Friday, February 27, 2009

Notes of Brimstone, Cherry, and Gouda

Ill try to make this as non-partisan as I can, because I have no illusions that the phenomenon doesn’t cross party lines — but I was incensed by Gov. Bobby Jindal’s swipe at volcano monitoring research, as were more than a few people with a greater-than-elementary-school education. [1]

Some of you may be mercifully unaware of the latest entry in the political grandstanding game of picking out a research project that has a strange title and making it seem like a waste of tax payer money — without bothering to find out what it’s actually meant to study. Gov. Jindal, in a rebuttal to President Obama’s televised talk on the state of the nation (if not a formal State of the Union Address), singled out $140 million to be spent on volcano monitoring, and asked, possibly intending an Andy Rooney vibe, what the heck volcano monitoring was.

Now, in view of how little they’re taught about — well, anything — gov. jocks have no right to be spouting out about science. Not without either educating themselves (as some do) or hiring somebody who knows something about it (as far too few do). But while I’m appalled at Gov. Sarah Palin panning fruit fly research [2], at least I get that. It comes from an ignorance of the science — and ignorance is no sin, only the attempt to cling to it is — meeting up with what, after all, is an odd avenue of research if you don’t know the scientific history. How anyone who’s old enough to remember Mount Saint Helens blowing a big chunk of Washington State to Kingdom Come could think that volcano monitoring, even if you haven’t heard the term before, isn’t a very serious matter of human life and death is far beyond my humble imagination.

One amusing, if disturbing, possibility: the 1971-born Jindal has speechwriters who aren’t old enough to remember Mount Saint Helens, and Jindal doesn’t review their work ahead of time. Sounds ridiculous, but it’s no more ridiculous than explanation #1.

Whatever you think of Gov. Jindal’s political positions, I hope that the fact you’re reading means you’d support his consignment to the infernal regions for crimes against science [3].

Still, every once in a while you come across a research paper that, while completely valid in its own right, simply begs for a punch line. Such as teaching mice to discriminate between different wines.

Noboru Takiguchi and friends at Hiroshima and Osaka universities found out that they could train 10 mice to 70 percent concordance (i.e., 20 percent better than a random, even-odds chance) in discriminating between red wine, white wine, rose, sake, and plum liqueur.

But the interesting result came when they took their mice (trained, remember, to tell between red wine and other alcoholic drinks) and tried to see if they could tell the difference between different red wines (Japanese reds Bon Rouge and Bistro Red, and Beaujolais Villages, from the admittedly more familiar wine-associated country of France). Six mice couldn’t do it better than random chance (50 to 67 percent concordance), and only two could do the job, with an average concordance of 75 percent.

I suppose nobody’s going to be firing their sommeliers and buying mouse cages over a 25-percent-better-than-guessing result, but the fact that the remaining two mice had a concordance lower than 30 percent is amazing.

Think about that: the mice were discriminating effectively between the choices, but making the wrong choice anyway, even though they were being rewarded for the right one. Takiguchi and his kameraden think it may have to do with the mice concentrating on the wrong odorant components. While it isn’t immediately clear to me how that would work — wouldn’t it explain inaccuracy rather than the observed “reverse accuracy?” — it’s a damned interesting idea that, not that I have to say it, is worth pursuing. No wait, I do get it. The idea is that the odorants you use to tell between wine and sake are different than the ones that distinguish between different wines. Even more interesting, because there’s a parallel issue in SAR dog handling: do dogs that are distingushing between the scent of individual humans pay attention to different odorants than those simply telling human scent from other animals? Still, we don’t know for sure; another paper by this group, showing that just about all mice can distinguish between reds if trained to do so, would help strengthen the argument.

Or is it shades of Punished by Rewards? Were the mice, resentful of the patronizing task of choosing between a decent French red and two less-than-inspiring domestics [4], trying to send the researchers a message? [5]

I see a result either way. Contrary to what politicians seem to think, research is its own reward.

[1] Also check out this Yale Daily News piece.
[2] Hey, fruit flies are funny.
Even Did a Cat Shit In Here? got into that game.
[3] No, I’m not going easy on Barack for the
who-invented-automobiles gaffe. It’s equally embarrassing; but at least it wasn’t the core of an attack on a line of research.
[4] Good sake is fantastic — but let’s face it, Japan will never be famed for its vineyards.
[5] Hey, the first reader who can tell me the title and author of an SF short story I read ages ago, in which a psychology researcher awakes to find himself in an alien’s
Skinner Box, gets an arbitrary, probably not valuable ... um, reward ... of my choosing.

