Z. Tierpsychol., 48,345-348 (1978) @ 1978 Verlag Paul Parey, Berlin und Hamburg ISSN 0044-3573 / ASTM-Coden: Z E T I A G
From the Max-Planck-Institut f u r Verhaltensphysiologie, Seewiesen
A Special Constraint on the Evolution of Composite Signals By WOLFGANG WICKLER Received: April 16, 1978 Revised and accepted: August lfi,1978
Abstract and Summary Many animals use composite signals, combining e. g. auditory and visual components. If the components travel at different speeds, the time lag between reception of the components increases with increasing distance from the sender. Problems arising especially from rhythmical repetition of composite signals are discussed and possible solutions are indicated.
Students of animal communication usually treat optical, acoustic, chemical, etc. signals separately, not only because of the different physical properties involved, but also because different sense organs act as receivers and different techniques of examination are required. However, many animals simultaneously use different signal channels in one communicatory performance. For primates, the “overwhelming importance of composite signals” has been pointed out by MARLER(1965). Humans make extensive use of the combination of auditory and visual components, here for brevity called AV-signals. This may be the reason why such AV-signals are especially conspicuous to us in mammals and birds. In this report only free combinations will be considered, i.e. combinations where no necessary causal connection exists between the components, e.g. between a movement and the sound produced by it (as in drumming woodpeckers or gorillas). It seems legitimate to assume that a signal will evolve only if selected by the recipient’s responses. This is reflected in semiotic theory (MORRIS1946), which at the “pragmatic” level of analysis defines the meaning of a signal with 1966), responses which the respect to the responses of its recipient(s) (CHERRY communicators intend to elicit by providing the signal. The response, then, indicates the significance attached to a given signal by the recipient; attaching significance to a signal has been termed “semantization” by WICKLER (1967). All kinds of signals, therefore, have to be adjusted by the sender to (and to be judged by us from) the receiver’s standpoint. Whether, in the case of a composite signal, a component or a combination of components should appropriately be termed a signal depends on whether the components can be shown to have particular message contents to potential receivers. U S.
Copyright Clearance Center Code Statement:
0044-3573/78/4804-0345 $02.5010
346
WOLFGANG WICKLER
Conspicuous movements accompany the vocalizations of many duetting birds. Field studies led us to conclude that in some species this combination ensures a differential influence on two types of receivers: only the acoustic component reaches territorial neighbours; the nearby mate in addition receives an optical component, which modifies the effect of the acoustic one (SEIBTand WICKLER1977). The range of the optical component is often but not always limited by dense vegetation. From studies of cases without this limitation a general problem arose which we would like to discuss here, although as yet we do not know how many of the implications outlined actually apply to the example chosen for demonstration. While optical and acoustic signals each have their own specific advantages, a combination of the two suffers through the different speeds of light and sound: reception of an optical signal will be practically simultaneous at all relevant distances from the sender, while for an acoustic signal the distance between sender and receiver (here called s-r) becomes an important aspect due to the relatively slow transmission of sound. For simplicity we assume that the receivers remain stationary during communication. If a receiver responds best or most to an acoustic component well synchronized with an optical one, e.g. by preferentially reacting to the Ons and Offs of a (composite) signal, there will be no problem for the sender as long as s-r is close to zero. But the time lag between reception of the two coniponents will increase by 3 ms per m s-r. A sender knowing the exact s-r could in theory adjust to it and begin the acoustic component a corresponding fraction of time in advance of the optical one. If he does not know the exact s-r, or if he wishes to address to several receivers at different distances simultaneously, synchronous arrival of both components cannot be assured. The problem becomes more pronounced if the signal is repeated rhythmically. If a series of identical visual signals (Vl, V2, V, . . .) are emitted a t regular intervals of t ms, each well synchronized with an auditory component (A,, A,, As . . .), the original synchronized AIVl-signal is perceived close to the sender, and pseudo-synchronized versions (AIV2, A2V3, or AIV3, AtV4, etc) occur a t integer multiples of t.0.333 m from the sender. In between, the A- and V-components are desynchronized to different degrees. The sender thus provides a synchronized and a non-synchronized version of its AV-signal. Which version is picked up by the recipient depends entirely on where he happens to be. (We will not speculate about the possible attenuation of both components over distance.) There are two ways for the sender to overcome this zonation of his broadcasting: I. he may change s-r erratically by moving around during signalling. Some animals in fact behave like this, though for still unknown reasons; others, including the bird species mentioned below, remain stationary during their displays. I n this case, 2., the sender must avoid a stable time relationship between the A- and V-component. If both require rhythmic repetition, the two rhythms should be independent. In fact it would suffice that they appear independent to the recipient, even if there were some Relative Coordination effect (VON HOLST 1939) in the sender. The result of 1. and 2. would be a random mixture of synchronized and non-synchronized AV, uniformly dispersed around the sender. Any recipient would be provided with the same mixture of the synchronized and the non-synchronized version, regardless of his location. This might be still more important for duetting birds with sex-specific duet components: if signalling were rhythmical and synchronized, a visual
