by Steve Moon
Kevin Berbaum PhD is an emeritus professor and human perception research scientist. He has conducted observer performance research at the University of Iowa for twenty-five years as professor in the Department of Radiology and adjunct professor in the Department of Psychology. Kevin continues to write and publish research articles. I’ve known Kevin since the mid-1980s, when I was the photographer for Radiology. Kevin and I often discuss scientific research, methods and theory, and usually in the context of Bigfoot research. In this discussion Kevin digs into the physiology, physics, and psychology of Bigfoot research. He offers his professional observations and suggests ways in which we might become more successful in our field research. Kevin is interested in all aspects of the pursuit of Bigfoot, including the potential ramifications of the discovery and scientific validation of the specie. This was an easy interview to conduct. All I had to do was sit down and turn on the recorder.
We were talking about the limitations on hearing. As far as the information that you analyze, you’re sort of in a position where you’ve got a lot more information than is very easy to analyze, and so you’re looking for ways to select for higher probability events. The problem is, it’s sorta like, you know, the old cartoon… A fellow walks down the street and he sees a guy under the streetlight pawing around looking for his keys, and he says, “Well where did you lose em’?” And the guy says, “Well, over there in that dark field. But it’s a lot lighter here. I’ll look here.” It’s sort of that kind of thing.
You don’t know what’s going to happen on windy nights or rainy nights… But it’s possible that somebody with younger ears could pick some things up that you don’t. And it’s a reason to at least check. You also kind of want to have a ground truth on your signal processing because you’ve got limitations of microphones, of recording technique. You could be missing things because you’re looking under the street light. It may be difficult to get a kid with good hearing to listen to ten hours of tape for you. But it may be valuable. You can demonstrate a sound in a quiet environment, and say “I’m trying to find out if these things are occurring under the background layer of this rain, or this wind.” And also, “Here’s a moan. Here’s a whoop.” It may be that such things can’t be heard, at least the way we record them, by just anyone, which would be very helpful to know. If you knew that, you’d say well, the hell with all that stuff, but until you know that, you kind of wonder. And of course it makes signal processing harder because there’s a layer of stuff on top of it. And I suspect that the best detector that you can come across is a pair of good ears. Because what you need is a high class pattern recognition system that learns very quickly. That’s not so easy to come by.
My knowledge of perceptual processing by computers is somewhat limited. I was taught it before the current era. And some pretty amazing things can happen. You’ve got phones that can transcribe stuff for you, or you can talk to them. So, there may be effective procedures available now that I know nothing about. But you also have to work from your own knowledge base. So, I would be looking for some collaborators that have young ears to at least spot check some of what you consider to be suboptimal recordings, but also to listen across what you’re taking as your data. Because you’re scanning those with computer processing that gives you tell tale signatures. But you’d like to know, OK, if I put a kid down and just have him listen, what are they picking up that I’m not. Are there sounds that I should be concerned with, maybe in the same category that I’m interested in already, but there’s no evidence of it in the visual record? Because there may be more information in the recording than can be picked up by the processing.
If it’s a spectrogram, it’s giving you time, intensity and frequency. And so, there are limitations to what it can do, there. Frequency is temporal frequency, which is based on decomposition into frequency; temporal frequencies… Joseph Fourier analysis. Because you’ve got temporal frequency as a function of time, it’s some funny stuff going on here. Because when you say, OK, I’m getting high frequency information, that means that you’ve got peaks close together, and you’ve got a certain amplitude of signal, and so forth. And maybe there are a bunch of those peaks and troughs. Fair enough. But those sign waves are assumed to be infinitely wide. They go infinitely into the past and infinitely into the future. But you’ve just got a moment here. So, that decomposition is flawed in all likelihood. What you actually have starts out not in the frequency domain, but in the time domain. A wood knock is a prime example. (Makes a knocking sound.) It’s kind of like a click, and what clicks are, they are peaks of energy in the time domain that have very short duration. Now if you decompose those things into the frequency domain, it turns out they have all frequencies, and what varies is the intensity at each one. But it turns into kind of a straight line in the frequency domain. But now the natural way to encode that may benefit from considering frequency characteristics.
