How I Listen

David Sibley was gracious enough to reply to my recent post on Buff-collared Nightjar, first in a comment on my blog, then yesterday in a post on his own blog.  He takes issue with me on at least one point:

Nathan Pieplow seems to suggest that, for decades, observers have misidentified Vermilion Flycatchers as Cassin’s Kingbirds, and then mistakenly written that Cassin’s Kingbird sounds like Buff-collared Nightjar.

I guess I did seem to suggest that, but I didn’t really intend to.  Instead, I think I meant, “an authoritative source once misidentified a Vermilion Flycatcher as a Cassin’s Kingbird, and for decades, authors have perpetuated the error by simply citing the published assertion that Cassin’s Kingbird sounds like Buff-collared Nightjar.”  Given what David wrote in his comment and his post, even the latter assertion by me is a little unfair to him at least, and perhaps to many of the other people I mentioned in that post, since David didn’t just blindly repeat the conventional wisdom; he had field experience to back it up.  (However, see his post for an interesting discussion of how the conventional wisdom might have influenced his field experience.)

Overall, though, regardless of who (mis)identified what, David has started a very interesting and, I think, important discussion about a sea change that may be occurring in how birders listen to bird sounds.  He writes:

I learned bird songs decades ago through countless hours of field experience, supplemented by listening to a few recordings, reading detailed descriptions, and talking to other birders. It was a subjective, holistic approach to bird songs that led to a sort of gestalt style of identification – after you hear a sound often enough the identification just becomes second-nature.

Now, it still takes countless hours, but birders have a wealth of technological aids, allowing them to study and compare bird sounds with an ease and immediacy that was never possible before. In the modern world of ipods, sonagrams, and websites like xeno-canto, birders can examine the bird sounds directly, objectively, and in great detail. This may lead (as Nathan Pieplow admits) to a slightly greater emphasis on differences in pattern rather than the more subjective and hard-to-describe differences in tone.

Given how suggestible we are, and how tiny things can influence our perception, the detail-oriented objective approach to bird sound identification is probably better and more accurate. A similar shift happened in sight identification a couple of decades ago, and that shift can also be linked to technology. In the 1980s it was rapidly improving photographic equipment and optics that allowed more detailed study and comparison of living birds than ever before, leading to a whole new approach to identification based on feather details, molt, etc.

It may be that with modern technology Cassin’s Kingbird is no longer such a source of confusion with Buff-collared Nightjar. If so it has merely been replaced by another species (Vermilion Flycatcher) that is less easily sorted by the modern style.

This gives me a lot to think about.  Perhaps the different ways in which we learned our bird sounds might provide insight into how David Sibley and I listen to sounds differently.  He learned sounds in the field; I learned them on the floor of my bedroom in South Dakota when I was in high school, playing the Peterson Birding By Ear tapes over and over again.  Those tapes (which remain the best resource I’ve ever seen for people who want to learn bird sounds on their own) didn’t take a holistic, all-at-once approach; instead they took an analytic approach, grouping similar sounds together and then pointing out key field marks or “handles” — here a distinctive tone quality, there a distinctive rhythm — to distinguish sounds within the groups.

I’ve used this same basic approach to sound identification ever since: recognize a pattern, then focus on a piece of it.  The pattern gets you to the right group; the pieces narrow the identification to species.  Tone quality is part of this analysis, but not the most important part.

In fact, in some ways I think I place a pretty low priority on tone quality.  For several years now, I have been convinced that tone quality is the slipperiest attribute of sound: the hardest to analyze perceptually, the hardest to describe.  And I think tone quality is responsible for most of the disconnect between most descriptions of sounds and the sounds themselves.  I de-emphasize it precisely because it is so difficult to categorize.  Other attributes of sound are much easier to describe and compare, so those are the ones I focus on.

For the most part, I’m just doing what works for me, but I hope it works for other people as well.  I really do believe in the objective, analytic approach.  On the whole, I don’t think I can say it any better than I said it at the end of my Birding article:

When we set about describing a bird sound in words, we should avoid the temptation to describe how the sound makes us feel or what it reminds us of, since those things exist in us, not in the sound.  Instead we should strive to describe what is there: what can be measured with a stopwatch, pointed out on a sonogram, and defined in an empirical fashion.  I cannot claim yet to have accomplished this with pinpoint accuracy.  But I firmly believe that it can be done, and I firmly believe we should start to do it.