Thursday, February 19, 2009

Who the F@#k Are You?

The charge was as ridiculous as it was unexpected. Last summer, Heather opened up a letter from the Pittsburgh Traffic Court that said I needed to come pick up my car from the impound lot, said vehicle having been towed for illegal parking.

If I remember correctly, she stuck her head out the front door to confirm that my blue 2001 Subaru Impreza hatchback was still parked in front of the house. Then she noticed that the impounded vehicle was a red Ford pickup truck. Also, on the date of the ticket we were both in West Virginia on an extended search-and-rescue training exercise.

Apparently, the ticketing officer had made a slight error in entering the license plate number — the tag number on the ticket was mine, which was why the letter had come to our house. But state records clearly show that that number doesn’t belong to a red pickup, so we were puzzled as to how the mistake had gotten so far without anybody catching it.

After talking with the traffic court folks on the phone, Heather sent a letter to the court explaining everything, and we forgot all about it.

Well, early this year the calls started coming from the collection agency. Not only had justice been blind enough to miss the more-than-reasonable doubt of the case, not only had she ignored Heather’s letter, but she had, Star-Chamber-like, carried out a legal proceeding without notifying the accused — remember, the summons for a traffic violation is the ticket, their letter was just an extra notification — found me guilty, waited until I was in default of the ruling I never heard about, and then passed me onto the collection folks.

The latter were hilarious; when I called with the evidence that it couldn’t have been my car, they not only said they couldn’t take it into consideration, but that I might wind up having to pay the ticket because it had gone into collection and so couldn’t be re-adjudicated by the court.

I’m normally a fairly placid guy (Heather tells stories of Sicilian fits of temper, but given her own short fuse, who really believes her?), but people who know me know I can be a stubborn son of a bitch when I think I’m right. I explained to the collection rep that I intended to call my attorney and make sure that they spent more getting the fine out of me than the state would pay them for the collection.

Amazing how talk of lawyers brings out the reasonable side of some people. She told me that I could certainly send whatever documentation I had to prove the vehicle in question couldn’t have been mine (for all I know, it’s still sitting in a Pittsburgh impound lot).

Well, I called the traffic court for some details on the actual violation, and when the woman there heard my story she said she could make the issue go away. I’ve got all the documents saved in case I need them, but since then I haven’t heard from the collection folks. We’ll see, anyway.

All of which is to say, identification is important, and a change in context — say, the same license tag on a different vehicle — shouldn’t mess up an accurate ID if you’re careful. This week’s entry is a study showing how the individual smell that each of us possesses remains identifiably constant despite the fact that our environment clouds the issue. The report comes from Jae Kwak and a team from Gary Beauchamp’s crew at the very interesting Monell Chemical Senses Center in Philadelphia.

First, some background. Eighteen years ago, Kunio Yamazaki (also from Beauchamp’s lab) reported that mutations in a gene normally associated with identifying pathogens and rejecting transplanted organs somehow made one mouse smell different than another.

What emerged in the next decade of research was a picture in which a system that everybody thought was meant only to help the body tag dangerous invaders (and the body’s cells they had co-opted) for destruction may actually have another, even older function: Picking and choosing among the many bits of protein, fats, and other chemicals that we ingest to create a unique bouquet of smells that belongs to each individual.

The genes involved are called the major histocompatibility complex (MHC) in vertebrates, or the human leukocyte antigens (HLA) in humans specifically. The produce a family of proteins that are the body’s ID cards, which infection-fighting white blood cells examine, like bouncers, to determine who belongs and who does not.

The MHC proteins, in their originally understood role, pick up fragments of the normal components in a healthy cell and display them in a kind of one-two code. If the MHC protein and the fragments are normal, the white blood cell does nothing. If either is wrong, though — say, the right MHC protein, but displaying a fragment of a virus that has infected the cell — the bouncer calls in the cops: “killer” white blood cells that destroy the infected cell before the virus can spread. It works like this:

Well, it turned out that the MHC proteins also seemed to be able to pick up fragments of normal cellular products, ingested food, perfume, who knew what, and released them from the body as smells. Since each individual (except identical twins) has a unique spectrum of these proteins, he or she also has a unique spectrum of smells. That works this way:

You can read a more complete explanation of how the system works on AMRG’s website.