But what’s happening in the world is something in the time domain. It’s a physical thing. Knocking on a tree; it’s basically a high intensity but brief signal, in the time domain. There’s nothing about it that says “these are high, special frequencies.” I mean, that may be what the signal looks like when you’re looking at them, but there’s probably a broad spectrum there, and this may account for why some things may not be getting represented in your spectrogram. And so having a youngster there to listen, they may spot things that are not apparent from the spectrogram. If for example you made a recording over a half hour epoch, and you thought there were three events, but you have this kid listen to it, and well, there’s a lot more than that. Now you’re probably picking up the most intense ones. But if the wood knocks had to do with some kind of communication, and there were different intensities which encoded information for them…
I saw something about a Greek island where there were people that were shipwrecked. It was hostile to them, and so they developed this language that had to do with whistles, that kind of sound like bird whistles. But there’s a complete language in them. And they carry a long way, so they can communicate a long way. There may be only five or ten native speakers of this language alive now. But one of the things they were trying to do was to be able to communicate over a long interval, distant from each other. Sound has to carry, and communicate intricate information, and disguise you as the background, so that people can’t detect you. Sasquatch maybe have similar interests; requirements. So, one of the ways you could try to tease it apart is whether these sounds represent intelligence, attempts to communicate, versus, say a tree in the forest that bends to a certain point and then slips, and so it makes a kind of a knocking sound. So natural versus communication, so try to look at the pattern, and… Now I know you’ve done some recordings where you’ve put microphones across sections of land. I don’t know how well you were able to sync those things.
I haven’t really gone to that extent yet. We’ve talked about it, but we have not put it into…
Yeah, well I don’t know exactly what the best way to do that would be. But even if you had four mics, say, I don’t know, a quarter of a mile apart, each from the other… A technical problem is, you want to have things coming into the same recorder on different tracks. Which would allow you to localize sound. Accuracy probably depends on expenditure of money and effort, but you could determine if all of these sounds are coming from the same location. With four mics you can get so far. With a hundred mics you could get a lot further, maybe. But the idea is, if it appears that some of these knocks all seem to be happening from the same location, but there’s another location where knocks are occurring from. For some of our knocks, the mic that’s closest to that one is getting a lot more signal, but it’s represented in these others and it may be delayed depending upon how much. Probably not very much because they’re probably fairly close together. And that creates an ease with which to relate them together, because you don’t have the speed of sound working against you. So, a higher intensity signal means that mic is closer. But then if some of the other knocks, a different microphone is getting the biggest signal, then that’s another location. Well, so then, what is the encoding that is being used? What is being indicated? Is there a conversation going on between different locations?
Another line of inquiry would be to try to locate the sources of the wood knocks. So, you know, maybe you have a fairly basic way to do that; four mics. So you set up one night, and you’re kind of trying to triangulate. You’re trying to see… Well, OK, are there multiple sources or are there just a few? And where are they? Which direction are they from me in this setup? And given I know these directions, next night, if I move… Suppose I say, well, I know it’s off this way, and by ear it sounds like it’s gotta be a mile away, or maybe it’s gotta be close, or whatever it is. You could say, well, OK, it seems fairly close. Let’s set up the next night on the other side of it. So that we’re going to move our whole setup over here, and then see if it’s the other way, so that we know it’s between us. Now this presumes, of course, that you’ve got a source that didn’t move overnight. But the signals that are not from an animal are gonna be stationary most likely. If there are sources for these things that are inherent in the wood in the woods, then they’re gonna be the same all the time.
And so, to the extent that you could say, OK, we’ve got these signals, and we think they indicate an event. But for us to know that they’re not something else we have to try to find them. And whether it’s whoops or moans… Whoops is another kind of a big one! Whoop is… That’s a more complex signal. It’s not just a, (knocks on wood) a knock. All sorts of information could be encoded in it, by the nature of it. A question might be, do all whoops sound the same? Or not? And that’s another question that you’d want to have some confirmation from younger ears about. That’s kind of the first thing I would be to ask somebody who’s younger. Also maybe someone who has some musical… because people who understand music, particularly singing, may be able to hear things that others do not. For me, I can’t sing things very well, or never have, because I couldn’t hear what I was trying to reproduce in an intelligent way. Well some people can. My sister-in-law for example. She taught people how to do accents for the stage and movies. Things that are vocal like that… Now, her ears can’t be too good either, but finding a singer, say, who’s not too old, who’s willing to listen to em’… In some ways that’s easier. That’s certainly easier than getting somebody to listen to ten hours of virtual nothing. I think that would be hard. That would probably take a lot of pizza. But, you know, you could stage it. Hour of listening, slice of pizza. You could sorta turn it into a party. Probably be helpful to have some Bigfoot enthusiasts among the young because they would have some more motivation going.