5 comments to How I Listen

  • Eric DeFonso

    There’s an aspect to this whole discussion that intrigues me and is in fact very pertinent to a project I’ve been working on for some time now. I’m writing a guide book on bird sounds in movies and television, and I frequently encounter situations where I have to identify a bird via a rather ambiguous vocalization. Certainly in the field this situation can arise, as is evidenced by this very blog discussion; however, in my case, I will never actually get a visual on the bird, I can never get closer or hear it from another angle even. I can’t even take an ecological approach and use the observed habitat to narrow down the possibilities, because in the movies any bird can occur anywhere at any time! It might not even be a real bird for all I know.

    In any case, this project of mine has made me think about how I listen to bird sounds. What’s more, in writing the guide I have to be acutely aware as well of how my future readers (the many millions of them :^) ) may be listening. Now, I’ve taken advantage of many of the newer technologies and available sound libraries, and these things help enormously to be sure; but in the end what matters is the approach that the listener actually takes when using them, and not so much the technologies themselves.

    I’ve found that when I do a ‘first pass’ on my movie surveying I use a very holistic approach – an aural ‘gist’ if you will. I see this as perfectly analogous to what was described in that Birding article a year or two back (I don’t have the reference handy) as well as that used the Crossley shorebird guide of which I am a fan. I also imagine that most birders do this for sounds anyway although perhaps not consciously. On a subsequent pass, I get much more analytic and listen very closely to quantifiable aspects to the sounds, if I need to narrow it further. This is absolutely necessary for distinguishing, say, Rose-breasted and Black-headed Grosbeaks. But in the end, when I write up an entry in the guide that the reader will follow when watching a show, I revert back to the more, well, subjective and perceptual style. I do that because I want this to be a fun, entertaining book as well as a friendly albeit hopefully subversive introduction to birdwatching for the unwashed masses. I fear that a more strictly analytic approach would scare off or annoy the very people I’m trying to attract (although I know that runs directly counter to approach you are advocating).

  • Dan Lane

    Hi Nathan,

    I am interested to read what you have written here, and find myself perhaps somewhat in between David’s and your camps in how I learned bird voices. I learned much of what I know from a mix of listening to published and non-published recordings as well as making my own (and, occasionally, learning a sound without recording it, which is my least-favorite method!). That said, I still remember my first time hearing the dawn song of Cassin’s Kingbird at Cave Creek in 1988, and thinking it was a nightjar because of time of day and tone. Clearly, since Buff-collared has never been heard there (nor is it likely to be), that didn’t enter my mind as a suspect, but the voice certainly did not register as a tyrant flycatcher. Happily, I recorded that sound and played it later to Kenn Kaufman, asking what the voice was, and he told me it was Cassin’s Kingbird. I think I was unaware of the concept of “dawn songs” and how different they can sound from the daytime songs of tyrants (and other birds), and the identity of the singer in that case really amazed me! I still think that, even though Phillips and the other authors may have misrepresented the similarities in pattern, a Cassin’s Kingbird sounds far more like a nightjar (in tone) than any Vermilion ever does, and Cassin’s Kingbird is far more likely to be misidentified as one.

    Now, after many years of identifying mystery bird voices, I find that it is more often the *tone* of a voice that gives me a clue to an unknown bird’s identity, and *then* I concentrate on pattern. Perhaps growing up in North America, an environment that is overwhelmingly dominated by the sounds of oscine passerines (and their dialect forming), has given me a leg up in uncovering the ID’s of neotropical birds (which are largely suboscines, such as tyrant flycatchers, and are less prone to dialect-forming), but I find it easier to wade through the confusing variation of song *pattern* of different populations within a species and concentrate on *tone* to make my first filter of what the voice can and cannot be. Is it a woodcreeper? A tyrant? A parulid warbler? Only then do I listen to the pattern. It is seldom the case that the pattern is more important than the tone in identifying a song. There have been a few instances where it is true, to be sure, but in my experience, these are far fewer than the other way around.

