« Life's Hard Enough As It Is... | Main | Heh... »

August 25, 2014

The Scientific "ManBearPig"

What do you do when you can't neatly separate out multiple influencing factors to establish causality? Run them all together and give the result a fancy name like... oh, we don't know... "neuropsychosocial":

When it comes to understanding ourselves, we tend to be splitters: mind and body, nature and nurture, or genes and environment. We take such a split for granted when we ask how the social becomes biological, but sometimes it’s not so useful to dichotomize the world into society and biology. Instead of looking for distinct social and biological influences (and believing that we can change one but not the other), we should recognize that the factors that drive our social behavior can, like a Zen koan, be two things at once.

Take the case of teen alcohol abuse. In a study published last week, an international team of researchers reported the “neuropsychosocial” factors that identify teens who are likely to abuse alcohol. The word “neuropsychosocial” does away with the common nature/nurture divide, and so did the researchers. Rather than asking whether teens abuse alcohol because of social influences or innate biology, the scientists looked at those variables that could be measured, regardless of whether the variables were social, biological, or a mix of both.

Yes, the Editorial Staff are making fun of this - a little. But it's actually a sensible approach to situations in which a large number of factors combine, in ways that are nearly impossible to predict, to influence an outcome:

As the authors write, their data “speak to the multiple causal factors for alcohol misuse,” and, in fact, any one variable, taken in isolation, had a small influence in their study. The predictive power of their computer model came from combining variables that were measurable—regardless of whether they could be neatly categorized as social or biological—into a single risk profile. This profile offers clues for how to find and help at-risk teens, and the most effective interventions may turn out to have little to do with directly treating some key social or biological cause of alcohol abuse. As we think about the connection between our social behavior and our biology, we should, like good scientists, be pragmatic, and abandon the distinction between society and biology when it’s not useful.

Part of what we do in our day job involves studying software productivity. Everyone wants to find a single, simple "fix" that will make teams and projects more productive (however that's defined: definitions seem to vary with the observer's priorities). But we're inclined to think that software development - like pretty much any other complex human endeavor - is influenced by a constantly shifting mix of management, technical, and human factors; none of which can be neatly separated out from the others and some of which are impossible to quantify with any objectivity.

At any rate, we found this amusing as well as thought provoking.

Posted by Cassandra at August 25, 2014 07:52 AM

Trackback Pings

TrackBack URL for this entry:
http://www.villainouscompany.com/mt/mt-tb.cgi/5276

Comments

Let me first say I am aware of the nuance applied to your assessment of what’s being got at here. But I see it as more dastardly therapeutic mumbo jumbo. It’s Deepak Chopra all over and he’s slipped his tethers and on the loose.

“we should, like good scientists, be pragmatic, and abandon the distinction between society and biology when it’s not useful.”

And just when might that be? At what point does obtuse become ‘hasn’t a clue’? We are, as a society, in the midst of a degenerative impulse so destructive that the imagination is overwhelmed in the recounting of it – and one of the report’s authors calls for pragmatism. I don’t know that I’d expressed it here in the comments before but, if so, it bears repeating. You may know the trend of society, a nation, or civilization itself thus: in its ascendancy it makes satire of policy; in its declension it makes policy of satire. Declinations of tradition, the natural order, nature itself, are all sacrificed for pragmatism. And here be a mighty satire.

Pragmatism is the new high expectations.

Posted by: George Pal at August 25, 2014 11:35 AM

George, you often seem to think that biology is destiny.

I think the human race is driven by all sorts of factors: biological drives, physical factors, culture, religious faith, upbringing, family genetics.... the list goes on and on.

I see some broad, general characteristics that could be said to apply *more* to men than women (or more to women than men). But my two sons and my husband aren't anything alike in all kinds of important ways (and none of the three conforms to the stereotypes I'm always reading about "all men").

Likewise, I was more boylike (but only in some ways) than my brother. At the same time, I loved to get dressed up. But only occasionally - most of the time my Mom had to catch me and braid my hair because I didn't really care how I looked.

So to me, it's actually eminently practical not to get too immersed in narratives like "Men are like this/women are like this". There's some truth to them, but there's also a ton of confirmation bias (we notice what confirms what we think and discount anything that doesn't, etc.).

"It's not useful" to worry about single-factor causation in a situation with multiple, interacting influences, none of which is clearly dominant.