A Special Constraint on the Evolution of Composite Signals
347
signal of, say, the 9 would be received synchronously with a female vocalization at some distances and with a male vocalization at others. This might give rise to misinterpretations which should be avoided. We wonder whether this helps to explain a phenomenon which otherwise we find difficult to understand: Mated pairs of several duetting bird species - Trachyphonus usambiro (Capitonidae), Halcyon chelicuti (Alcedinidae), Tockus hemprichii (Bucerotidae) - produce rhythmical vocal elements accompanied by conspicuous rhythmical movements; Trachyphonus wags its raised tail sideways like a windscreen wiper, Halcyon suddenly opens and closes its wings, Tockus bows and stretched, head and legs. In their natural habitat all three species perform these displays from tree tops or protruding branches, so that both A- and V-components are received at large distances. Contrary to the observer’s intuitive expectation, the marked optical and acoustic rhythms d o not develop a stable relationship or come into phase. (For humans it is extremely difficult to simultaneously perform two actions with different rhythms.) T o deliberately prevent AV-synchronization would make sense if in fact repeated perception of synchronized A- and V-elements would result in a special semantization by the recipient. More detailed data - from ALBRICHTand WICKLEK (1968), WICKLER and UHRIC (1969) - from the duet display of Trachyphonus usambiro (formerly included in T.d’arnaudii) are mentioned here merely to increase plausibility of the suggested functional relationships, not to prove them. The Q gives t w o short notes, “A-a”, 115 ms apart; another 115 ms after the beginning of “a” the 8 utters a harsh call “M’, during which the 9 gives a call “B”, which starts rather precisely 80 ms after the beginning of M. The whole sequence (A-a- M - - A-a. . .) is repeated over several min, identical elements recurring about every 0.65 s. The 9’s wagging tail needs about 0.32 s to complete one left-right-left cycle and thus could beat in synchrony with the 9’s A and B. This does not happen, however; in fact the tail rhythm appears to be deliberately broken by slower movements and brief standstills, so that it does not achieve synchrony with any of the vocal elements, which keep rhythm and precise 8-9 timing throughout the duet. If the tail movement were synchronized with the 7’s B, it would appear synchronized with the 8’s M at an s-r of 2 6 m and with the 9 ’ s “a” a t an s-r of 6 5 m , etc. Duet communication between pairs is frequently observed over 50 m o r more.
Following this reasoning one would expect what MARLER(1965) actually found in primates (and which is true for humans as well), that “visual or auditory signals are used for communication over larger distances”, while “composite signals are a special feature of close-range communication”. Not, however, because subtle properties (such as dynamic facial expressions) “would be lost if perceived from a distance”; available knowledge makes it very likely that many birds have sufficient visual acuity to easily recognize wing or tail flicks at distances of 50 m and more. The more likely reason is that visual signals cannot be kept bound to their sound counterparts if the composite signal travels over such distances. T o counteract the separation of signal components the elements could, of course, be stretched in time; if a long sound is emitted together with a long optical signal the two components still overlap to a large degree when perceived from far away. This could work with recipients more sensitive to multiple sensory input than to On/Off; but it would drastically reduce the amount of information transmitted per unit time, making it more efficient to choose high information density within one sensory modality. chemical Similar reasoning applies to other composite signals (optical etc), except where the different components travel at the same speed. Such electrical, acoustic vibratory, gustatory exceptions would be e.g. optical
+
+
+
348
W. WICKLER, A Special Constraint on the Evolution of Composite Signals
-t- olfactory signals. The latter two combinations have been shown to occur, and typically it proves difficult to separate the contributions of each modality to the communication process. Zusammenfassung D a sich Licht und Schall verschieden schnell ausbreiten, ergeben sich Probleme fur zusammengesetzte optisch-akustische Signale, besonders wenn sie rhythmisch wiederholt werden. Synchron gesendete Bild- und Ton-Elemente desynchronisieren sich mit wachsender Entfernung vom Sender oder pseudosynchronisieren sich mit anderen Elementen. Um das Signal fur jeden (ortsfest gedachten) Empfanger in beliebiger Entfernung einheitlich zu halten, mui3 der Sender entweder, wenn er synchron sendet, dauernd seinen O r t wechseln, oder, wenn er ortsfest bleibt, die beiden Rhythmen unabhangig halten. Mogliche Beispiele fur letzteres werden angefuhrt. Dasselbe gilt fur andere Modalitaten-Mischungen. Fur Fernkommunikation sollten deshalb entweder Signale nur einer Modalitat verwendet werden oder Mischungen aus Komponenten gleicher Ausbreitungsgeschwindigkeit. I am most grateful to Miss C. RECHTEN for crit,ical comments on the manuscript.