You could play the same game with the whoops and the wood knocks, trying to see… Well, OK, is it a conversation? Are there different locations that are interacting, that are trading information in some way? Those noises have to be there for some reason. Possibly, “we need to find each other.” That would be a big one. But maybe not. Maybe “I need to tell you that stuff is going on.” “I need to communicate that I’m picking up human activity,” or “there’s danger.” “You need to stay where you are or move further off.” How is that encoded? Well, with knocks, if they really are knocks, then you’d think that it would have to do with… you know, the encoding has to be multiple knocks. And if there’s interaction… You know, it could be that all they’re sayin’ is “I’m here.” And so another knock, “Oh, yeah, I’m here, too.” Maybe that’s all the information that needs to be communicated. I’m here. I’m here. I’m here. So we know where we are. That may be all there is to say at some particular point. It could be the whoops, within the whoop itself, is encoding information, that is sorta like “danger” or “good eats over here.” “Wanna hook up?” Who knows what they’re talking about? But if it’s encoded in that way, then there oughta be distinguishable differences in those whoops.
One could say OK, I got a hundred examples of whoops. Here they are. And yes, I got more information about each one of em’; you know, where they were taken. Do some of these sound more alike than others? In other words, someone with a good ear, can they sort em’ into groups? And when those groups are looked at with their background information, do they sort into different times and locations? Are the differences in these whoops indicative of different individuals? Or are they… Another way to go is to look at all in one location at the same evening. So, here’s a whoop, here’s a whoop. Here’s a whoop, here’s a whoop. What is the pattern of location telling you? How many speakers do we have here? And what does their conversation look like? Is there a juxtaposition of whoops in different locations, or knocks and whoops, or knocks and whoops and moans, such that, the pattern starts giving you some indication of the language being used? Knock “I’m here.” Knock “I’m here.” Whoop of a certain kind; “Anything goin’ on over there?” “Uh, no.”
All of these are ways to expand, make the data more specific, so that you can begin to separate out what might be a naturally occurring sound from one that actually originates from an animal. And to the extent that you can get different locations interacting, you begin to see a language. Now, it wouldn’t have to be Bigfoot. I mean, it could be the sounds made by certain animals that we didn’t know they made. Or at least it’s not commonly known that, say, that a possum can moan. But your scientific pursuit would be advanced by knowing, for example, can we find these things… To what extent can some of these things be found because they’re static? Is there a tree that’s making these sounds? Can we find a tree that makes wood knocks that are not animal based? Because that eliminates a whole bunch of stuff that you were interpreting in an unhelpful way. And, science moves by elimination. And I think you do the same thing with the animals. You’d like to be able to… A skeptic would say “OK, you’ve got all these sounds, how do I know that that’s what you think it is? Of course, if you had a picture of sasquatch, or a movie of him, knocking on a tree, and then going “whoop, whoop, whoop, errrrr”, yeah, that would be swell. But it’s hard to get that, apparently.
And there’s not a lot of visual data. There’s sightings, like you say. Look down, and a kid told us there’s something down there, and we looked down there, and it looked to us like there was something down there. But there’s not a photographic record. (Kevin is aware of the PG film, but we have yet to discuss it.) If there was a photographic record, that would be the most convincing thing, of course. The fact that there is not, is probably what holds back acceptance more than anything else. And of course a lot of that is that our brains and visual system are very constructive. We tend to see faces in random patterns. We tend to see meaning when there is no meaning, because that’s the way or perceptual systems work in general. So how do you eliminate… It was dark, I saw a shape, and it looked to me like a form that was big, of a kind I sorta expected, and I saw some features that made me think “Hu! That really does look like that. It’s not just a patch of black.” You know, with the moonlight shining on a couple of leaves that look like eyeglow. It made me think that. So, why didn’t I immediately grab my camera and go “pow?”
And when I say pow, I mean a camera that doesn’t make noise, that has a big receptor on it so it can gather a lot of light, and it can autofocus in the dark. So that I get at least an image and a location, so that I have a chance at actually getting some capture that way. Of course a movie would be super, but I suspect that all of the evidence that you would gather that way would be hand held because it’s gonna happen… If you’ve got an animal that can hide from us they must be pretty good at it. So the idea that you’re gonna get some kind of movie is probably somewhat limited. It’s hard to get a camera that can record in the dark, for video. There are some now that people report that there’s something. And suppose you got a snapshot of something. And it looked to me like it was something. Well, and so if you know where the camera location is, and somebody says “well, is it still there?” You look and say “no, it’s gone.” If you take a picture of the same location when it’s gone then you’ve got a comparison. You’ve got comparison imagery. Before and after, close in time, so you would think that everything else is the same, so any differences between these could reveal characteristics of interest.
