Bird Song at the Edge of Music and Science: An Interview with Peter Marler

By David Rothenberg

Print Friendly, PDF & Email

About Neurobiologist and Ethologist Peter Marler

Peter Marler
Peter Marler.
Photo by Bethany Daniels, courtesy College of Biological Sciences, UC Davis.
Peter Marler, who spent a lifetime studying the call of the wild—first the calls of birds and later of primates—and applying the calls to neuroscience, died on July 5, 2014 at the age of 86.

Marler, who held emeritus status in the Department of Neurobiology, Physiology and Behavior at the University of California Davis, conducted groundbreaking research in using songbirds as a model for learning language. Passionate about birdsong since he was a boy growing up in England, Marler noticed that some songbirds show local “dialects.” His research showed that birds raised in the lab learned their songs from other birds; the songs were not innate to the birds from birth. This was the only example in the animal kingdom, other than humans, whales and possibly bats, of cultural transmission of communication. He found that birds show genetic constraints in how they approach their song development—some learn faster, developing a larger repertoire than others, some are more open to improvising, and others remain faithful to their tutors. Using playbacks of tape-recorded calls in the field of vervet monkeys, Marler also helped show how the monkeys had distinct alarm calls for different kinds of predators, such as snakes or leopards. And he conducted the first spectrographic analysis of the vocal repertoire of wild chimpanzees.

After previous positions at UC Berkeley and Rockefeller University in New York, Marler joined the faculty at UC Davis in 1989 and helped establish the Center for Neuroscience. He retired in 1994. He was a fellow of the National Academy of Sciences, American Academy of Arts and Sciences, and the Royal Society, the British equivalent of the National Academy of Sciences.

Bio courtesy University of California Davis.

Introduction by David Rothenberg

In 2004, in the course of doing research for my book Why Birds Sing, I had the opportunity to interview Peter Marler, possibly the greatest bird song scientist of our time. Nearly everyone in the field worked with him at some point in their studies. He was one of the last surviving links to the time when scientists first started visualizing animal sounds with sonograms, a technique invented by the military near the end of World War II. With his passing this year at the age of 86, I thought it was time for the whole interview to be published in full. If you want to know more about any of the names mentioned here, check the pages of Why Birds Sing or Nature’s Music, edited by Peter Marler.



David Rothenberg: Do you think bird song has anything to do with music?

Nature's MusicPeter Marler: There was once a conference in Washington on the biology of music, and there was a certain kind of anti-science mood, I think. I found little common ground with what I could glean from it, which is not a necessity, but I was looking for something. I’m uncertain there could be more common ground. The only hope I’ve sensed anywhere is what I’ve learned from Olivier Messiaen. It seems to me that there are little borrowed fragments of song incorporated in various compositions, and that to me seems to be a cheap solution, whereas Messiaen seems to have steeped himself in the structure of birds singing, not just little fragments here and there. I have a sense that in some of his compositions you have a sense of the rhythm of birds singing, even though the detailed structure is not very lifelike. There are all these problems of key, tonality, and timing that we cannot match. We can’t even hear them. He was aware he was putting the song through a radical transformation. The reiteration of quite complex phrases with progressive rhythmic modulation from rendition to rendition. Even though the result doesn’t sound like any bird particularly by the time it’s in instrumental form.

David Rothenberg: Nearly all birdsong scientists say they are amateurs, not attuned to human music. 

Peter Marler: It may very well be that there are all kinds of obstacles, all kinds of formalisms there.

David Rothenberg: Why do people love the hermit thrush song?

Peter Marler: As is clear from lots of your essays, there’s a huge contribution to the aesthetic from the ambience, the mood of the whole situation. I must admit that for me, when you hear a bird’s song, you relive the whole circumstance of the song, the evening, the reverberation.

When I was at Cornell a few years ago, I began to get interested in this music topic, and I searched through their archives for recordings with musical qualities, and I found that they favor the parabolic recordings with more pure bird songs, but in a strange way, it’s somewhat degraded, because the reverberations are missing. Regular microphones produce recordings that offer a much richer aural experience.

David Rothenberg:
But what about the Morton ranging hypothesis?

Peter Marler: Of course ranging is only one very limited aspect of song’s function. It’s visible because it’s something you can test. How much are you ready to rely on intution. That’s my dilemma when it comes to the aesthetic side of things. Once you get committed to behaving like a rigorous scientist, which I got converted to when I was quite young, I found myself mentally dividing life into a limited subset of things that were accessible to scientific investigation, and the rest.

David Rothenberg: How did you get converted?

Peter Marler: I’ve never thought about it as ontogenetic. The only way I can try to reconstruct it is by thinking of different phases in my own history. I really began as an amateur birdwatcher, back in high school. One pivotal event which was nothing to do with science was an extraordinary meeting that was held in the Zoological Society of London. I went to hear the great bird recordist Ludwig Koch, and Max Nicholson. Nicholson was a very interesting person, a classic British civil servant, very senior, an absolutely committed naturalist, was especially interested in bird song. I don’t know if he was somehow instrumental in getting Koch to leave Germany, but they became close friends. Nicholson arranged this meeting, and here was Ludwig Koch presenting his work, a strong Austro-German accent, very endearing, I was absolutely thrilled. He was especially interested in chaffinch song, and somehow that was an a ha experience for me. I’m not sure if I had already begun transcribing chaffinch song, but by then or soon after I developed this primitive script, long before sonograms or tape recorders. Very primitive—amazing really how ill-equipped we were, and how rapidly that whole situation was transformed, first wire recorders, then tape recorders. I sort of lived through it.

I spent two or three years working on bird song on my own. I was a botanist professionally at the time, in the late 40s. I did my first bachelor’s degree in 1945, then I became a graduate student in botany, I worked on the chemistry of mud in the Lake District, but I tracked chaffinch dialects around the lakes in my spare time. Another significant experience was attending David Lack’s short Christmas conference for student who were interested in the biology of birds. They were quite extraordinary experiences for me. Lack had already pointed out the special interest on birdsongs on islands, where you have a depopulated fauna, he talked about the Azores. I made two trips there pretending to be a botanist, but I was more interested in birds. This Lack experience was important for introducing some kind of rigor into my enterprise.

The second time, Nikolaas Tinbergen had just moved to join the faculty at Oxford, and several of the students there became my lifelong friends and colleagues.

Chaffinch. View spectrograph and hear song (video).
Photo by Charlesjsharp, cc.

David Rothenberg: Did you find you had to listen carefully, learn to discriminate carefully, since you didn’t have technology?