    So, I guess this is a long-winded way of saying that we humans hear, learn, and identify sounds in very individual ways. No one person’s way can be repeated by another. This is frustrating, and hampers our ability to communicate about sounds (which is ironic, since we so often communicate *with* sounds!). But I think it is a truth. That said, I see your emphasis on pattern over tone to be a little too black-and-white. It may be easier to quantify pattern, but it is also easier to be fooled by it. I remember, as a child, looking at the spectrograms published in the Golden Guide and wondering how on earth those could help a user? Frankly, I still do when I look at them today! In most cases, they are using one song of one individual of a species that probably has a large repertoire, and also probably has many dialects, and expect it to be of use to the average birder. Even worse, it’s hard to know the starting pitch of a sound as represented on a spectrogram, particularly since sound is perceived logarithmically, but spectrograms are not usually published that way (so particularly low-pitched sounds are hard to distinguish in frequency on spectrograms). This is not to criticize the usefulness of spectrograms– heck, I use them all the time! But to me they still cannot capture the important essence of a sound, one that is of utmost importance to me in trying to identify it.

    Bret Whitney has said that if you can identify a song to name and musician based on only a few notes, you have the same skills that it takes to ID bird sounds. To take this analogy one step further, if you can identify a song you’ve never heard before to a familiar artist, that is the equivalent of identifying a bird song irrespective of pattern (concentrating on tone, especially of voice, and, admittedly, perhaps “micro-patterns” such as chord progressions or guitar flourishes that a musician commonly uses). It is a wonderful (and heartening!) way to explain ear-birding to those who claim not to be able to do it.

    I find that likening a sound to another familiar sound is a good way to remember it, but that kind of link may only be useful to me, since I’ve had a unique experience and memory of both sounds that resulted with my likening one to another. This problem became painfully obvious to me while I was writing the voice descriptions in Birds of Peru a few years ago, and realized I could not compare any bird voices from Peru with bird voices from North America without alienating a large portion of the readership. It’s a shame, too, since it’s much easier to identify Black-billed Thrush (Turdus ignobilis) if you simply say “sounds just like an American Robin.” But South Americans and Europeans (and anyone else who is not familiar with American Robin songs) would both think that’s completely useless as a way to identify a sound, and rightly so! Thus, I restricted myself to comparing voices of birds only to other species found in Peru. Still, comparisons to everyday sounds that are familiar to most people who will have Birds of Peru in their hands– such as a rusty hinge, an angry traffic officer’s whistle, the down-shifting of a truck (yes, I used all these in my descriptions)– creates a better “aural search image” for the user; and these are largely tonal comparisons. I suspect I perceive the aural world differently than you do, but the tone really is one of the first, and most important, characters to assess for identifying a sound.

    Ok, I think I’ve babbled on long enough! But I hope that some of these comments are of interest to you and your readers in dealing with hearing and describing bird sounds.

    Good birding!
    Dan Lane

  • Nathan Pieplow

    Very interesting comments, Eric and Dan. Dan, I very much appreciate your point of view, but I can’t say I agree with all your points. In particular, I have to disagree with this one:

    We humans hear, learn, and identify sounds in very individual ways. No one person’s way can be repeated by another.

    If you’re talking about the purely subjective experience of sound perception, then I can buy that claim — but if it were true that nobody could learn another person’s method of hearing, learning, or identifying bird sounds, then my blogging would be futile. The whole time I was sitting on my bedroom floor listening to the Birding By Ear tapes, I was learning somebody else’s method of hearing, learning, and identifying sounds. I’m sure I was putting my own personal spin on it sometimes, but in a fundamental sense, I was internalizing somebody else’s analysis, and it worked well for me.

    tone really is one of the first, and most important, characters to assess for identifying a sound.

    Personally, I would love to replace the “is” in that sentence with a “has been.” Just because tone has been of foremost importance for the great earbirders of our age doesn’t mean it will be for the next generation. It will always be a very important, even crucial aspect of identification, but it requires an excellent auditory memory — a better one than I have, I dare say. I think there are other (maybe not better, but easier) ways to learn sounds.

    Finally, although I must say that I have nothing but respect for your field experience and David Sibley’s, when I hear you both argue that Vermilion Flycatcher may be less likely to be mistaken for a (Buff-collared) Nightjar than Cassin’s Kingbird is, I scrunch up my brows and think, “really?” I have difficulty fathoming it. Your and Bret Whitney’s great analogy would work well here: Buff-collar and Vermilion may be different artists, but they’re covering the same tune. Which is easier: recognizing a familiar vocal artist, or recognizing a familiar tune (even a new cover of it)? My argument is that, for beginners at least, it’s definitely the second, and that’s why Vermilion is the confusion species, at least for the uninitiated.