This discovery resonated strongly with what search-and-rescue dog handlers had known for years: both our genetics and the smells that normally surround us contribute to an individual scent that a trailing dog can follow, if given a scent article that individual has touched.

But the problem was, how can a dog recognize an individual human by smell — or any animal recognize another — when, potentially, you could give the smellee a substantially different bouquet just by changing environments? In other words, change utterly the palette of potential externally derived smells form which the MHC proteins select?

Well, Kwak and the Philadelphia gang showed that even when you change a mouse’s environment radically, the MHC proteins are able to scrape together enough familiar odor-carrying chemicals that another mouse trained to identify that individual could still tell that the smells were coming from the same mouse.

It’s a classic case of scientists confirming something dog handlers pretty much knew, but which is still vitally important in understanding the basics of how scent work happens. And, not inconsequentially, it establishes that, provided you do the training right [1], further demonstrates that there’s a healthy scientific basis for what dog handlers do to help find lost people.

Anyhow, Pittsburgh Traffic Court could take some lessons from tasks that dogs (and mice) carry out just about every day. We’ll see who turns out to be smarter.

[1] Do not, do not get me started.

Thursday, February 12, 2009


Broken cutters, broken saws,
Broken buckles, broken laws,
Broken bodies, broken bones,
Broken voices on broken phones.
Take a deep breath, feel like you're chokin’,
Everything is broken.
– Bob Dylan, “Everything Is Broken

The first person I ever helped to rescue killed himself, years later, after being accused of molesting children.

What do you do with that? How do you make sense of it?

This past week has offered a number of strange and maybe-not-so-coincidental connections, which started with “The Smell of Fear,” a Did a Cat Shit In Here piece about the link between fear responses and the protein underlying mad cow disease and CJD, the human brain-wasting disease. They ended with my vivid memory of a chilly morning in the woods of New Hampshire, the smell of a little survival fire hanging in the air, when a search team I led saved someone who may have gone on to hurt people.

It begins with Prp-Sc, a mangled version of Prp-C, itself a protein produced in the brain but until recently having no known function. Prp-Sc is a bizarre type of infectious agent called a prion — no living thing, no parasite with its own agenda, but rather a protein that folds incorrectly. Such a small thing, a single protein molecule: but Prp-Sc has a terrible ability to cause its “healthy” Prp-C neighbors also to misfold, the end result being a tangle of broken protein that kills brain cells.

Slowly, inch by inch, a human being — or a sheep, or a cow, or a deer — goes away. Never to return.


Well, nature doesn’t create anything for no reason. The simplest explanation, perhaps, is that Prp-C has a function so important that having it is worth the risk of it converting to its harmful Prp-Sc alter ego. And Bruno Lobão-Soares and colleagues at the University of Sao Paulo found a hum dinger of an important function: mice that lacked it failed to display the normal fear of a nearby predator.

Now, fear responses are often triggered by smells. But the connection to smell turned out to be far stronger than that: Claire E. Le Pichon and crew at Columbia University and elsewhere have now reported evidence that mice without Prp-C lose their sense of smell. They lag behind normal mice at finding hidden cookes; they can’t find cookies any better than a strain of mice known to have no sense of smell; and reintroducing Prp-C into the brain restores their sniffers. The mutation seems to disconnect activity in the brain’s olfactory bulb from the rhythm of breathing — normally they’re pretty much synchronized — and generally dampens the normal electrical oscillations in that vital first routing point of the sense of smell.


Well, we have two further connections: one predictable, if frustrating; the other blindsided me, and brought me back to that New England hillside.

As reported on Pet Connection — rapidly becoming a go-to source for information on both pet and human food safety — the FDA is playing down the potential risk that meat from an elk with chronic wasting disease may have entered the food chain. Now, I’m no FDA basher: the story of how rookie FDA medical officer Frances Oldham Kelsey stretched her authority to the breaking point to prevent thalidomide from being given to pregnant women in the U.S. is a hero’s tale that screams for a dramatic treatment, and ought to be told to our young around the cooking fire. It set a tone that, despite wobbles, has kept the FDA in the “drug cops” role, protecting us from the most egregious effects of over-eager (to be charitable) pharmaceutical salesmen.

