Random Observations

“For man is but the servant or interpreter of nature; what he does and what he knows is only what he has observed of nature’s order in fact or in thought; beyond this he knows nothing and can do nothing. . . . All depends on keeping the eye steadily fixed upon the facts of nature and so receiving the images simply as they are.”     Francis Bacon  1620

The article titled “Dog Bites Dog Story “from the June 2007 issue of Scientific American caught my attention as I did a bit of web browsing this morning.  As with many good things, there was a lot more to this story than first met the eye.  As I began to read through it, I thought I might learn something interesting about canine behavior.  I was both surprised and delighted to find that the article was really about — science!

There are experimental sciences, and then there are historical and observational sciences. The experimental sciences, like chemistry and physics, are easy to spot. When stuff blows up or systems don’t work right, you’ve got yourself an experiment.

Historical and observational sciences can be a little tougher to get a handle on. The researchers in these fields must adopt the Yogi Berra stance—“You can observe a lot just by watching”—and then interpret reality.

Oh yes, the long-standing battle between experimental and anecdotal research continues.  Empiricists assert that true knowledge can only arise from experience, not from revelation.  Anecdotal evidence, being based on human observation doesn’t always qualify as scientific evidence for many reasons.  However, some subjects – like the study of human and animal behavior – don’t necessarily lend themselves conveniently to laboratory study.  I mean seriously, how can you study ‘normal’ animal (or human) behavior in a lab setting?

Experimental study is a wonderful thing – but anecdotal research is useful too.  When we’re faced with the necessity of using anecdotal data, how can we avoid the many pitfalls this kind of research seems determined to lure us into?  Well — like good experimental data, good anecdotal data must be collected and evaluated carefully. 

A nifty little article from Quirks.com called “Seven Rules for Observational Research” was referenced in Sciam’s “Dog Bites Dog” article.  It’s important stuff, so here (with just a bit of editorial liberty) are the rules:

1. Look for the ordinary, not the extraordinary.  Ordinary events are the foundation of science.  If you build a theory on anything else – its likely to collapse.
2. Nothing people or animals do is ‘natural.’  There is a reason or trigger for every behavior.  “Once you recognize that everything people do is the result of something, you can begin looking for that something. Maybe it’s something about them. Or the people they’re with. Or the environment they’re in. Or something else.”
3. If it’s really obvious, then it must be really basic.  Mastering the obvious is the secret to success.  Obvious things that everybody else had heretofore ignored is what led great minds like Galileo and Newton to insights that changed society.
4. God is in the details.  The truly important revelations of many situations are often found in the small, often-repeated details of the environment, not in the flashy, unique ones.  If something only happened once – it’s probably not important in the study of day-to-day behavior.
5. The observational perspective defines the study.  Or – you can only find what you look for.  The temporal and spatial scales of observation, point of view of the observers (both literally and philosophically speaking) and details of observation all affect the results. “Observation, like all qualitative techniques, takes some Zen. If your task is too tightly defined, all you’ll see is what you expected to see.” 
6. Context is key. “Think of activities as rings of context.  The whole activity is a set of behaviors that includes these small units plus at least one layer of context.”  Identifying which behaviors and/or situations control others is not always a simple task.
7. The most obvious things become obvious only in hindsight.  Context doesn’t appear until it has a big picture to figure into.  You often need to flesh out a lot of detail before you can separate the trees from the forest.

So, what does any of this have to do with dogs?  Well, unfortunately, junk science and junk journalism seem to have become the norm in reporting on animal behavior in today’s world.  Scientists and journalists seem not to even be aware of the seven simple rules outlined above. 

Sensationalism trumps science at every step.

The issue is especially problematic in the ‘study’ of dog bite statistics.  Random observations by opinionated, uninformed, emotionally over-wrought humans is selectively collected, sorted and sensationalized then reported as ‘news’ — or worse yet, as ‘science.’

Descriptions of dogs provided by members of the public at large, animal control officers, police officers, EMTs and others with little or no experience in breed identification are taken as ‘facts’.  This is compounded by the fact that breed identification isn’t a simple thing even for those who are experienced.  Not only is there an enormous variation in the appearance of members of many breeds of dogs – but the large number of mixed breed dogs present everywhere – and the fact that these ‘identifications’ do not use, and in fact commonly ignore, pedigree information, makes them utterly useless as data.

And just how to we define an attack?  If I get knocked over by a large dog and skin my knee, was I attacked?  If I smiling dog turns his head to face me just as I turn mine to face him, a large canine tooth bumps my tender forehead and cuts me – was I bitten?  And if I repeatedly torture my neighbor’s dog through the fence until said dog gets loose and exacts his well-deserved revenge – is he a vicious animal?