Posted by: Cassandra at August 25, 2014 12:00 PM

Clarification: if I've misstated your position, please let me know. The above is just how it seems to me, but I don't always accurately interpret your comments :)

Will happily accept correction if I'm wrong!

Posted by: Cassandra at August 25, 2014 12:11 PM

Academic specialization is a difficult problem, really. On the one hand, it's necessary to specialize to train people to the level of expertise necessary to the practice of modern sciences. On the other hand, the specialization itself creates voids -- you can't think about some problems because you haven't been taught how, and the people who have don't have the specific expertise that you have. The journals in which you might publish any findings you do manage don't want your findings either, because they are structured to serve a narrow field as well.

Efforts to unify the branches of knowledge are thus necessary to any proper understanding. It's a difficult thing to do, but important.

Posted by: Grim at August 25, 2014 12:30 PM

Those buggers are dressing up a simple proposition in fancy language in order to appear to be 'smart' academics, despite being unable to make a scientific conclusion supported by data. Let me paraphrase and translate:
'We can't answer the question we wanted to, so this stuff must be too hard to understand because we are really smart.'

Science offers many tools and approaches, best of all a good scientist is always ready to admit that any one of their own projects failed to produce any conclusion supported by data (in fact any project using inferential statistics based on population samples ought by definition fail to replicate the population in a small percentage of the samples; that's the math!).

Posted by: CAPT Mike at August 25, 2014 01:03 PM

“George, you often seem to think that biology is destiny.”

I am disposed to that, yes – and will pay no deference to PoCo sensibilities – not that you’d be one to be demanding it.

I am not biophobic - the irrational fear of biology and the belief that mentioning genetic/biological differences makes one a troglodyte. I have no expectations of Somalia ever becoming Switzerland, of finding world-class sprinters among the Ashkenazi, and will not wait for women to build the next Brooklyn Bridge, or fill the Louvre and Prado with masterworks. It’s not a prejudice that bows to biology it’s reality.

"It's not useful" to worry about single-factor causation in a situation with multiple, interacting influences, none of which is clearly dominant.

How right you are but you are right in a small way, in a personal way, in a human way. We humans, in the episodes that make up our life do not resort to calculations no matter our reputations. We intersect with each other at... well, at a human level. For however we are we are always intimately aware of the other’s humanity. My engagement with a daughter, a wife, a mother, or even a woman who is a stranger cannot, if I am sane and not evil, distill into anything less than an admission that she is, in the greater and most vital determination, as much as I and quite possibly more. But when we look at the bigger picture, the panorama, the great prospect, there, I believe, in that tableau, biology is in the forefront.

Posted by: George Pal at August 25, 2014 01:39 PM

I apologize for abandoning the previous discussion. Life Intruded™.

I did want to pop-in on this because this is also my day job (same question, different industry).

This is an area that actually trips up many a statistician (especially those modellers who are as much in love with the math as the question to be answered): Predictive v/s Explanatory models.

In an explanatory model it is the goal to say what caused a certain result (the result itself is secondary). That is: my objective is to say "If I change X by Y amount holding everything else constant, the result changes by Z". This type of model requires all predictors to be independent of each other (eg. Paint brand and weather conditions: Behr does not make it rain more often than Glidden).

Explanatory models break down when the predictors are correlated. You can't tell, at least not directly, how much of the change is attributable to each of correlated factors: X + Y = 2. X can be 1 (and thus Y = 1) or X can be 4 (and thus Y = -2). Both of these solutions work equally as well. Your guess is as good as mine which one is the correct one. This model is darn near useless for explaining how important X or Y is. This is the problem with explanatory models of the "neuropsychosocial" type. You can't get independent variables. Changing one's attitude necessarily changes the attitudes of those around you. You can't detangle the effects of the subject's positivity from the positivity of those around them. It's simply not possible to "Hold all else equal".

Predictive models have no such issue. Here, the purpose isn't to make inference on the predictors. What happens to the result as you change the subjects attitude by X amount while holding the attitude of everyone around the subject constant? Don't know. Don't care. Doesn't matter. Upbeat subjects surround by upbeat people do X and depressed people surrounded by depressed people do Y.

This isn't "obtuse/hasn't a clue". Neither is it pleading that "it's just too hard to understand".

It's a matter of deciding what's important (explanation of the causes v/s prediction of the results) and picking your tool accordingly.