Still, whether we’re talking poisoned dog food, peanut products for human consumption, or E. coli-tainted spinach, I think it’s hard to avoid the conclusion that the FDA still doesn’t get food safety right (mainly because of scant funding).

The system is ... broken.

And finally, what did my wondering eyes encounter in the January 22 issue of Nature but an obituary of Daniel Carleton Gajdusek, the microbiologist who led the team that discovered prions. Gajdusek’s discovery hasn’t paid off yet — we still don’t know enough about prions and how they cause disease — but when we do come up with an agent that can rescue people from the long, slow dissolution of self that prions threaten, it’s a good bet that his work will underlie it.

Thing is, Daniel Carleton Gajdusek was also a convicted child molester. Something about him, if you trust the courts, was undeniably broken.

What do you do with that? How do you make sense of it? Is anybody going to refuse his elderly parents a treatment based on these discoveries because of what and who their discoverer was?

But that’s not the whole reason I bring this up. Here, in the interest of fairness and in-context understanding, I present the entirety of the relevant paragraph of the obit:

“Eccentricity was the source of Gajdusek's genius as a scientist, and of his notoriety late in life. In 1997, he was imprisoned on a child molestation charge involving one of the more than 50 Micronesian and Melanesian children he had adopted and brought to the United States. On his release in 1998 he moved to Europe, which he regarded as less puritanical than his home country.”

“Eccentricity?” “Puritanical?”

I think — I hope — that implicit in this paragraph was the author’s belief that the charges were unjust, and Gajdusek wasn’t in fact guilty of molesting a child. The author is from the Netherlands, to which Gajdusek emigrated after getting out of prison, and so there may be an issue of understanding the full implications of the English text, however well the author may speak the language.

I view this more as an utter failure of editing: How can an editor encounter that paragraph and not jot, in the requisite blue ink, “Please clarify — this sounds like you’re making an apologia for child molestation”?

Would it be impertinent to suggest that the editorial process here was ... well, broken?

I’m sure interested in checking out Nature’s Correspondence section in the coming weeks.

But returning to my original question: How do you make sense of all of this? I guess it depends on what kind of sense you’re looking for. If you want the sense that science can offer, then it has to do with the difficult tradeoffs in evolving complex organisms. Somehow, the crucial function of Prp-C — the structure that its chemical function dictates — brings with it the potential to go rogue. Somehow, giving a human being powerful reproductive urges brings with it the potential for those urges to be appallingly misdirected. (And that is my link to today’s celebration of Darwin’s birthday.) [1]

The problem with what the scientific explanation can offer us lies precisely in its objectivity. We don’t want child molestation to be a matter of objectivity; we want those who commit it to be evil, we want to judge. But in the words of Colonel Kurtz:

“You have no right to call me a murderer. You have a right to kill me. You have a right to do that; but you have no right to judge me. It’s impossible for words to describe what is necessary to those who do not know what horror means.”

What is it like to be broken? How, if we ever reduce the phenomenon of human misbehavior to a matter of chemistry, do we make sense of that?

But I’m ready to answer my first question. My answer may be a nonsensical, fuzzy mix of Hegel and Tolkien. But it got me through the long nights after the phone call that told me of Scott’s suicide, and I still draw from it. The fact is that I’m not smart or wise or good enough to make the call of who lives and dies; I’m broken myself, in my own, hopefully small ways, and that gives me no right to ask, certainly not before the fact, whether this or that lost person is “worthy” of being found.

We do our best to find them, to bring the lost sheep home, because that’s what human beings, as broken as we may be, do. If you need more meaning than that, you’ll have to look elsewhere.

Wednesday, February 4, 2009

D. melanogaster, We Hardly Knew Ye

I get the most adorable dominance displays from people with modest scientific backgrounds who assume I’m “just a dog handler.”

The incident of the day had started out mutually respectful, at least. The individual in question — an animal behaviorist [1] who’d come to a SAR conference to lecture to us on drives — had actually given me a lot of interesting new ideas to think about and employ. Mostly it took the form of an expanded list of instinctual drives that, in their combination, help decide an animal’s reaction to different situations and stimuli — prey drive, social drive, defensive drive, etc. Heather later told me it was pretty standard fare in dog training terms, but it had been new enough for me.