Literature cited ALBRECHT, H., and W. WICKLER (1968) : Freilandbeobachtungen zur ,,Begruhngszeremonie" des Schmuckbartvogels Trachyphonus d'arnaudii. J. Orn. 109,255-263. CHERRY, C. (1966): O n human communication. 2nd ed. MIT Press, Cambridge, Mass. HOLST, E. VON (1939): Die relative Koordination als Phanomen und als Methode zentralnervoser Funktionsanalyse. Erg. Physiol. 42, 228-306. MARLER, P. (1965): Communication in monkeys and apes. In: Primate Behavior. (DE VORE,I., ed.) Rinehart and Winston, New York, 544-584 MORRIS,C. W. (1946): Signs, language and behavior. Prentice Hall, New York. SEIBT.U.. and W. WICKLER (1977): Duettieren als Revier-Anzeiee 2. Tier" bei Vogeln. " psychol. 43,'180-187. WICKLER.W. (1967) : Verdeichende Verhaltensforschuna und Phyloaenetik. In: Die Evolution der Organismen. ( H E B ~ R EG., R , ed.) Bd. I, 420--508.-G. Fische;, Syuttgart WICKLER, W., and D. UHRIG (1969): Bettelrufe, Antwortszeit und Rassenunterschiede im Begrufiungsduett des Schmudcbartvogels Tracbyphonus d'arnaudii. 2. Tierpsychol. 26, 651-661. \
,
Author's address: W. WICKLER, Max-Plank-Institut fur Verhaltensphysiologie, D-8131 Seewiesen, W-Germany.
From the Max-Planck-Institut f u r Verhaltensphysiologie, Seewiesen
A Special Constraint on the Evolution of Composite Signals By WOLFGANG WICKLER Received: April 16, 1978 Revised and accepted: August lfi,1978
Abstract and Summary Many animals use composite signals, combining e. g. auditory and visual components. If the components travel at different speeds, the time lag between reception of the components increases with increasing distance from the sender. Problems arising especially from rhythmical repetition of composite signals are discussed and possible solutions are indicated.
Students of animal communication usually treat optical, acoustic, chemical, etc. signals separately, not only because of the different physical properties involved, but also because different sense organs act as receivers and different techniques of examination are required. However, many animals simultaneously use different signal channels in one communicatory performance. For primates, the “overwhelming importance of composite signals” has been pointed out by MARLER(1965). Humans make extensive use of the combination of auditory and visual components, here for brevity called AV-signals. This may be the reason why such AV-signals are especially conspicuous to us in mammals and birds. In this report only free combinations will be considered, i.e. combinations where no necessary causal connection exists between the components, e.g. between a movement and the sound produced by it (as in drumming woodpeckers or gorillas). It seems legitimate to assume that a signal will evolve only if selected by the recipient’s responses. This is reflected in semiotic theory (MORRIS1946), which at the “pragmatic” level of analysis defines the meaning of a signal with 1966), responses which the respect to the responses of its recipient(s) (CHERRY communicators intend to elicit by providing the signal. The response, then, indicates the significance attached to a given signal by the recipient; attaching significance to a signal has been termed “semantization” by WICKLER (1967). All kinds of signals, therefore, have to be adjusted by the sender to (and to be judged by us from) the receiver’s standpoint. Whether, in the case of a composite signal, a component or a combination of components should appropriately be termed a signal depends on whether the components can be shown to have particular message contents to potential receivers. U S.