Peter Marler: I, unlike Luis Baptista, have never had a very good ear. I have always had some trouble recovering an acoustic image in my head. The verbal stuff is important to recreating it, even though I’m sure my ear was infinitely more sensitive then. It’s one thing to hear the most subtle differences, and quite another thing to preserve that image so you can carry it to some other context and time. However, one of the charming things about bird songs is that if you work at it, you can develop either mental machinery or visual aids for making that transfer because of in a large part the tonality of bird songs makes them accessible.

There is a kind of proto-musical element there, which you only realize if you move your attention to, say, monkeys, who have such noisy sounds, and it’s much harder to engage with them. I think the most important single element is that the sounds happen in a narrow frequency range, it’s easy for our brains and our technology to process. Monkey vocalizations are extraordinarily difficult to process by ear.

David Rothenberg: Is there any clear evolutionary reason why bird sounds are accessible to us?

Peter Marler: One of the commonest vehicles for privacy in acoustic channels is pitch. Take my old student, Doug Nelson, he did a very nice analysis of a whole community of birds, different species, all living at our field station in New York, he got recordings of all of their songs and reasoned that there must be some reason to be sonically distinct from each other in the same community. He asked which feature is the most efficient in classifying the songs to species, when he had like 15 different features. Frequency came out very clearly at the top. He then selected one bird, the field sparrow, and tested the ranking of different features that his statistical analysis had yielded by creating families of artificial modified songs, and the quickest way to turn a song into something alien is to change the frequency.

It’s an idea that only makes sense if you look at it in the larger context of what acoustic neighbors are singing.

David Rothenberg: You have so many bird songs that have the same kind of structural elements that you find in music. Is there some universality in that if you hear a pattern of sounds that is clearly a bird song, are there certain universal patterns of structure that all bird songs have?

Peter Marler: Have you looked at work on the relationship between song structure and the sound transmission qualities of the environment? There’s definitely something there. There was a French ornithologist who was comparing open country and forest, and Eugene Morton, before the ranging hypothesis, his thesis was making that kind of comparision which we then picked up again, he never carried through with it technically, we refined the whole thing a bit more and found some new principles, but did not significantly harm the main hypothesis, if you’re trying to signal someone in a forest there are very specific problems, there’s an emphasis on tonal songs with sustained patterning in the forest, wider bandwith, temporal patterning out in the open, which makes some logical sense in terms of degradation, that’s already muddying the water a bit about generalizing about the musicality of birdsong.

David Rothenberg: Now did you first work on calls, the chaffinch as language?

Peter Marler: Actually I was focuses on the song before that. You have those accidents of personal history. When I went to Cambridge I was a graduate student in William Thorpe’s lab, I wasn’t completely free to focus on song development, he was already doing this, so there are these subtle pressures, but I was determined to focus on the chaffinch, so I was diverted to a field study on a limited population of marked birds in a very nice oak woodland, so I sort of milked song as a theme fairly quickly, then I gradually became immersed in all the other things the birds were saying, that’s how I became familiar with their entire vocal repertoire.

For Thorpe’s experiments we were also raising chaffinches by hand, while he worked on the song, their calls were also developing. Very few people had even tried to track the development of bird calls before this. That got me interested in the whole topic of repertoires, signal repertoires, and how they change in the course of evolution, and the obvious comparisons to be made are with other finches, so I began looking at other finches and comparing their call repertoires, I found some very interesting things that I was never able to pursue because I got the job in Berkeley, I left all those European finches behind.

I was just looking the other day at this strange book, Darwin’s Biological Work, this was before I took the job in Berkeley I was approached by Peter Bell, one of my botany teachers, who was organizing this book as a sort of Darwinian celebration, one of the figures in this article is a rather unique map comparing the calls in different finches, and then a developmental sequence of the calls. 

David Rothenberg: Did you discover the parallel hawk-flying-overhead calls shared by several species of British birds?

Field sparrow
Field sparrow.
Photo by PookieFugglestein, cc.

Peter Marler: Well, I don’t think anyone had ever thought about it before. It is a bit stupid in retrospect to call chaffinch calls a language, but I was young!

David Rothenberg: The calls, with specific meanings, are innate, while the songs are learned. This is sort of counterintuitive.

Peter Marler: Well actually, some calls are learned, but they are a minority. Your generalization does stand. The diversity is just prodigious, this is an interesting scientific challenge, you have to pretend when you present a generalization, that things are much clearer than they really are. A lot of people are uncomfortable with doing science in this fashion. Especially people who are mathematically minded and are more likeley to favor a hypothetical-deductive kind of reasoning. There are people who are never comfortable unless they get into that mode. They’re seriously uncomfortable if they are required to work in this inductive mode of trying to reach some conclusion from surprises that you observe. Some students simply could not work in a disciplined fashion inductively.

Ethology was always very much an intuitive pursuit. That didn’t work very well as it spread. So much depended on the personal history and mental assiduousness of the people pursuing it. It could be very productive for the person of the right mindset and background, for better or worse, over a period of 20 or 30 years, there was a growing dissatisfaction with this methodology. Then there was this quite radical shift when John Krebs and Maynard Smith realized the potential power of applying economic theory to animal behavior. That was really the birth of this other approach, very much hypothetico-deductive.

David Rothenberg: Game theory? Is every animal really a strategist?

Peter Marler: Well it has its own unique power as a methodology because as long as you accept a set of premises and think hard about them, there are inevitable predictions that can be made. It is potentially a very powerful way of approaching, but at the same time it is in its way quite reductionistic, because you focus on some particular set of behaviors, you begin asking questions about functional efficiency, and you remain very much focused on that piece of behavior. You answer questions that fit into the general method. You don’t just go out and record what you find. You don’t go in as a sponge.

David Rothenberg: But when you went in to study chaffinch language, you went in as a sponge. How did technology change this? Your early sonograms were a pointillistic Seurat approach, I’ve seen that nowhere else.

Peter Marler: The original spectrograph produced messy images that were very hard to reproduce, so I came up with this method of tracing them. Later came the more Japanese ink-painting filled-in approach. This was worked out specifically with the Oxford group of scientific ornithologists. Reg Moreau was editor of Ibis, he could not handle sonograms, so we had to work out a way to present them in the middle 50s.

David Rothenberg: Today’s computer sonograms are less beautiful than the ink sonograms.

Peter Marler: Well eventually you became skilled at generating them, so you can make them more beautiful. You choose a bandwidth for the analysis, and having made that choice, you bias the way in which the information will be portrayed.

The sonograph was originally created to help the deaf to learn. The idea was that the deaf could see how sound should look, and you could see how it should be done. It absolutely never worked. Sonograms of speech are very difficult to read. No one can do it in real time with any precision.