Posted by: Yu-Ain Gonnano at August 25, 2014 03:08 PM

YAG,

This isn't "obtuse/hasn't a clue"

Very well then, it's a flimflam scam in search of more grant money.

From the link, in chronological order and escalating humbug:

The word “neuropsychosocial” does away with the common nature/nurture divide, and so did the researchers. Rather than asking whether teens abuse alcohol because of social influences or innate biology, the scientists looked at those variables that could be measured, regardless of whether the variables were social, biological, or a mix of both.

Humbug #1 - “neuropsychosocial”.


It’s important to keep in mind that the study was not designed to discover which factors cause teen drinking, a much more difficult task. Instead, the researchers focused on what they could measure—an approach that, while it has its limits, is one of science’s most successful strategies.

Humbug #2 - researchers focused on what they could measure... one of science’s most successful strategies.


With these results, we can we say about the biological basis for the social phenomenon of teen alcohol abuse? Not much more than, “it’s complicated.”

Humbug #3 – with these results... it’s complicated.


I embark now on a study of human migraines and will measure all measurables including how many times sufferers had come across the word “neuropsychosocial” or phrases such as “neuroempathy disorders” in their lifetime.

Posted by: George Pal at August 25, 2014 04:19 PM

George:

At the risk of getting into trouble for teasing you (she said with a smile) I think you're doing the classic "thinking like a man" shtick :p

One thing I see men doing far more than women (and yes, I do know women who do this) is simply pronouncing something to be true without actually demonstrating that it *is* true.

Since you mention migraines, allow me an on-topic digression.

For many years (I didn't have migraines - or at least didn't have them often - until I hit about 30) everyone I knew sent me articles, the theme of which was always, "Doing more of/completely avoiding this one simple thing completely cured my migraines."

And I would try doing more of/completely avoiding that one thing. And it didn't not cure my migraines.

After well over a decade of this nonsense, I finally realized the problem: there isn't one simple thing that makes me have a migraine. In fact, I've identified 8 separate things that trigger them:

1. Dehydration.
2. Irregular sleep.
3. Allergies.
4. Sudden changes in altitude/barometric pressure.
5. Stress (as in, sudden relief of).
6. Hormonal fluctuations.
7. Magnesium deficiency.
8. Strenuous exercise.

There isn't a single one of these things that will reliably give me a migraine. Typically, when at least two of them happen at the same time, though, I will get one.

I didn't figure this out for a very long time because the entire world (including my doctors at the time) was telling me that there was "*a* cause", and if I could just find it, I'd be better. Turns out there were many causes.

And all I really needed to know was that, for instance, if I exercise, I have to be make sure I'm hydrated, take extra magnesium, *and* get plenty of sleep. Or if a big front is coming through, I'd better take decongestants, and drink water, and take it easy (this isn't a good time to exercise).

The "model" is still quite useful, and arbitrarily declaring that things are simpler than they really are wasn't useful at all. I've been completely off daily medication for several years now - essentially, since I figured this out. That's after decades of taking toxic medications with bad side effects. So, not humbug at all. The simple model wasn't an accurate one. Nor was it useful, or scientific. It was more like wishful thinking.

Posted by: Cassandra at August 25, 2014 04:36 PM

One thing I see men doing far more than women (and yes, I do know women who do this) is simply pronouncing something to be true without actually demonstrating that it *is* true.

This is like a sentence-long onomatopoeia! :)

Posted by: Grim at August 25, 2014 04:37 PM

I think this particular study is not very helpful. It doesn't seem like a useful mode of trying to combine separate disciplines so as to attain new insight.

On the other hand, I agree with you (Cass) that it's a good idea to try to bridge fields when possible. The answers to human problems (insofar as there are answers) are quite likely to span several traditional fields of scientific study.

Posted by: Grim at August 25, 2014 04:40 PM

This is like a sentence-long onomatopoeia! :)

Well, in my defense it's pretty much impossible for me to prove to anyone what I have observed :p

I did refrain from saying it's universally true; only that I have observed it!

/flouncing off into the sunset

Posted by: Cassandra at August 25, 2014 04:42 PM

I think the study could well be useful for exactly the same reasons figuring out what my triggers are was useful.

If you know that "this combination of factors" materially raises the likelihood of a particular teen abusing alcohol, you can be on the lookout for that combination (and you won't waste time looking for a single causative agent when one doesn't exist).