  • K. Cowcill

    Eric…that sounds like an interesting book. Will you include a little blurb about whether or not the species in question would be found in that habitat or location? E.g. loons heard in the desert, eastern phoebe in an African jungle, Red-tailed Hawk calling everywhere and at all times…. :)

    Nathan…I also tend to listen for a pattern to hone in on an identification, and then use tone if needed. I have worked with two birders who used pitch…turns out they possessed “perfect pitch”. They were both quite extraordinary when it came to identifying species on chips, zips, zllps, and seeps, or picking out Golden-crowned Kinglets from Blackburnian Warblers when the Blackburnian imitates the kinglet’s pattern. I don’t know how much of their skill was due to their hearing ability, but seeing what was possible inspired me to approach ear-birding more seriously.

    I’d like to see newer field guides include more information on bird song. David Sibley’s North American guide is the best I’ve found in that aspect. Unfortunately, when the book was divided into eastern and western it seems a lot of the song/voice information was omitted, probably for space reasons. This may be something more easily incorporated into a program like Dendroica rather than a field guide.

    Dendroica (Beta version, 1.0) is an aid to learning Canadian bird songs and was built by Antonio Salvadori, Charles Frances, Margaret Campbell, and distributed to many people for free when we were participating in the Ontario Breeding Bird Atlas surveys. It combines bird pictures with a variety of song-types from numerous recordings around Canada and North America so you can compare species dialects with just one click of a mouse.

  • Amigos, Dan Lane forwarded this discussion to me, so I thought I’d offer my perspectives on sound learning. First, I think that all of the posts on this subject are right-on in various respects. I’ll start with a sort of clinical perspective: The human brain is a totally amazing computer! The average or normal brain comes with the software to learn to recognize auditory stimuli on a level that the most sophisticated computers cannot match. More on that later, but for starters, it’s as simple as that. I think of it this way: Everything that happens as one grows up is processed and prioritized in the brain, and the pathways that are worn down the hardest, especially early in life, are the ones that are likely to endure and become highly developed. Pathways get worn down through repitition, especially frequent repetition, and it is through this process of repeated exposure that a person sharpens the ability to understand and, more importantly, interpret new information that hits the pathway. Ta-dum.

    For people with all five senses functioning normally, auditory information is always, always, associated with visual and other sensory information. That is very important to keep in mind, because the fixation and variability of auditory (or other) information is to some degree linked to supporting information. For birds, sounds of birds, that ancillary information varies hugely, but it’s almost always an integral ingredient of the memory. Bird sound learning in the field is very often supported strongly by visual images of the bird itself, of the habitat, of the weather that day, of the people by your side, ad infinitum. The single most powerful technique for learning bird sounds is to make recordings of birds yourself; there is no better way to learn bird sounds than to make and review your own recordings, and then to review your material with other recordings, and other individuals of the same species that you record in the field. Bird sound learning in a building is less richly supported by other stimuli, most notably the absence of other sounds and of the bird itself, but is nonetheless usually supported by images of the bird in a book or on a computer, perhaps spectrograms or personal notations/drawings of sounds, etc. — and immediate and repeated exposure to similar bird sounds, easy to do at home but not in the field, can certainly reinforce the “signature” element(s) of sounds. Something that is immensely helpful in learning bird sounds is to be able to imitate them by whistling. Whistling bird sounds provides strong and frequent reinforcement for learning because you are required to recall the sound over and over again as you attempt to imitate it. Even if your imitation is poor, the reinforcement step was probably beneficial, and if your imitation is reasonably good, you have probably truly learned that sound. All that really matters, whether you learn sounds in the field or through recordings made by others, is that the next time a bird sound hits your pathways, your experience with it and other sounds has been sufficient to properly recognize it. That’s where the level of sophistication of the human brain far surpasses any computer.

    The best ear-birders among us develop a certain “plasticity” in assigning a sound to a species group. Thus, an ear-birder, after repeated exposure to many millions or billions of sounds, from insects to amphibians to birds to mammals (in the natural world) to human voices, horns, motors, and gunshots (in the rest of the world), has developed the ability to immediately catalog a natural sound as such, and to assign it to cascadingly smaller or tighter groups by intrinsic properties of the sound, foremost among these pattern and tone, not necessarily in that order. People with deep levels of experience with bird sounds, far deeper than is possible to attain by listening to recordings, even all the recordings in the world, may develop the ability to identify even the most bizarre variants of bird songs to the correct species, calling in to play everything that characterizes the signature of that species for them (always a combination of tones and patterns). These highly experienced birders have learned limits of how “far-out” a sound can be before it leaves the realm of species A and slides toward species B, or perhaps away from everything it is similar to; it happens largely in the subconscious. It is this level of experience that enables skilled ear-birders to home in instantly on a strange call-note, a sound that is not part of the avian community where they often bird, or even to identify a species unknown to science because it is beyond the plasticity limits of other similar-sounding or closely related species, the judgement almost always supported by knowledge of biogeography.