Copyright Clearance Center Code Statement:
0044-3573/78/4804-0345 $02.5010
346
WOLFGANG WICKLER
Conspicuous movements accompany the vocalizations of many duetting birds. Field studies led us to conclude that in some species this combination ensures a differential influence on two types of receivers: only the acoustic component reaches territorial neighbours; the nearby mate in addition receives an optical component, which modifies the effect of the acoustic one (SEIBTand WICKLER1977). The range of the optical component is often but not always limited by dense vegetation. From studies of cases without this limitation a general problem arose which we would like to discuss here, although as yet we do not know how many of the implications outlined actually apply to the example chosen for demonstration. While optical and acoustic signals each have their own specific advantages, a combination of the two suffers through the different speeds of light and sound: reception of an optical signal will be practically simultaneous at all relevant distances from the sender, while for an acoustic signal the distance between sender and receiver (here called s-r) becomes an important aspect due to the relatively slow transmission of sound. For simplicity we assume that the receivers remain stationary during communication. If a receiver responds best or most to an acoustic component well synchronized with an optical one, e.g. by preferentially reacting to the Ons and Offs of a (composite) signal, there will be no problem for the sender as long as s-r is close to zero. But the time lag between reception of the two coniponents will increase by 3 ms per m s-r. A sender knowing the exact s-r could in theory adjust to it and begin the acoustic component a corresponding fraction of time in advance of the optical one. If he does not know the exact s-r, or if he wishes to address to several receivers at different distances simultaneously, synchronous arrival of both components cannot be assured. The problem becomes more pronounced if the signal is repeated rhythmically. If a series of identical visual signals (Vl, V2, V, . . .) are emitted a t regular intervals of t ms, each well synchronized with an auditory component (A,, A,, As . . .), the original synchronized AIVl-signal is perceived close to the sender, and pseudo-synchronized versions (AIV2, A2V3, or AIV3, AtV4, etc) occur a t integer multiples of t.0.333 m from the sender. In between, the A- and V-components are desynchronized to different degrees. The sender thus provides a synchronized and a non-synchronized version of its AV-signal. Which version is picked up by the recipient depends entirely on where he happens to be. (We will not speculate about the possible attenuation of both components over distance.) There are two ways for the sender to overcome this zonation of his broadcasting: I. he may change s-r erratically by moving around during signalling. Some animals in fact behave like this, though for still unknown reasons; others, including the bird species mentioned below, remain stationary during their displays. I n this case, 2., the sender must avoid a stable time relationship between the A- and V-component. If both require rhythmic repetition, the two rhythms should be independent. In fact it would suffice that they appear independent to the recipient, even if there were some Relative Coordination effect (VON HOLST 1939) in the sender. The result of 1. and 2. would be a random mixture of synchronized and non-synchronized AV, uniformly dispersed around the sender. Any recipient would be provided with the same mixture of the synchronized and the non-synchronized version, regardless of his location. This might be still more important for duetting birds with sex-specific duet components: if signalling were rhythmical and synchronized, a visual
A Special Constraint on the Evolution of Composite Signals
347
signal of, say, the 9 would be received synchronously with a female vocalization at some distances and with a male vocalization at others. This might give rise to misinterpretations which should be avoided. We wonder whether this helps to explain a phenomenon which otherwise we find difficult to understand: Mated pairs of several duetting bird species - Trachyphonus usambiro (Capitonidae), Halcyon chelicuti (Alcedinidae), Tockus hemprichii (Bucerotidae) - produce rhythmical vocal elements accompanied by conspicuous rhythmical movements; Trachyphonus wags its raised tail sideways like a windscreen wiper, Halcyon suddenly opens and closes its wings, Tockus bows and stretched, head and legs. In their natural habitat all three species perform these displays from tree tops or protruding branches, so that both A- and V-components are received at large distances. Contrary to the observer’s intuitive expectation, the marked optical and acoustic rhythms d o not develop a stable relationship or come into phase. (For humans it is extremely difficult to simultaneously perform two actions with different rhythms.) T o deliberately prevent AV-synchronization would make sense if in fact repeated perception of synchronized A- and V-elements would result in a special semantization by the recipient. More detailed data - from ALBRICHTand WICKLEK (1968), WICKLER and UHRIC (1969) - from the duet display of Trachyphonus usambiro (formerly included in T.d’arnaudii) are mentioned here merely to increase plausibility of the suggested functional relationships, not to prove them. The Q gives t w o short notes, “A-a”, 115 ms apart; another 115 ms after the beginning of “a” the 8 utters a harsh call “M’, during which the 9 gives a call “B”, which starts rather precisely 80 ms after the beginning of M. The whole sequence (A-a- M - - A-a. . .) is repeated over several min, identical elements recurring about every 0.65 s. The 9’s wagging tail needs about 0.32 s to complete one left-right-left cycle and thus could beat in synchrony with the 9’s A and B. This does not happen, however; in fact the tail rhythm appears to be deliberately broken by slower movements and brief standstills, so that it does not achieve synchrony with any of the vocal elements, which keep rhythm and precise 8-9 timing throughout the duet. If the tail movement were synchronized with the 7’s B, it would appear synchronized with the 8’s M at an s-r of 2 6 m and with the 9 ’ s “a” a t an s-r of 6 5 m , etc. Duet communication between pairs is frequently observed over 50 m o r more.