Peter Marler in the field
Peter Marler in the field.
Photo by Ingbert Grüttner, courtesy The Rockefeller University.

David Rothenberg: Did people take notice that you were calling bird calls language?

Peter Marler: I don’t think anyone paid all that much attention to it. I don’t know how as a student you register reaction to your work. I suppose I’m rather naïve about it all, it’s not like Watson and Crick figuring out the genetic code, with blazing headlines everywhere. In some ways it was a rather lonely enterprise, it was a new thing to do, it had not been possible until that time to begin thinking about such issues in anything other than intuitive terms which had already become the targets of skepticism and distrust. It didn’t trouble me, I was so enthusiastic it didn’t bother me in the slightest. But it was not immediately contagious! Especially working on the calls, it’s quite a daunting task because with song you can get yourself a tape recorder, go out and get good material in a few days, but to obtain a reliable estimate of the call repertoire, you have to really immerse yourself and follow the birds around. Seasonal changes, sexual dimorphism. It’s not an accident that there are still relatively few complete estimates of call repertoires even now. There’s a lot of nice work on chickadees now, that’s probably one of the best known. The best estimates are birds that I’ve been involved with. Chaffinches. Chickens!

David Rothenberg: What I’d really like to see is more detailed data than what actually gets published. People say they have them, but they don’t present them. Why is there so little analysis of long, complex songs?

Peter Marler: Starling songs, yes, they are at the very limits of human comprehensibility. Nearly a minute in length, all these subthemes that are not completely stable, all these complicated decisions about when to go from one type to the next. If you’re doing quantitative song type analysis you have to be relatively sure another researcher would see the same structure. My goodness, it was challenging to generate data from recordings of starling song.

With a bird like a mockingbird, you’re hard put to be sure if you’ve represented an entire repertoire. You’d have to do cumulative asymptotic plots.

David Rothenberg: Well, in a way, those plots tell you very little.

Peter Marler: Well, they’re very dumb, but they’re very important. It’s one way of explaining why others should trust your data.

David Rothenberg: Mockingbirds move from one sound to another, morphing; no one has written about that except Dowsett-Lemaire.

Peter Marler: Well, you know it’s very deceptive. What you’re saying was in my mind in the early days, but it hasn’t been pursued. We had no idea how a thrush song was structured. After looking at 50 consecutive songs the bird was repeating itself, reusing the same units in different ways. That was a revelation, it was brand new.

There was one student, Peter Merritt, working on mockingbird song, one day we took over a large laboratory in Florida, and we laid out end to end about 500 sonograms of consecutive mockingbird songs of one bird, recorded over a period of 40 minutes, and early on, you had the repeated impression that the bird was morphing in a more or less continuous fashion, but as you looked at more and more songs, those morphs were actually being treated as separate units, because they would appear again identically, but in a different acoustic context. But we didn’t find that.

David Rothenberg: I think you could say that they are improvising in the sense of putting together what they have previously learned in unique, novel ways.

Peter Marler: But I still think the song has become crystallized in some sense. The whole ordering them is in itself very complicated. I must show you some of my very old papers, using interval distribution, trying to decide what the units of a song should be. You get a multimodal distribution, and you then have an objective basis for distinguishing intervals between and within song syllables, and intervals between song. These distributions are almost unoverlapping.

Once you’ve derived some estimate of how long the bird has to pause before you can assume it is changing songs, you start to look at the structure of the song, and you find those units recurring. The whole unit question is absolutely fundamental to song analysis, and a lot more complicated than most people think.

David Rothenberg: I can understand how hermit thrush songs are separate, but with mockingbirds?

Peter Marler: There will be short pauses if you look closely, and you can ascertain where the separation between songs happens. That’s what we did in those early papers on the mistle thrush “20 basic themes in 47 songs”

Photo: Andreas Eichler, CC BY-SA 3.0, wikimedia.

David Rothenberg: Why are there so few papers like Sotavalta’s on the nightingale, where the structure of complex songs are really analyzed?

Peter Marler: Wallace Craig, also, was an absolute pioneer. Extraordinary guy, that wood pewee paper is quite different than anything else he had ever done. Really a 200-page book on a four-note bird song. The other things he did were quite revolutionary, in a completely different way. May I digress for a minute? You know, Otis Whitman wrote these huge tomes on the biology of pigeons and doves. Wallace Craig was a student of his. He wrote some extraordinarily profound very general essays on principles of behavioral organization, about aggression for the The International Journal of Ethics. Craig stated that you have two ways to break down the universe of behaviors: those that are endogenously generated, appetative, something the animal decides to do for internal reasons, and reactive behaviors that don’t occur until the animal encounters a specific situation. Where does aggression fall? He asserted very strongly that aggression is a reactive behavior, animals don’t go out and seek opportunities to aggress, it is primarily a reaction, triggered by social stimuli. This was a complete contradiction to Konrad Lorenz’s argument, Lorenz said there are recurring endogenous needs to be aggressive.

Lorenz was interested in Craig for other reasons. Craig anticipated much of what Lorenz was thinking about.

David Rothenberg: The wood pewee song certainly has something musical about it.

Peter Marler: Yet the few people who have looked at that aspect of song have tended to get very formalistic. There was a Canadian mathematician, I’ll have to look it up. Some of it became quite technical. The nuggets that I came away with from it were that if you look at long sequences, recordings of singing in a bird with a finite repertoire, and you’re asking, does the bird have a program for proceeding through this repertoire, at first blush, it seems unlikely because there are so many deviations from any fixed order you might propose. But if you begin to think of it in terms of the underlying brain mechanism, if there were a program, it would require something like a clock, which means when you’re looking at this sequence of songs, you better take into account what’s happening in real time, when certain pauses are longer than others. If you then pose the program question again, and restrict it to cases when you’re looking at a succession of closely spaced songs, you find that it is orderly, there is a program. When it begins to deviate, there are larger pauses. Something has happened, the clock has ticked.

David Rothenberg:
I still think that to generate order out of mockingbird song would be easier: teacher teacher teacher to witchety witchety witchety. What kind of connection is recognized?

Peter Marler: If I react to what you’ve said as a scientist I’d say, “fascinating hypothesis, let’s test it.” Because of course when you’re listening, when you hear two imitations juxtaposed, and they happen to have similar properties, something registers. Boom! The fact that the two imitations harmonize in some sense may be true, but maybe it’s just something that you recognize.

David Rothenberg: But even the very idea that they should be random is a different kind of preconceived notion….