Posted by: Cassandra at August 25, 2014 04:43 PM

Just as an aside, I resisted the "multiple triggers" theory for quite some time. The big light bulb when on when I was telling my neurologist what I had been doing to try and get past the exercise problem and how nothing worked, and I said, "It's almost like there has to be a confluence of factors".

Now, that's mainstream theory on migraines. It also neatly explains why I had so many severe headaches, and why they were so resistant to medication.

Nothing argues for success like success.

Thanks for the explanation, YAG. I enjoyed reading it :)

Posted by: Cassandra at August 25, 2014 04:47 PM

Very well then, it's a flimflam scam in search of more grant money.

Again, not so. Identifying who is at risk for a certain outcome without needing to attribute causes is a perfectly legitimate endevour. We see this, and you likely accept this, when it comes to crime. Where are the high crime neighborhoods? Poor socio-economic status, high drug utilization, low-education are all markers for higher crime areas. Which, if any, are causative? Would alleviating poverty by X% return a Y% decrease in crime? Would increasing education by X% decrease crime by Y%?

A predictive model need not address these questions to determine which areas are at risk for higher crime rates. And these questions need not be answered to base decisions on it about where to buy a house or where to put cops on patrol.

"Neuropsychosocial" is simply a buzzwordy way of saying we are building a predictive model and not an explanatory one. But academia is beloved of explanatory models and doing something as low-brow as a predictive model gets you shunned at parties. :-) Sure, it's about securing grant money from the hoity toity, but it's not doing so by less than ethical means.

I'm not sure why "measurability" is a Humbug. If you can't measure it, it's philosophy, not science.

The last Humbug, is really just the first, restated. Sure, "it's complicated", most things are. Making explanatory inferences on predictive models is bad science, but leaving the "Why" of the question unadressed while delivering on the "What" is perfectly good science.

Posted by: Yu-Ain Gonnano at August 25, 2014 05:49 PM

If you can't measure it, it's philosophy, not science.

Strictly speaking, all science is philosophy. It's a subset of natural philosophy.

Posted by: Grim at August 25, 2014 06:06 PM

I’ll not apologize for 'thinking like a man' – after all, what are my options, think like a woman? Wildebeest? Wombat? I ask you.

As for pronouncing something to be true without actually demonstrating that it *is* true, well... I’d hoped I’d been doing the opposite, pronouncing things that had not been demonstrated as true as mere BS. I am disheartened that I have failed, in a credulous world hysterical with idealism, to come across as the indispensable man, the sane man - the cynic.

Posted by: George Pal at August 25, 2014 06:13 PM

I’ll not apologize for 'thinking like a man' – after all, what are my options, think like a woman? Wildebeest? Wombat? I ask you.

Thinking like a human who has more than one tool in the old toolbox?

Yes, I'm kidding. Mostly :) But not entirely.

Posted by: Cassandra at August 25, 2014 06:23 PM

Making explanatory inferences on predictive models is bad science, but leaving the "Why" of the question unaddressed while delivering on the "What" is perfectly good science.

I would argue that it's far better (and less hubristic) science than pretending we know all the answers. Contra our current president, science is far from settled on most questions.

Posted by: Cassandra at August 25, 2014 06:25 PM

The problem is that you haven't made the case for why "Screw the causes, They don't matter. I can predict the effects" is BS.

All you've said is that you won't believe something if the causes aren't explained. That's your perogative to hold that opinion, but you haven't explained why you think it's correct.

Posted by: Yu-Ain Gonnano at August 25, 2014 06:26 PM

All you've said is that you won't believe something if the causes aren't explained. That's your perogative to hold that opinion, but you haven't explained why you think it's correct.

The causes have been explained, spiritually. That’s why it’s correct. No venturing as to why this will not do for the materialists - it is obvious. The philosophers were first great when they asked – what are we? Modernity, true to being modern, i.e., a mode has inundated us with a succession of hypotheses and theories each a welcome fad for too many. One after another they had failed and now, in the gloaming of our discontent we turn to the newest fad – science. What are we? Let me check the genome map. No, something new, the neuromancers. We are our encephalograms!

What had been known about who we are was sufficient to us by the thirteenth century. Then came the modernists. Descartes laments our state, our predicament, our ignorance. That which had been self-evident had come under attack and, some time ago, overthrown. The usurpers now promise to uncover the Rosetta Stone of perfecting our imperfection at any moment. We are, already, less than sacred in their eyes – less than human by extension. When they claim to have found it, it will not be good.