Following this reasoning one would expect what MARLER(1965) actually found in primates (and which is true for humans as well), that “visual or auditory signals are used for communication over larger distances”, while “composite signals are a special feature of close-range communication”. Not, however, because subtle properties (such as dynamic facial expressions) “would be lost if perceived from a distance”; available knowledge makes it very likely that many birds have sufficient visual acuity to easily recognize wing or tail flicks at distances of 50 m and more. The more likely reason is that visual signals cannot be kept bound to their sound counterparts if the composite signal travels over such distances. T o counteract the separation of signal components the elements could, of course, be stretched in time; if a long sound is emitted together with a long optical signal the two components still overlap to a large degree when perceived from far away. This could work with recipients more sensitive to multiple sensory input than to On/Off; but it would drastically reduce the amount of information transmitted per unit time, making it more efficient to choose high information density within one sensory modality. chemical Similar reasoning applies to other composite signals (optical etc), except where the different components travel at the same speed. Such electrical, acoustic vibratory, gustatory exceptions would be e.g. optical
+
+
+
348
W. WICKLER, A Special Constraint on the Evolution of Composite Signals
-t- olfactory signals. The latter two combinations have been shown to occur, and typically it proves difficult to separate the contributions of each modality to the communication process. Zusammenfassung D a sich Licht und Schall verschieden schnell ausbreiten, ergeben sich Probleme fur zusammengesetzte optisch-akustische Signale, besonders wenn sie rhythmisch wiederholt werden. Synchron gesendete Bild- und Ton-Elemente desynchronisieren sich mit wachsender Entfernung vom Sender oder pseudosynchronisieren sich mit anderen Elementen. Um das Signal fur jeden (ortsfest gedachten) Empfanger in beliebiger Entfernung einheitlich zu halten, mui3 der Sender entweder, wenn er synchron sendet, dauernd seinen O r t wechseln, oder, wenn er ortsfest bleibt, die beiden Rhythmen unabhangig halten. Mogliche Beispiele fur letzteres werden angefuhrt. Dasselbe gilt fur andere Modalitaten-Mischungen. Fur Fernkommunikation sollten deshalb entweder Signale nur einer Modalitat verwendet werden oder Mischungen aus Komponenten gleicher Ausbreitungsgeschwindigkeit. I am most grateful to Miss C. RECHTEN for crit,ical comments on the manuscript.
Literature cited ALBRECHT, H., and W. WICKLER (1968) : Freilandbeobachtungen zur ,,Begruhngszeremonie" des Schmuckbartvogels Trachyphonus d'arnaudii. J. Orn. 109,255-263. CHERRY, C. (1966): O n human communication. 2nd ed. MIT Press, Cambridge, Mass. HOLST, E. VON (1939): Die relative Koordination als Phanomen und als Methode zentralnervoser Funktionsanalyse. Erg. Physiol. 42, 228-306. MARLER, P. (1965): Communication in monkeys and apes. In: Primate Behavior. (DE VORE,I., ed.) Rinehart and Winston, New York, 544-584 MORRIS,C. W. (1946): Signs, language and behavior. Prentice Hall, New York. SEIBT.U.. and W. WICKLER (1977): Duettieren als Revier-Anzeiee 2. Tier" bei Vogeln. " psychol. 43,'180-187. WICKLER.W. (1967) : Verdeichende Verhaltensforschuna und Phyloaenetik. In: Die Evolution der Organismen. ( H E B ~ R EG., R , ed.) Bd. I, 420--508.-G. Fische;, Syuttgart WICKLER, W., and D. UHRIG (1969): Bettelrufe, Antwortszeit und Rassenunterschiede im Begrufiungsduett des Schmudcbartvogels Tracbyphonus d'arnaudii. 2. Tierpsychol. 26, 651-661. \
,
Author's address: W. WICKLER, Max-Plank-Institut fur Verhaltensphysiologie, D-8131 Seewiesen, W-Germany.