Peter Marler: Agreed! I had a student who proposed that kind of question with a varied thrush, a bird with a fairly limited repertoire, a rather simple song. He wanted to pose the question you’re asking. Are they sung in random sequence or is there some underlying rule? He recorded a lot of them , and got some support for the notion that having sung song A, in choosing B, there has to be more than some minimal contrast. You see what I mean? It’s certainly nonrandom, that’s the way science tends to work. You get a hypothesis, you look for some way to address the question in a straightforward way, and you veer toward the simple end of the continuum. If you were to attempt that with a mockingbird, you’d have to grapple with an enormously difficult unit problem.

David Rothenberg: There’s enough to examine in looking at how the phrases are put together. Go back to Sotavalta, there’s something here others have ignored….

Peter Marler: Well does it satisfy the requirements for rigor and reproduceability. The starling song has been understood to some degree, and now people are starting to use them as a neurobiological preparation, and very often they gloss over the detailed song structure. It’s not uncommon for people in laboratories to just measure the duration of songs, that becomes the dependent variable. This does such violence to all of the information that exists about the fine structure of these songs. People fiddle with the brain and are not very interested in the behavior. It’s a slippery slope there, and I do battle with people there sometime. I’m particularly sensitive about starling song. I suspect that the yield from some of the neurobiology would be much richer if they took the trouble to consider it. However, they’d have to put someone on that question full time; you can’t just do it on the side.

I’ve succumbed to the same temptation I think: the white-crowned sparrow song is terribly simple as songs go, and yet we spent a lot of time with it, just because it is manageable. You go on about mockingbirds, the fact is, there has never been a really full-blown study of what they are up to!

David Rothenberg:
Jeff Baylis suggested that up to 80 percent of mockingbird songs might be from extinct species!

Peter Marler: Well, that’s ridiculous. If he meant instead birds experienced long and back in the past of the singer’s experience, that would be credible. There’s certainly a lot of people who are skeptical that a major part of a mockingbird’s repertoire is based on personal experience. You cannot underestimate the capacity of a mockingbird to invent sounds, to create sounds, especially if it has been deprived of the ability to imitate sounds. Now there hasn’t been a really convincing study of this, now there’s an old paper by Miller who raised some mockingbirds in the early days and he was impressed by the number of songs that they created, as I remember. Talk about understudied abilities, the whole phenomenon of the inventive potential of birds is understudied, very difficult to study, and I’m quite convinced that the possibilities are remarkable if you could only document them convincingly.

David Rothenberg: What kind of possibilities?

Peter Marler: You have this interesting tradeoff between the richness of auditory experience and the richness of vocal output. These aren’t necessarily the same. It’s as if a species as a threshold of input beyond which it is unable to hear any more. But if it doesn’t hear enough, it may become even more inventive. I’m not talking about mimics, but birds who learn from their peers. If you shrink the set of models the bird hears, he will have to do something else.

Professionally I’ve tended to shy away from professional mimics, except as isolated examples of some extraordinary acoustic feat.

Now Dowsett-Lemaire’s marsh warbler project is a remarkable story. I don’t know of any other account that has quite the same charm and breadth, the idea of being able to reconstruct the pathway of migration by the imitations that recur in the song is quite remarkable. You need to be as intimate with the songs of all these African birds as you need to be with all these European birds. Now she’s quite hesitant about some of her identifications, see that’s the kind of thing that brings my scientist’s hat out—how sure are you that none of the marsh warbler’s song is its own, beyond imitation.

Sonogram of Eastern Phoebe.
Sonograms of the Eastern phoebe.
Image courtesy Peter Marler, birdsong lecture (read lecture).

You can recount the various theories of mimicry, even though none has been proven. You don’t want to dismiss the very idea of these theories.

Now on nightingales, Henrike Hultsch in particular has a profound understanding of nightingales and nightingale song. She’s the more intuitive of the two, while Dietmar Todt is a formalist and modeller. He builds these little interpretive structural models which I usually don’t warm to, but some people like them. They form a perfect research pair. He’s just retired and she became director of an institute for child development studies in Berlin, although she keeps up the nightingale work. But unfortunately, funding cuts have hit Germany quite hard. You should try to visit them. Henrike’s quite remarkable, she’s written a chapter in my upcoming book, Nature’s Music.

David Rothenberg: Yes, it’s going to be the bird song book of this generation.

Peter Marler: Don Kroodsma’s chapter is one of the most amazing. We invited him to take a comparative overview. He’s the only person in the world who could have done it, he has an amazingly broad knowledge, much of it personal. But he’s almost getting to the borderline (don’t tell him I said this) of scientific credibility. Another of the scientific challenges is to somehow strike a balance between particulars and generalities. Like Jane Goodall, for example, when you ask her a question about chimpanzee behavior she’ll give you a series of amazing anecdotes about animals. She’s not able to stand back and overlook some of that detail, even contradict it in a minor way in order to get a generalization out of it.

David Rothenberg: Well, we never like to generalize that way about human beings.

Peter Marler: An outsider could do it, but not a professional! Don is getting so immersed in the detail that he loses the ability to generalize. He’s been a pivotal person in the development of the field.

To be reviewed in the New York Review of Books would be a climactic thing for me. Every issue I devour! It’s the leading place for popular intellectual discourse.

David Rothenberg: How do you feel about the shrinking number of behavioral experiments in biology?

Peter Marler: Well that’s really too bad, and the audience of new students I hope to be the main audience for Nature’s Music. These people sometimes rarely read anything outside the zebra finch literature. People are getting farther and farther from behavior, and it doesn’t so much matter in this generation but it will in the next generation. Things will become intensely reductionistic. The nature of the field will change.

David Rothenberg: Is anyone going against the grain?

Peter Marler: Well, yes somewhat. The Cold Spring Harbor people have pointed out that it is a small step to move away from the zebra finch and pick other birds with a larger repertoire. It’s a small step in the behavioral direction, but you know as soon as soon as there’s a decent genetic map for the zebra finch, that’s going to be all but irrestible for all kinds of developmental questions. It’s too bad that it has some peculiarities that are a serious disadvantage from the behavioral point of view. The whole song acquisition process is so rapid you can’t always tell the acquisition phase from the production phase. It’s going to be fascinating to see what comes of the Ofer Tchernichovski methodology. I must say I am quite cautious about it. Like when people first started collecting video data of birds learning. It came to almost nothing because we were overwhelmed by information. That’s my worry with Tchernichovski. He’s an engineer really, a computer nut, not really knowledgeable about behavior. If he collaborates with people who have behavioral skills, it could be quite fruitful.

David Rothenberg: Do you remember the moment when the sonogram first appeared?