    On this thread, I use an analogy to help explain what I mean. Imagine having a beginning knowledge of bird sounds such that each sound you know well is a thin book on the shelf of a bookcase. As you learn more and more sounds, new books for the shelves, you naturally place each one in its position relative to all others, sometimes way out on the edge of what you have previously come to know, but increasingly each new book seems to fit in close to another book already on the shelf, and they begin to hold each other up: new sounds are cataloged and archived relative to other sounds on your personal bookcase. After many years of repetition with many of the books on the shelves, it becomes easier to hear, learn, and archive completely new sounds after hearing them just once or twice because your shelves are so packed with books that anything new, whether it’s much like something else or not, is quickly cataloged and stored away. In fact, if you heard that sound only as a recording on an iPod on an airplane, you think about it carefully in the context of everything you know, concentrating on elements of it, perhaps talk about it with friends — all of which your brain incorporates into the process of etching the sound into your memory — and the sound is learned.

    Dan mentioned my analogy of learning music and bird sounds. I’ve said that I believe that any person capable of recognizing a Beatles song after hearing only a few notes is capable of learning to identify bird songs. I chose the Beatles because everybody has heard many Beatles songs so many times in their lives that they are indeed very familiar tunes. That’s the critical point: Repetition is the key to learning to recognize sounds, and things like specific album covers or who your date was or which wild party you attended while those songs were playing are powerful supporting stimuli, especially for remembering sounds from the first time you heard them. Any person has the mental capability to learn bird sounds, but only those that truly dedicate themselves to it will be great ear-birders. My own “pathway-burning experience” began when I was birding like crazy at the age of 6-8 without binoculars. I had to sneak up on birds to see enough detail to identify them. A by-product was that I quickly figured out that if I knew all the sounds a species made, I could pass it by and concentrate on any bird that made even a very slightly different sound. As a matter of practicality, my brain began laying the pathway for distinguishing subtleties in bird sounds. I was in the field for most of the daylight hours for months and years on end, so the repetition was, well, repetitive, and I habitually whistled bird songs and calls — all deeply reinforcing stuff.

    Ted Parker and I had a fun conversation about how each of us found it much easier to learn songs (not calls) of suboscine passerines than most passerines. (I recall that we were driving around Baton Rouge in my VW Rabbit listening to The Pretenders “Middle of the Road” just after it came out in about 1984; we loved the lyrics). We decided that it had to be heavily influenced by what kind of music you liked the most. If you were a rocker and identified with the blues, like us, the distinctive tonalities and hard-wired patterns of suboscine songs sank into your head. If you were more into jazz and classical music, then the oscines should come easier because so many of them have inventive and highly variable repertoires. It was the way your deepest pathways got laid down.

    One other kind of fun, recent anecdote that relates to the discussion… I love to listen to birds in soundtracks of movies, in part because they are often slapped in there just to have a cool natural sound in the scene, never mind that it’s from another continent or it’s a nocturnal bird at noon, etc. I was watching “There Will Be Blood”. About 3/4 of the way through the film, there’s a scene in the barren desert of west Texas where guys are digging a trench to lay an oil pipeline. Just wind in the shrubs, shovels in the dirt. Then a weakly audible bird sound comes in, just for a second, then comes back, also very weakly and briefly, a couple of seconds later. The first snippet made me think “Raven?”, but not quite “Raven!” That tiny window automatically opened the door to everything remotely raven-like, though I was still influenced at least subliminally by the landscape. When the second sound came in, it suddenly clicked that it was… a Hyacinth Macaw! I had to crack up, which was weird at that moment in the film. I imagine that someone involved in the film had a pet Hyacinth Macaw and decided to sneak its voice in there as a kind of watermark, a way-way-insider bit of fun. Hey, if you like warblers, rent “The Big Chill”. I haven’t seen it since it came out on the big screen a loooong time ago, but I think I remember that the first sound you hear is a singing Prairie Warbler, and they had Kirtland’s in the yard on at least two days. I think it was supposed to be springtime in Michigan, so an admirable attempt to be authentic.