Humans are very good at comparing. We’re less good at absolute judgements, but comparative judgements, we’re really good at. If I hand you one beer, you’re likely to say “Hm! Pretty good.” If I say “Which one do you like better?” you’re probably gonna be able to pick it out. But, and then I say “Well which beer is the one you said is pretty good?” Eh, I don’t know. And it could be the basis for some computer; some traction. There are algorithms that one could use to try to isolate what it was that you were looking at. And you need some algorithms that are not totally frustrated by the fact that location has changed a little. And of course that kind of presumes that you have some ways of recording that are very sensitive. Cause, this all seems to happen at night. That is certainly one very good reason why there wouldn’t be much of a visual record, is because all of these things are being observed by dark adapted individuals. We’re able to see in the dark, essentially, once we’re dark adapted. And it’s hard to have a recording medium that can follow that.
I think Sony has some video cameras that are very light sensitive. Because they have a (35mm) full-frame sensor, but what they’re trying to do with it is not create higher and higher resolution. It’s not like a fifty megapixel sensor, it’s like a twelve. What that means is the pixels are great big, and being great big, they capture more light. The Sony A7-S series are often used in dim illumination these days because they can capture dark stuff. You can shoot in dark environments. Whether it would be enough, it’s hard to know. And of course one nice thing about Sony is they turn out cameras with high frequency, so that picking up a last generation at a bargain price becomes an option. Because you may not need all that it can do in the current generation. You may need just the big pixels. I think at some point they went to back illuminated sensors, too, which is kind of another factor.
I’d think about that. How do we get some visual information? It seems like the things that have been tried, you know, a different spectrum. Let’s do heat. The problem is those are fuzzy. A lot of things can interfere, mimic, and nobody believes it. Show me a fuzzy color image that could be anything. Well, clearly you believe it, but…
One of the problems in doing all this too, is that it’s really hard to get a team in the field that all understand what we’re trying to do. This is not like a snipe hunt. This is not a Cub Scout trip. We’re not trying to tell each other ghost stories here. We’re looking for evidence that is convincing.
What would drive me crazy, if I were doing it, is if I had some assistants or collaborators that… First off, well, we’re trying to get activity going here, so let me do some (knocks and whoops) with the notion that that’s going to start things going. Now, the likelihood is very good that it won’t start things going. Cause you don’t know the language. You don’t know what that’s for. And they probably know it’s you. You’re a human. And they don’t want to talk to you, so, that’s just bullshit. Don’t do any of that. I don’t wanna having my recorder going over here and having you make whoops. I’m recording, and I turn around and record what I think are evidence of Bigfoot, that you manufactured, and maybe you’re not admitting to, because… You know, we’re not all on the same page here. We need to be able to trust each other’s behaviors. If you’re going to emit a sound, we all need to know that that’s what you’re going to do, and that we know exactly what time you did that, and we don’t have people out there just fuckin’ around because they’re off on their own, and nothing is going on. They’re out there in the middle of the night, and so far as they can tell, they’ve been out there for three hours and nothing has happened. And they were lookin’ for a little excitement, and they got nothin’. So they’re trying… It’s not like an indictment of their personality. It’s human nature. So I think to be successful it requires a lot of planning, and probably a lot of instruction for the group as to what exactly we’re trying to do here, so that we have a code of behavior. So that we can get information that is reliable, that contains nothing that is a contaminant.
Contaminated data? You might as well throw it away. You’re in a situation where it’s hard to analyze the data to begin with. It’s probably on the limits of our ability to do it. So if there’s something overlaying that, because of your own people… All it takes is you could have ten people out there and one of them is doin’ that. That would be enough. And, I don’t know how you’d pick that out if it was that sort of situation. Because maybe they’re new, maybe they didn’t do it ever before but decided to then, or maybe they had this notion that, “oh, we need to get this going,” sort of thing, or whatever else it is. One person doing that, and twenty others listening, are all going to come back with different stories about what they experienced. And that’s going to be over the top of whatever else happened out there. And so, coming back with a bunch of experiences… I would find that extremely frustrating.