Peter Marler: Oh it was clear from the moment that we opened up the package that this would be something important. In 1949 there was only one other sonograph in Britain, at the Naval Laboratory, where they were using it for what it was created for, underwater surveillance. Thorpe was lucky enough to get one, it was clear that it almost had magical potential. We were still in a very primitive state when it came to creating our own recordings. Thorpe, in his wisdom, had arranged to get a complete copy of the BBC bird song library, which was huge, going all the way back to Ludwig Koch. I was immediately able to start feeding all these songs into the machine. I was intoxicated by the potential to the point of kind of skipping from one thing to another rather superficially I’m afraid, which was rather different from what happened here.

There was a parallel development at Ohio State. Donald Borror must have been working at the naval laboratory during the war, he knew about the genesis of the sonograph while it was still under wraps.

Visible speech illustration
Page 6 of On the Nature and Uses of Visible Speech.
Graphic courtesy Library of Congress.

David Rothenberg: So it was a secret World War II thing?

Peter Marler: Have you looked at Visible Speech? If any of those people survive, it would be interesting to talk to them. I think what happened was somehow at the end of the war a semi-classified project was set up at Ohio State, something to do with the emerging interest in sound broadcasting through the substrate, the whole thing in Wisconsin where they had these very powerful sound generators, but it didn’t work. Borror acquired the machine by default, and he immediately saw its potential. There was a linguist at the University of Chicago, Martin Joos. And the great ecologist W.C. Ali was interested in the interface between ecology and social behavior, he had a student, Nick Collias, another of the pioneers, in the early 1950s published a paper on the vocalizations of chickens, which was a pioneering event.

At first people thought the oscilloscope would be useful for bird songs, but frequency information is hard to illustrate there. The machine turned out to be more useful for insect songs. So for a little while the insect people were ahead of the bird people, because they had their machine first. That’s why René Busnel’s book Acoustic Behaviour of Animals has so little about birds. The whole issue of species specificity of sound was first displayed in crickets. But after we got the sonograph we sort of overtook them.

David Rothenberg: Tell me about the secretary in W.H. Thorpe’s lab at Cambridge, Joan Hall-Craggs.

Peter Marler: She first entered my field of you when she convinced me that there was a blackbird in her garden who listened to her piano playing and he changed key when she did. She made recordings and one of my first jobs was to analyze these recordings to see if she was correct. It was a terrible job. The final answer was that the intervals are not consistent, as you might expect, it doesn’t match up, as Messiaen says, he tells you that. There are individual birds with individual melodies that conform, but not in any consistent harmonic ways.

David Rothenberg: But she wrote a series of papers on the musical aspects of the blackbird song evolving through the season, and no one else seems to have done anything like that.

Peter Marler: Yes, she was another lonely soul. She suffered in a way because she was not entirely respected, she was not a scientist. She did not have all that much data, and she made very rich generalizations from the limited material that she did have.

It’s only quite late in my career that I’ve come to appreciate what W.H. Thorpe did for me. At the time there was a kind of competititveness between the two of us. He didn’t show me that his book was in press until it came out. His interests were wide, moving to science and religion, philosophy. He may be citing me in his book but the work was done in his laboratory. He was quite fair in that sense.

David Rothenberg: What did he do and what did you do?

Peter Marler: You know, all his entomological work is quite fascinating and still of current interest. The whole idea that insects might parasitic insects might imprint on cues from their hosts. As a consequence, then they seek out the same host when they’re ready to oviposit, kind of an elementary learning process. He got into that because in the war he was working on biological insect control and he was a pioneer in that kind of work, there’s that whole issue of learning in insects, it’s now of great current interest. He was also a very aggressive advocate for ethology, he championed Lorenz and Tinbergen and he gave a set of lectures at Harvard in 1949 or something the target of which was to champion European ethology although it always had his own distinctive flavor. He was extraordinarily generous to me, he just sort of let me loose although there was always this side pressure that song work was his prerogative, that’s how I got on to the calls.

David Rothenberg: What did he think about his work with Hall-Craggs?

Peter Marler: I honestly don’t know, I don’t think he was ready to champion it aggressively. But he didn’t have any more students after me, so he may have found her work attractive and interesting, I don’t know. You know he had another big diversion studying duetting birds.

David Rothenberg: In his Bou Bou shrike work he does the rare thing of comparing interval distribution in Bach, Webern, folk music, and bou bou duets!

Peter Marler: Another colleague who was stimulating to Thorpe was a South African guy who did a lot of recordings of South African birds. I think that might have been what got him started on the bou bou shrikes.

David Rothenberg: Only in Webern is the augmented fourth, a very dissonant interval, extremely popular.

Peter Marler: Thorpe wasn’t that knowledgeable about music, so that was probably Hall-Craggs’ influence.

Southern boubou drawing
Male and female Southern bou bous.
Illustration by Claude Gibney Finc-Davies (May 24, 1875 – August 4, 1920).

David Rothenberg: Do you believe that bou bou shrikes sing a victory duet when they succeed in chasing intruders away?

Peter Marler: I’m not sure I find that plausible, I’m trying to think of parallel examples. There is a goose call that Lorenz labelled “triumfgeschrei” I have never felt motivated to really immerse myself in that duetting literature, I don’t know why. Partly because I have something of a conviction that behavior close to duetting is extremely widespread in birds, and what you’re dealing with there is a special derivative of something that’s much more general: countersinging and countercalling between individuals, probably involving general recognition, this is probably just the perverse scientist in me raising his ugly head again, OK, you spend years studying duetting in these shrikes and where does it lead you? What’s the next big thing? My bias has always been to favor themes where you could see future prospects, the illumination of other domains, which leads one to sidestep extreme peculiarities which are sort of a dead end in a scientific sense.

David Rothenberg: Like lyrebirds? How much do you want to investigate something that’s so crazy.

Peter Marler: Yes, although years ago someone gave me some lyrebird song tapes and asked me to analyze them. They wanted to have a folio of sonograms of this entire tape, which they could then inspect, listening to the recordings themselves, filling in the raw material. I would have provided a service function. I would have done it, but I never got around to it. As things involved, the whole equipment story changed several times quite radically, driven largely by military research and once real time analyzers became available, that opened up a whole new vista, you didn’t have to work in these little two and a half chunks anymore.

There were some interesting encounters in the 1960s when the technology was just evolving, you sensed that you were approaching the boundaries of what the visiting engineer was permitted to tell us. Of course eventually it was all desclassified. It happened to us several times, there was a whole profession of people working on cryptosurveillance engineering. A lot of the gear was miniaturized, we spent a long time trying to develop wireless microphones, I’ve always thought for years and years that it would be an absolute revelation if you could obtain recordings of a complete day’s output of a singing bird. You couldn’t follow the bird around, you have to rig up a bird with a wireless setup. It still hasn’t been done because of the problem of miniaturizing things. We did manage to rig up chickens, I have an unpublished manuscript about that. It must have been in the mid-80s.