I trust this answer will not meet with your approval for being too broad. But we have been too long at dehumanizing humans and I expect things to get worse when our baneful urges get the stamp of approval from ‘science’.

As for the small episode such as pointed out by Cassandra in this post, allow me a sneer. It keeps me happy and sane. And no-one will be hurt by it as it is not a theory.

Posted by: George Pal at August 25, 2014 08:24 PM

The causes have been explained, spiritually. That’s why it’s correct.

Far be it from me to not allow some things to be taken on faith, but that still doesn't explain why the model's results (a model that, by the way, does not contradict your explanations as it offers none itself) should be discounted.

Posted by: Yu-Ain Gonnano at August 26, 2014 02:23 PM

I had not said faith, I said spiritual - that which is the basis of Western civilization’s understanding of 'human' (body and soul-pneuma) from the time of the classical ancients to the medieval Christian apologists.

Once again:
“the study was not designed to discover which factors cause teen drinking, a much more difficult task. Instead, the researchers focused on what they could measure”.
This is the tripwire that modern science routinely sets off with their studies – some more so than others, obviously.

a model that, by the way, does not contradict your explanations as it offers none itself.

That the 'measurables' are, as often as not, imponderable, they nevertheless become the matter from which theories are formed – and I hate theories – and I smell a theory.

If you should ever be so inclined, you might read APING MANKIND: NEUROMANIA, DARWINITIS AND THE MISREPRESENTATION OF HUMANITY by Raymond Tallis – I highly recommend it. We will not reinvent the wheel; we should abstain from reinventing the human – we’re neither, smart enough, or good enough. There is no next big thing in understanding the human; only in understanding the anti-human.

Posted by: George Pal at August 28, 2014 11:49 AM

“the study was not designed to discover which factors cause teen drinking, a much more difficult task. Instead, the researchers focused on what they could measure”.

On reflection, this does sound rather like the old joke about "looking for my car keys where the streetlights are, not where I dropped them, because I can actually see over here."

Posted by: Grim at August 28, 2014 12:03 PM

But accepting that the causes are spiritual is something that cannot be proven true or false. They must be taken on faith (I did not mean to imply religious faith).

Still not seeing, and your aren't explaining, why focusing on identifying *who* is at risk rather than *why* they are at risk is a tripwire for anything other than practicality.

Nor are you offering any explanation for why "measurables" are "imponderable". I can measure Household Income, # of parents at home, school grades, etc. I can pretty easily ponder all of those. I would agree that theorizing cause and effect might be "imponderable" (in fact I've said that already). It very well may be that teen alcoholism causes bad grades and not bad grades causes alcoholism. But as I'm not trying to construct theories of causes, that doesn't matter.

To return to the crime model: Income, marriage rates, owner/renter mix, employment rate, educational attainment are all markers for crime rate. And yes, the elephant in the room, ethnicity, is as well. The greater the proportion of minorities the higher the crime rate. Adding ethnicity to a predictive model is perfectly legitimate. What you absolutely, positively, in no manner, way, shape, or form can do is then infer that ethnicity causes crime. Your inference is bad and you should feel bad. You cannot change ethnicity of a neighborhood and hold all else constant. Income, educational attainment, marriage rates, all of it, necessarily changes when you change ethnicity. For that reason, no theory about the causes of crime can emerge from a predictive model. Constructing a well executed explanatory model that could be used for such inference is a vastly greater challenge.

Grim, your analogy isn't wrong per se. That said, if you have 10,000 lost car keys and only 10 people to look for them and one hour to find as many as you can, looking for keys where you can't see them (and thus wouldn't be able to find them even if they are there) isn't exactly efficient.

Posted by: Yu-Ain Gonnano at August 29, 2014 09:26 AM

YAG,

“Nor are you offering any explanation for why "measurables" are "imponderable".

I’ll not make a point of your misdirection – perhaps you missed some dots and cannot now connect them. Household Income, # of parents at home, school grades are certainly not imponderable, but neither are they in the classification/category of neuropsychosocial - emphasis on the neuro and psycho (mind).


”no theory about the causes of crime can emerge from a predictive model”

You mean like never? You mean there had never been such goings on? You mean the predictive models of global warming had led to no theories, hypotheses, premises, laws, treaties, flimflam? Had I been imagining the last decade?