The changeover came after the talk, when I approached the lecturer to ask a question. I asked, got my answer, and then, perpetual optimist I am, offered a philosophical observation:

“Of course, we’re prisoners of our terminology, aren’t we?” I said. “We can’t be sure if any of these drives actually exist.”

I think I read a mix of emotions on her face: amazement at my temerity, engaging her as if I could grasp the material at her level; anger at what she perceived was a challenge to her entire career; pity at the possibility anybody could be so ignorant.

“These are the drives,” she said. “Of course they exist.” Then she got herself the hell away from me.

Of course, my statement was a bit over-reaching. The drives do exist, as a statistically valid entity; as a concept that can help us aim and craft our training goals; and I’d never disparage the power of metaphor for understanding the world. But on the day that we can put a portable combined PET/MR scanner on a dog’s head and actually see what her brain is doing when she is defending her litter, or cuddling with her pups, or chasing a ball, I’ll lay a bet that what we see is not three different brain states but a murky amalgam of subtle shifts in activity and unexpected overlaps that offer no neurological justification for what we thought were the basic modules of behavior.

My animal behaviorist couldn’t see that possibility, because her own training had been limited to just her craft. She’d never stepped outside the paradigm, and so could not see the ways in which the terminology she was using — the names she had been given by her teachers to use — had taken on a reality that might not be entirely, well, real.

Names are the crux of a brouhaha, recently reported by Rex Dalton in Nature, that’s, well, brewing among species taxonomists over the not-so-humble common fruit fly, Drosophila melanogaster. More data have been generated from this tiny beast and how its genes govern its development, anatomy, and behavior than possibly any other.

Thing is, it may have the wrong name.

It all has to do with how species are defined in the first place. When a scientist thinks she’s discovered a new species — whether it’s a live critter or fossilized bones — she needs to produce a type specimen. This is an example of the species that will serve as the benchmark for any other critters found. If the newcomer is too different from the type specimen, e voilà, chances are it’s a different species.

Just as species — in the fruit fly’s case, that’s its second name, melanogaster — are defined by a type specimen, the genus — the first name, Drosophila — has a type species. All candidates for membership in the genus are judged against that species. If you’re too different, again, you may wind up in a different genus.

You might think that D. melanogaster, having the high-powered, monkey-navigated support that it does, would be the type species for Drosophila. But genuses and species are defined historically, first come first served, and in this case another fruit fly, D. funebris, has precedence, having been identified by a dude with the suspicious-sounding name of Johann Fabricius in 1787.

Well, on second look, D. melanogaster doesn’t look as much like D. funebris as it ought. [2] It’s different enough, in fact, that some taxonomists are seriously thinking of renaming it Sophophora melanogaster, promoting what had been a subgenus to the status of a full genus. [3]

As I said, D. melanogaster has friends in the kinds of places where opposable thumbs can do a lot for you. The heresy of renaming one of the most important research species on the cinder was too much for some researchers, who’ve proposed that the type species for Drosophila be changed from funebris to melanogaster.

It isn’t clear what changing such an important animal’s name would do to the scientific literature. In the early Pleistocene, when I went to grad school and you had to look up articles in paper reference books, it would have meant utter chaos. Modern electronic referencing systems may make the switchover relatively invisible. But ya gotta love any science story that quotes a researcher — in this case, Therese Markow of UC San Diego — as saying:

“If this were some obscure beetle, you could rename it Godzilla and it wouldn’t make much difference.”

Well, at the time of publication, the International Commission on Zoological Nomenclature had not yet ruled on the issue, and had even suggested they might just leave it up to the researchers to sort out without an official nod. So maybe a fly is a fly is a fly after all.

For more discussion on the concept of species, and how it’s getting a little wear-worn, check out some interesting back-and-forth on Stephen Bodio’s blog. But please, remember: There is no such thing as “just a dog handler.”

[1] Yeah, the title itself loses any real meaning, ranging from Skinnerian operant conditioning technicians to ethologists -- interestingly, the Wikipedia entry is almost entirely ethological and hasn’t garnered a peep of protest from the Skinnerites.
[2] Or rather, their genes don’t look enough alike — the new relationships have been coming out of studies of the species’ DNA sequences.
[3] Taxonomists — don’t get me started.