Through this research we began talking to these surveillance people, they are a strange lot. Much of their work was barely legal. Sorry, I’m getting punchy here.

David Rothenberg: I tend to find glimmers of people interested in what I’m interested in, but the mainstream is elsewhere.

Peter Marler: I think that is an insight that should be pressed. I think it is a very fertile train of thought. Whether it can ever break through the barriers into hard science I’m not sure, but that’s probably not something you should worry about too much.

David Rothenberg: I do think it’s part of the phenomenon that hasn’t been considered enough in a scientific way. 

Peter Marler: You see Thorpe was an artist at dividing his life into separate domains, and I suspect that’s what he was doing then. He saw an aesthetic dimension, which is fascinating, but then he leaves it. My gut intuition for what it’s worth is that you might be able to use some of the insights from the compositional material of Messiaen and then persuade someone to re-examine birdsong from the compositional point of view, then you might establish a foothold on the scientific side and reveal something novel and exciting, and reopen questions as to whether there’s a natural basis for music or whether it’s purely cultural.

David Rothenberg: I think there are plenty of examples of the song itself that can be examined to have a real musical-type structure. The function does not explain this structure.

Peter Marler: I don’t think anyone halfway intelligent would think of it as random walking. You know about Charles Hartshorne, right? The philosopher who wrote a book on bird song. I was lucky enough to sit by him on an ornithological field trip, on the bus.

David Rothenberg: He picked up on these same things, saying: look at all these musical qualities in bird songs.

Peter Marler: You know, that’s not a very powerful scientific insight, because you have so much diversity to choose from, so you could confirm almost every imaginable hypothesis.

David Rothenberg: Isn’t the very fact of that diversity something important?

Peter Marler: I’ve always thought that his simple interpretation of continuity, if you sing too much, your listener gets bored. You have to introduce diversity. How are you going to do it? If there’s any aesthetic primordial in the bird’s brain, it would become manifest.

Varied sonograms.
Song sparrow sonograms.
Image courtesy Peter Marler, birdsong lecture (read lecture).

David Rothenberg: Like Kant says in the Critique of Judgment: if a human musician were to play a single note over and over again, we would get bored, but if a bird does it, we do not get bored, showing that birdsong is somehow outside of music, some essential purity beyond….

Peter Marler: That’s very interesting. I must say I’m absolutely ready to believe that there are common principles, it would be amazing actually if there were not some underlying principles, but whether they converge at all with human aesthetic principles, that would be something else.

David Rothenberg: What did people in different centuries hear when they heard theses birds singing? Did they hear them as more or less musical in different ways?

 Peter Marler: Well there must be ethnomusicologists who have opinions on that.

David Rothenberg: Certainly music has been obsessed with birds for centuries.

Peter Marler: Well in the Western tradition that hasn’t been especially helpful…. You’re sort of coming full circle with the concept of beauty: music is becoming contorted by music as conducive for a trance state, therefore you favor it because of what it does to you.

So there is some dischord there on the kinds of criteria for what counts as beautiful in either bird or human music. I just discovered that old Daines Barrington had a rather different ranking of which bird songs were the most beautiful, with some surprising omissions, no European thrushes at all. He was most familiar with captive birds, and European thrushes don’t sing well in the cage, although the shama thrush does.

My impression is that top-flight bird fancyers in Germany would immediately name the shama as the outstanding performer, but that may be a rather small population with a very specialized experience.

I don’t know if you’ve stumbled on the couple of papers by this German ornithologist Tretzel who found two cases of dialects among shamas that were picked up from human sounds and then transmitted from one bird to another. We visited his home and he had pet shamas there singing their heads off. He loved them more than he did his children, I dare say. He was a veteran from the war….

Generally though, the rankings, there’s a lot of coincidence among different rankings, as long as people are familiar with songs from different parts of the world

David Rothenberg: I’m fascinated by how much is written about nightingales in England, and actually when I first heard one, traveling from the States, it was definitely not what I expected.

Peter Marler: Well, it’s a classic case of ambience. I once heard them at one end of Berlin, the bridge where people from the other side of the Wall would come over, a nightingale singing his head off on the bridge. A strange place for him to be, but I can still remember the moment. It must have been 15 years ago, the Wall was still up. It must have been the first visit I made to see Dietmar Todt and Henrike Hultsch. Makes my hair prickle on my neck just to think about it.

I don’t know if even I could generalize to other birds and other circumstances, certainly with the more potent songs you have that sense of remembering the moment, becoming aware of other circumstances that you might otherwise overlook. You relive the particular circumstances. You must be in a contemplative state, but it’s interesting that it happens at all.

David Rothenberg: I’m interested in how learning more about a bird song changes the way you hear it. Right after reading papers on the song of the starling I heard it differently.

Peter Marler: At that level of what you’re able to perceive, the consequence is immediate, that’s why I do believe Hartshorne. His ear wasn’t all that good, in many cases he thought that a bird had a limited repertoire, but in fact it was much greater than he realized. On the other hand, there are cases where even a simple song is immediately appreciated, like I did with the chaffinch.

You know, there is the experiment that Steve Nowicki did, exploiting his discovery that if you put a bird in helium, that if you take recordings of helium air songs, and tutor birds with them in a regular environment, they very often restored the correct tonality, the helium air emphasized higher harmonics, when they imitated these songs, very often those harmonics were deleted again. There’s this whole little literature that I’ve never been able to understand fully about whether birds have absolute pitch, mainly based on the guy in the psychology department at Johns Hopkins, and his student Jeff Cynx, and he wrote a box for our book on this subject. He thinks he has it all explained, but I’m not sure.

If you play a bird back a song he recognizes at a different pitch, he often fails to recognize it. Usually this kind of experiment is done on starlings.

David Rothenberg:
Why in some birds does song “improve” after the functions of territory defense and mating are finished?

Peter Marler: Well certainly this idea has been floating around for a while: if song is released from rather strict functional constraints you sometimes see a blossoming of melody and so forth. That’s just one aspect of the larger issue of inventiveness: Why would they ever embark on this in the first place? They’re driven, that’s something you can’t deny about song, it is a behavior subject to strong internal motivation, it’s like walking, something a child has to do, it becomes impaired if it’s prevented, it’s a driven activity. So what’s the nature of that motivation? Is it something like an emotion, something with subjective connotations to it, which is something we tend to assume: a bird is being joyful? This may or may not be true, that’s the puzzle that’s interested me, it’s obviously, to me, analagous to an emotional display. It’s not triggered by something outside, it can be modulated by things outside, but that’s not the primary force, so it’s not in any way analagous to an alarm call.