It would be enough that I would probably want to… If I knew we were going to have an event at a place, if I could do it, I’d probably go there on my own the week before, and without anybody there, to try to get an idea of, well, “What do I expect to happen next week?” So that if the experiences I got were very different from what I expected, you know, once you have the group out there… In science it’s all about eliminating alternative explanations. That’s how it moves. One explanation is that with a group, there’s always going to be somebody. It’s not because they’re a bad person. Not that we have the time to figure out what people are like. We aren’t going to know everybody that well. And I wouldn’t tell a soul that I was going to do that. I would just do it. Because first of all, one person can move around a lot less obviously than a gang. You can be as quiet as you can manage. And so you go out the week before and spend a night or two and just going around to areas you expect to explore, and making your own recordings, maybe basic recordings, and just experiencing… Well, OK, these are the kinds of events that seem to be happening, so that you have some basis for interpreting what happens the week after when you have the whole group moving around. And so, if it turns out that there’s a lot of events, or even some events, that didn’t happen the week before, then the question becomes, “OK, how much of this is our own activity?” You could even have stuff that‘s unintentional. Somebody’s walkin’ in the woods, and they step on a branch, and it makes a snap, that kind of sounds like a knock.
Even in physics the act of observing changes what you’re observing. Well there are a lot of mechanisms here where that could happen, for your investigations. And so managing that element would be a constant concern, I would think. And I probably wouldn’t make it all the same if I were trying to gather information. There are some things you can do with a group of twenty, but that you could also do with a group of four. So, were I concerned that contamination was happening, and I wanted a person on each mic, say, and I’m trying to get location information, well that means a group of four, and I wanna pick four people that I know are not going to add anything. And I’m going to give them a briefing that makes them know exactly what it is we’re after so that there’s no inclination to do it. If everybody knows that whatever we record is going to be interpreted looking at all four mics together, and that you probably can tell, in terms of, we don’t expect sounds we’re getting to be closer than an eighth of a mile. And in fact we may go out in advance and make some recordings which are control condition recordings where we set up an array of mics, and then we just go around knocking in the woods, to get… Here’s one where we knew it was an eighth of a mile away. We whacked on the tree, with a stick or something, and we got this recording, and so it’s a known. It’s ground truth. This mic was this high, this one was this high, and so we know it was in this direction. And we know from the amplitude how far away it was. That then gives you a basis for a recording where you aren’t making any knocks. And gives you more information about it. I don’t know… I think I’m running out of ideas.
Two weeks after the conversation above was recorded, Kevin wanted to contribute some closing thoughts.
The thoughts I’ve had since we talked last… What you got from me last time was sort of, um, thirty-five years as a scientist, and scientific thinking, and how a scientist thinks if you’re going to be in that business. You’re kind of thinking about, OK, what makes a good experiment? What makes something that I can get funded? How am I gonna get the next batch of money? I was pretty successful at that game. I think I had thirty years of grant support, another twenty years that I participated in. And so, that’s kind of the mindset.
What occurred to me is that that may not be the appropriate mindset for you guys. I mean, I know you want to do science, but it’s kind of at what level you wanna do science. Keeping in mind that should you be able to document these creatures, and prove it to the general public, well that would be extremely bad for Bigfoot. It would! I mean, Jesus, there are people who would want to hunt em’.
There are people who want to hunt em’ now.
But think what it would be if they weren’t laughed off. I got a guy that works for me sometimes. His names Ray. And he loves to hunt. Loves hunting. Goes on hunting trips. He has a pair of binoculars that cost twenty-seven hundred dollars, which is justified in his view. OK, so if I say “Do you want to hunt for Bigfoot?” (Kevin laughs) Oh sure! You know, which is really good. But that’s not the limit of it. Think of all the gawkers. Think of all the people… If you documented how they hide, where they live, what they do, how they mask their presence… Well then, you know, the human race is destructive. Look at how many Asian Elephant are being hunted for their skin. They make beads out of em’.
So in some sense, you have to ask yourself ethically, well, “Is that really the goal?” Maybe built into your investigations is a consideration that, well, “I’m doing this because I want to know. Because it’s of interest to me and I know people who are also interested, and it’s OK for us to know, but it may not be OK for everyone to know.” That’s something that you may have to build into all of your thinking about it. Yes we’re going to do science, but it’s gonna be… it’s not for everybody. As scientists we have some responsibility to protect these creatures, and probably humans against their own stupidity.
It is a major consideration of ours; preservation.
If you got a picture of Bigfoot, well you might not want to share that. You could show it to certain people. In a sense there’s a reason for you to kind of be a secret society. Really! I mean if it’s just pure science, and accomplishing scientific goals, that might lead to their extinction. In short order. So, probably you don’t want to write grants to the defense department, even though it could give you some equipment. Think of what the military might do. So that sort of thinking has been what’s sort of haunted my mind these last few days.
Thanks, Kevin!
