ( 6 ) : 1-84.
Mueller, H. C. 1971. Displays a n d vocalizations of t h e sparrow hawk. Wilson Bull. 83:
249-254.
Nero, H. W . 1963. Comparative behavior of t h e yellow-heaclecl blackbird, red-wingecl blackbird, a n d other lcterids. Wilson Bull. 75 : 376-413.
Oring, L. W. 1968. Vocalizations of t h e green a n d solitary sandpipers. Wilson Bull. 80:
395-420.
Petrinovich, I,., ancl H. V. S. Peeke. 1973. Habituation t o territorial song in t h e white- crowned sparrow (Zonotrichitc leucophrys). Behav. Riol. 8 : 743-748.
Power, D. M. 1966. Agonistic behavior a n d vocalizations of Orange-chinned paralteets in captivity. Condor 68 : 562-581.
Root, R . B. 1969. The behavior ancl reproductive success of t h e blue-gray gnatcatcher.
Conclor 7 1 : 16&3 1.
Rowell, T. E., ancl R . A. Hincle. 1962. Vocal communication b y t h e rhesus monkey. F'roc.
Zool. Soc. London 138: 27!)-294.
Samuel, D. E . 1971. Vocal repertoires of sympatric barn ancl clif'l'swallows. Auk 88: 839-855.
Schaller, G. B . 1964. Rreecling behavior of the white pelican a t Yellowstone Tdalre, Wyoming.
Conclor 66 : 3-23.
Schott, D. 1975. Quantitative analysis of t h e vocal repertoire of squirrel rnonkeys (Sairniri sci?rreus). Z. Tierpsychol. 38: 225-250.
Shalter, M . 1975. Tdaclc of spatial generalization in habituation tests of fowl. J. Cornp.
F'hysiol. F'sychol. 89: 258-262.
Smith, S. M. 1975. I n n a t e recognition of coral snake pattern b y a possible avian predator.
Science 187 : 759-760.
Smith, S. T . 1972. Cornmunication ancl other social behavior in P a r u s cnrolirrerrsis. P u b . Nuttall Omithol. Club, no. 11. 125 pp.
Smith, W . J. 1966. Communication ancl relationships i n t h e genus Tymnnus. Pub. Nuttall Ornithol. Club, no. 6. 250 pp.
Smythe, N. 1970. Ecology ancl behavior of t h e agouti (Bnsyproctrr p?rnctntc~) ancl related species on Darro Colorado Islancl, P a n a m a . Ph.U. thesis. University of 14larylanJ.
130 pp.
Stokes, A. W . 1967. Behavior of t h e bobwhite, Colinus virfinintzus. Auk 84: 1-33.
Ternbrocli, G. 1968. 1,ancI mammals. Pages 338-404 i?? 1'. Webeok, ed. Animal commnni- cation. Indiana University Press, Bloomington. 686 pp.
Thoreson, A. C. 1964. The breeding behavior of t h e Cassin aultlet. Conclor 66:456-476.
Van Valen, L. 1973. Body size ancl nurnbers of plants a n d animals. Evolution 27:27-35.
Wemmer, C. 1977. The comparative ethology of t h e large spotted genet, C:etaetttc tigrinu, ancl relatecl viverricl genera. Smithsonian Contrib. Zool. (in press).
Willis, E . 0 . 1969. On t h e behavior of five specles of Rhegmutorl~iva, ant-following antbircls of t h e Arnazon Basin. Wilson Bull. 81 : 363-395.
Willis, E . O., a n d Y. Oniki. 1972. Ecology ancl nesting behavior of t h e chestnut-backed antbird (iblyrmeciyn ezsul). Condor 74: 87-98.