That’s the kind of criteria that’s often used to distinguish an emotional display from language. Where does birdsong fit in the spectrum of emotional states? It’s not strictly sexual obviously, for it’s relatively rare for a bird to sing as an accompaniment to copulation. They do sing when they’re fighting.

David Rothenberg: The Catchpole catch: Why is there one song for both mating and fighting?

Peter Marler: Ask Kroodsma about this, he’s spent more time than anybody listening to American warblers, who have two song types. What he tells me is that he thinks of both of them as affective displays, but with a different flavor. Sometimes there is a compound of aggression and sex, which becomes separate in two songs, not in any sense a label for a sexual object or an aggressive target is boiling up inside.

David Rothenberg: Francoise Lemaire told me that as soon as the female appears, the male stops singing. She may never get to hear what he’s capable of.

Peter Marler: Especially with that huge repertoire.

David Rothenberg: She also says they sing for fun, and none of her colleagues seemed interested in that.

Peter Marler: I don’t know she has any proof of how the songs are acquired. Makes a nice story, it doesn’t really matter, but they do indeed map their migration route in their song. They only imitate the songs of birds heard en route.

David Rothenberg: How much more can be learned with birds in the lab as opposed to in the wild?

Peter Marler: We had an elaborate experiment in Millbrook, disguising redwings as orioles, it’s mostly about the epaulets. Would redwings who learned oriole songs switch to that song? It wasn’t very conclusive, they’re all very spooky birds, they’d never really settle down and be comfortable in the laboratory.

David Rothenberg: Is that true of many bird species?

Brown-headed cowbird
Brown-headed cowbird.
Photo by Bear golden retriever, cc.

Peter Marler: Well, I’ve never really thought about it in a systematic way. Cowbirds are perfectly tameable. That’s another of the stupid things that influences your subjects for experimentation. I don’t really know what it means. Quails never become personally tame, and yet chickens and guinea fowl become almost like pets. I suppose it has something to do with the impact of early experience. I never pursued it any further. It’s always been one of the great inspirations in the song work, getting to know skilled aviculturalists, they have so much valuable experience.

David Rothenberg: Do birds perceive sound at a faster rate than we do? Is that why we often hear more when they are slowed down?

Peter Marler: Well, it’s an interesting issue. Pomfrey, Cambridge professor of comparative physiology, developed several arguments that birds are much better than we are at temporal discrimination, that became part of the lore, you know, you take a complicated song that goes by so fast we can hardly hear what’s going on. Then people began to do proper psychoacoustic tests, and they didn’t find much appreciable difference between birds and us in time discrimination. It’s only recently that Bob Dooling has reopened this issue, see his chapter in the book. Birds are much better than people at time discrimination completely separated from frequency.

Al Lieberman determined that there’s rather a small set of human phonemes, maybe just 16, out of which all sounds in all human languages can be produced. Most remarkable, but we mustn’t digress.

David Rothenberg: Why the emphasis on counting song types rather than connecting them?

Peter Marler: I think you’re asking for more than the field is yet ready to provide. Many people are still in the state of being slightly bemused by the complexity of their undertaking. How many song types does a song sparrow have? What do you do with the variations of song type A? The fact is there are variations at both levels. To emphasize one at the expense of another is just not good science. People will tend to set aside within category variations if they’re lumpers, not splitters.

David Rothenberg: I want them to look at the quality and make some generalization about them. So many of these papers just count the song types. These kinds of papers on complex bird songs all look the same.

Peter Marler: You’re looking for order at higher level. I think it’s understandable, you bring a new set of questions that will require you to reexamine the issues.

David Rothenberg: Are birds intricate machines?

Peter Marler: I’ve always relied on this trick of separating the world into domains. Some are subject to scientific investigations, others are not. [Note from David Rothenberg: This is the same thing Marler says Thorpe was able to do, he probably learned it from him.] To apply scientific thinking to them is not only unproductive but potentially harmful, the ultimate sort of thing you argue as a graduate student, suppose you could somehow develop scientific understanding of some of the mythical miracles that have supposed to have taken place—what would you gain?

Think of the nine classical emotions. Paul Ekman agrees the classification is circular. It originated with some understanding of the human face, linked with internal experience. This became the pedestal upon which understanding of emotions was based. No one questioned in. Then there were numerous experiments. Sheckman. Subjects were injected with epinephrine and asked, “do you experience a particular emotion here?” It all turned out to be context-dependent. Now if you turn out to people who are doing the basic brain experiments on emotion, you find them admitting that they have some sense of the system that underlies avoidance and escape behavior, but they have no idea how to find physiological correlates for other classical emotions, even something as basic as aggression.

And it’s even harder to identify such things in birds, especially with the absence of our familiar markers, which almost removes the temptation to do so. But I’m no authority, I kind of gave up work in this area, as a nonspecialist, I could hardly keep up with the literature.

But the original dogma was that all bird sounds, even calls, were expressions of primitive emotional states. Sebeok went so far as to argue that alarm call behavior meant that there were hawk emotions, leopard emotions, and snake emotions. He clung so strongly to the emotional view that he was ready to fragment his notions of emotions into a very radical revisionism. But of course he’s always been defensive about animal studies of anything remotely language-oriented. He clings to the uniqueness of man.

David Rothenberg:
I’d like to hear more about where you think this field is going, and what you feel the consequences are of so much emphasis on studying the brain rather than behavior.

Peter Marler: My first response is there’s no question in my mind that the neurobiological work is here to stay. Vocal learning is the entry into a general study of learning. So very much more is known about the bird brain in this area than any other. It’s amazing how little is known about the human brain from this point of view. It’s very difficult to do this kind of intrusive experimentation on humans, and you depend so much on pathological findings, Broca’s area, Werneger’s area, etc., almost certainly as imaging techniques are refined there will be a blossoming of a new kind of study of human brain activity during speech behavior of various kinds.

There is a certain threshold of new intriguing discoveries, there seem to be certain neurotransmitters that seem to be especially involved when there is neural plasticity, and these seem to be involved all across verterbrate biology. There’s also a growing appreciation of the value of having a behavior as complex as song as the readout for any manipulation or set of records of neural activity that have been made in different parts of the song system, that’s one of the amazing things that’s happening now, people realizing for the first time that issues of timing in a complex song may be processed quite separately from pitch and tonality.

David Rothenberg: Are they suggesting that they are especially separate in birds?

Peter Marler: In a practical sense, everyone thinks of the song system in birds as an ideal model system for pursuing some of these fundamental questions, without thinking for a moment that it could be translated to a mammalian brain. But you assume that it is generalizable, but only implicitly.

David Rothenberg: Did you ever want to go in this direction with your research?