I know very well what you mean - not 'can' but ought. You can stop making cases for the efficacy of science, statistics, and methods. You again miss my point which is not that they don’t matter but that science, as presently constituted, is sufficiently corrupt to deem it, at best, unreliable, at worst, treacherous.

To my larger point of the pretense of science (as presently constituted remember) - as it pertains to the human - I cannot put it any more simply than... The map is not the territory, the picture is not the nature, the directions are not the road, the measure is not material – and, ultimately, neither is the 'answer'. For that, the neuromancers will, for me, forever be suspect.

Posted by: George Pal at August 29, 2014 10:54 AM

Neuropsychosocial isn't a classification of "measurables". It's a buzzword that says that the researchers didn't care if the measurable was in the neurological realm (seratonin levels), the psychological realm (% of days feeling depressed), or social realm (% of friends who diagnosed with depression).

They simply said that the distinctions (which overlap in quite substantial ways) were not relevant. They were not going to allocate causes to each of these areas as if they were distinct.

You mean the predictive models of global warming

Global Warming models are not predictive models. They are explanatory models. The researchers are pretty clear up front that they are trying to determine exactly which factors are causes and exactly how much those factors matter. It's why they go through Principle Component Analysis: to yeild predictors that are independent and not correlated. Explanatory models also make predictions, but as they seek to do more than just make predictions more work and vastly more care must be taken. It's a much more difficult task. These can and do work, but they are also easier to screw up (and not always through malice).

My problem with GW models is that I know how much work, documentation, and audit trails matter on the pedestrial predictive models. If I lost the original data on my predictive models like East Anglia did, I wouldn't make it past internal review much less external review. To do so for an explanatory model, with its higher standards, is intolerable. Science isn't faith: "Trust Me, I Did It Right" isn't acceptable.

I get what you are saying by "The map is not the territory" and "the directions are not the road". But that doesn't mean that maps and directions aren't useful. I applaud these researchers for saying "I don't need to re-create the territory, I don't need to discover why the road is here or that stream is over there, but let's draw a map and write up directions so that the next guy can reach his destination easier."

That is an entirely legitimate, valid, useful, and yes, ethical endevour.

I don't know why you think that is sinister.

Posted by: Yu-Ain Gonnano at August 29, 2014 01:06 PM

Neuropsychosocial is more than a buzzword. In its precision, in the ordering of the compound, it makes the neuromancer’s premise palpable: the primacy of the brain; the mind as a subset of the brain; the individual personality as nothing more than a part of the herd, hive, or swarm; the action - the unwilled result of a neurological imperative; the human less than human. The problems, then, will be neither in our stars nor in ourselves, but in our misfiring synapses. I shudder at the utter lethality of this nocuous sciencism.

Global Warming models are not predictive models.

I’ll take that you are correct about the computer models and they are explanatory. But the models are invariably extrapolated to predictive. Scientists have been complicit in the process.


That [drawing maps, writing up directions] is an entirely legitimate, valid, useful, and yes, ethical endevour.

Yes, indeed. But it’s the presumptions with which science too often commences I object to. Draw the maps, write the directions, but it better had not be the road to hell.

I don't know why you think that is sinister.

Once again I haven’t a problem with scientific methods only with the notion that science is beyond good and evil.

To sum up, an example:
MIT professor and pioneer in the computer modeling, Jay Forester - and member the Club of Rome. He and other brights, by most any measure of brightness, assumed the task of planning the future of humanity, and had the computers model the world economy using his techniques. The book THE LIMITS TO GROWTH (1972) predicted calamity resulting from exponential growth of population and industry. Computer-generated graphs drew a picture: shortages, crashes, pollution. It was wrong, monumentally wrong, go-away wrong. And it was wrong not because of data sets, poor modeling, insufficient RAM, it was wrong because it was presumptive; it had presumed...
“The world has a cancer and the cancer is man.” - Club of Rome (Mankind at the Turning-Point)

I think that sinister. I think that – to one degree or another – prevalent.

Posted by: George Pal at August 29, 2014 05:20 PM

Post a comment

To reduce comment spam, comments on older posts are put into moderation 5 days after the last activity. Comments with more than one link also go into moderation. If you don't see your comment after posting it, try refreshing the screen. If you still don't see it, your comment is probably in the moderation queue.




Remember Me?

(you may use HTML tags for style)