Now back Scientific American’s Dog Bites Dog Story.  It describes a heart-warming tale taken from the files of the Associated Press about how a ‘plucky little terrier’ protected its child owner from an attack by a pair of ‘marauding pitbulls’.  To wit:

“A plucky Jack Russell terrier named George saved five children from two marauding pit bulls…. George was playing with the group of children as they returned home from buying sweets.” So far we have an anthropomorphized terrier—plucky, and he was playing with them, mind you—and the Little Rascals returning from the candy store, when:

“Two pit bulls appeared and lunged toward them.” Next comes a quote from one of the kids, an 11-year-old animal behaviorist: “‘George tried to protect us by barking and rushing at them, but they started to bite him.’” Note that she goes beyond description to narrative herself—George’s primary interest was her safety. Now comes the resolution of the situation, according to the 11-year-old: “‘We ran off crying, and some people saw what was happening and rescued George.’”

The headline and the article thus conspire to portray a brave little dog that tried to rescue human children. And that may indeed be what happened. Based solely on the facts reported in this piece, however, we may construct a somewhat different narrative. The pit bulls appeared and moved in on the group; the terrier rushed at them; the pit bulls focused their attention on the terrier; the kids ran away. In other words, the same reported facts could have led to a story that carried the headline “Five Frightened Kids Flee as Tiny Dog Is Attacked.”

Oh yes.  Or maybe even more accurately: “Obnoxious, untrained, underexercised mutt runs into neighbor’s yard and attacks two boxers who were minding their own business.  Rude, undisciplined children amped up on a sugar high who were supposed to have left the dog at home, run off in a panic realizing that the neighbor who owns the boxers has been overcome with rightous indigination and will tell their parents what they’ve done.” 

Watch for the follow up story: “Children’s parents, after seeing overblown media blitz and realizing neighbor has excellent insurance decide to sue.  Local authorities (egged on by AR activists) propose a breed ban.”

Pavlovian Lullaby

This adorable little video was recently posted to one of the dog lists I subscribe to:

Along with the utter sweetness of puppies piling up and dropping off to sleep, there are some really interesting things going on here.

Nods to a friend on the list who posted a link to Greenpa’s LittleBlogintheWoods, where he wrote about how he did this with his children.

It started with my first child- when I was youngish and raring to go forth and prove that education is useful.  As far as I know, I just thought this up – by using what I knew – People expect singing to soothe the child; but I set out to intentionally CONDITION the baby to fall asleep when I sang.

Oh yes!  Being able to sing a tired, fussy baby or puppy to sleep would be a godsend to every parent, breeder and new puppy owner in the world!  As Greenpa posted: “Sometimes, when you are dead tired and want sleep- the little stinkers won’t.  And won’t let you do it, either.”

So, is this just a case of choosing the right song?  Will the right song or the right musical tones ease a fussy young one to sleep?   Um, not usually.  Try singing to a crabby toddler or a pup who won’t settle in his crate and what you’re likely find is that “The dirty so-and-so WAKES up to listen.”

So, what would Pavlov do? 

Classical Conditioning or Pavlovian Conditioning is a form of associative learning. It is typically induced by repeated pairings of a neutral (or conditioned) stimulus with a second (unconditioned)  stimulus that evokes an innate, usually reflexive, response. With repeated, perfectly timed pairings, the two stimuli become associated with each other and the organism begins to produce a reflexive response to the conditioned stimulus.

In Greenpa’s plain English the process is:

With somnolence aforethought then, I made an EFFORT.  I made it a point to be there, when the durn critter was already falling asleep because of sheer exhaustion. 

DON’T start to sing- until the child is nine-tenths gone.  Sing softly.  Continue singing for a good minute after the creature is clearly asleep.

Then do it again.  And again.  Then start singing when the kiddle is HALF asleep.  Repeat.

Then start singing just as they’re getting sleepy.

You should be getting the idea by now.  You are getting sleepy…  drowsy… so warm and comfortable, it’s hard to keep your eyes open…  you will send me money, lots and lots of  money…

Unending patience and perfect timing are the key.  It will also be helpful if you choose a specific song to sing exactly as the young creature drops off to sleep.  

When the child is actually conditioned, singing in the accustomed way can MAKE the child sleepy, and put them to sleep; ready or not.

And it works for things other than sleep.  This is pretty much the same process I’ve used to teach my dogs to urinate on command.  Say the word (not being one to employ baby language or other silliness I simply say ‘urinate’) just after the dog adopts the position (i.e. squats or lifts leg) but before it releases the stream.  The goal is initially to have the dog associate the cue with the sensation of urgency it feels right before it starts to urinate.

There’s even a nifty little book out on the process (teaching a dog to eliminate on command – not singing puppies to sleep). It’s called, appropriately “You Can Teach Your Dog to Eliminate on Command.”  (Note – if you check out the book on Amazon, be sure to read the reviews.  The third one – Bitterly Disappointed –  posted is a riot.)

Thank you, Pavlov.