Peter Marler: No, I’m too clumsy manually for surgery. But I encouraged my students to go in this direction.

David Rothenberg: So most people doing this work have to be expert with the knife?

Peter Marler: Yes, that was one of Masakazu Konishi’s magical skills. Right from the beginning he was quite amazing in creating the necessary tools and using them. When he was a beginning student, he wanted to try some surgery but he didn’t have the right scalpels. He took a light bulb, broke it, and removed some of the wires inside to create a blade and a hook to pull out the cochlea of a bird from the inner ear. I used to wonder privately whether there wasn’t something from growing up in a society where calligraphy is part of daily life that led to his ability in this. Nowadays you can tell a male from a female bird you catch in a net by sending the blood off for analysis, but in those days, you had to cut a tiny slit and look at the gonads inside. Konishi, when no one was watching, used to do it without any anaesthetic, even though you’re supposed to use a topical when you do such minor surgeries.

David Rothenberg: Is that something people are concerned about, the treatment of the animals?

Peter Marler: No one as far as I am aware is deeply concerned about this, partly because there has never been a really fierce attack on anyone doing this kind of procedure, as there has been with monkeys and even cats.

David Rothenberg:
Nevertheless, the field talks about “sacrifice” rather than killing birds.

Peter Marler: Oh, that was probably an aspect of grantsmanship. I think it’s just a circumlocution, you’re slightly embarrassed when you say you’ve killed the animal. People wince if they hear you using such words. You know the whole animal rights thing pervaded all comparative research to some degree, if only because NIH became so paranoid about it, creating rules for experiments in which any animal is used.

There are tremendous individual differences in how people feel about this. The American response is this is another obstacle to be overcome, as long as you yourself behave responsibly in terms of animal pain and discomfort, then you have no reason to worry. But there other people who have quit after a time. There’s a great deal of pressure, and it fed back on your funding prospects. There’d be someone on the review committee assigned to scrutinize all aspects of animal care in the experiment proposal. I know several projects that were simply terminated, the most tragic example I know of was at Rockefeller in New York, where there was a guy at the Museum of Natural History, Lester Aronson, he did very interesting work on the sexual behavior of cats…. He was able to show that if the penis is de-enevated, then the mating behavior of the male cat is transformed into new patterns. Very clever, thoughtful work, all derived ultimately from Frank Beech. Aronson was challenged by the animal rights people and it ultimately destroyed his personal life. They systematically telephoned him in the middle of the night, that kind of intimidation. There’s much less of that kind of activism going on here these days, although it could still happen in Britain.

Scientists have become more sensible and canny in terms of how they describe their work. Experiments have become much more proficient at defending themselves. David Hubel spent a couple of years campaigning on the subject and defending people who were doing work on animals. These kind of protests ritard the progress of science! By isolating a bird from all of his natural social forces, you oversimplify his life, but the critics don’t understand how huge the advantages are. You isolate one variable in a situation and then you are able to study it. When I am criticized about this it annoys me because I did some of the first work on the role of social interactions in such matters.

The criticisms tend to come from people who work in the field, rather than in the laboratory. Michael Beecher, University of Washington. He works on song sparrows, and he has made some of the more well known critiques of all of my lab work on song learning. Take a young song sparrow, raise him in the laboratory, give it a very rich array of songs to listen to during the sensitive period, and one of many things we found was that they learn a great deal, and one interesting behavior that you often see is that they take a pair of tutor songs and parse them down to syllables and phrases, and then recombine them in different ways to produce a new song.

Beecher works with wild song sparrows, and he says this happens only very rarely, suggesting that it is some artifact of captivity. Wild birds tend to learn complete songs from their tutors and keep them intact. He almost cast ridicule upon me! But now he was working in Washington. It turns out that song sparrows in Washington are very different than those in New York. My colleague Susan Peters found out that this fragmentation actually does occur in Eastern birds, but if not in Western ones. You’d be surprised how many people in field ecology resist any attempt to intervene with what they observe—you will change the very things you are studying, almost a kind of religious constraint! And yet much more has come from lab studies than from field studies, because they are so much easier to control. But we have learned important things in the field, such as the extraordinary potency of song matching in the wild. Matching is rarely friendly, it’s confrontational.

You know there’s quite a lot of literature on the indigo bunting, the work of Robert Payne, who also worked on the African indigo bird. Young birds tend to imitate the song of a successful male in any given community. There are a set of available tutors, they choose a male who is successful. I am quite convinced that this is another case where the process of song crystallization is delayed into even the second year. It is not novel acquisition as much as a selective bias winnowing out unnecessary syllables. If this correct, then what you need to look at is the circumstance in which this repertoire was first acquired, not when it crystallized.

I get a little bit hot under the collar with this whole theme, as it is sometimes presented in a rather disruptive way. The neurobiologists hardly listen to this story at all. They’re not distracted from the basics, they want to have reasonable confidence that they understand the underlying mechanism first.

David Rothenberg:
So it’s going to be a long time before neuroscience will have much to say about how songs are produced at the complex level?

Peter Marler: Yes, yes. Undoubtably there are all kinds of nuances that will vary greatly from one species to another. We already have several hints there. The starling is really very different from the sparrow. They don’t begin to learn until much later. It almost looks as though they remain continuously open once they begin to learn. There may not even be any seasonable limit to this ability. And with canaries it’s a different story. And then there is a radical behavioral insight derived from neurobiology about the stability of crystallized song. Everyone used to speculate about Lorenzian fixed action patterns, and what kind of physiological controls they are subject to, well we have gradually realized that that’s not a correct view of the situation. What happens during crystallization is a great increase in the refinement of various feedback mechanisms, which you only discover once you disrupt them.

David Rothenberg: If you disrupt it it will just keep trying to re-crystallize?

Peter Marler: Well the research on this is just beginning. It started in Kathy Nordine’s lab. When Konishi did his original deafening experiments, there were radical effects on young birds, but not such large effects on adult birds whose songs were already crystallized. That became dogma, and then Nordine took another luck. Indeed, there wasn’t much immediate effect in adult birds. But if you waited a week or two, abormalities began to creep in.

David Rothenberg: So where is the field going?

Peter Marler: Well, let me think about that. I do think bird calls will be one important area. With calls we may be able to study what birds understand by meaning, looking at connections between the song system and other physiological systems, calls may be the way to go. It may be possible to ask new kinds of questions with regard to motivational links.

David Rothenberg: Well Peter, thank you so much for generously giving so much of your time.

Header photo of yellow warbler singing by Double Brow Imagery, courtesy Shutterstock. is the first online literary journal of place, publishing award-winning literature, art, editorials, and community case studies since 1998.