Wikipedia:Reference desk/Archives/Science/2009 January 3

= January 3 =

What is it eating?
This Nestor meridionalis meridionalis of the South Island of New Zealand was featured on the Main Page in the "Did You Know..?" section. When I first saw the photo, I thought it was eating a snake and I was consequently inclined to start a civilization; but on closer inspection, I found that the parrot in question doesn't eat snake. I'm not too sure, but that doesn't look like a Longhorn beetle grub, or a flower bud, or anything identifiable to me (maybe a melon fragment?). So, what is it eating? Nimur (talk) 04:43, 3 January 2009 (UTC)


 * Looks like a piece of capsicum to me.Mattopaedia (talk) 05:09, 3 January 2009 (UTC)


 * Given that New Zealand doesn't have snakes and they would be another serious danger to our wildlife, it's a rather good thing it isn't eating one. Even more so on Stewart Island. Nil Einne (talk) 05:29, 3 January 2009 (UTC)


 * Are we sure this is a wild parrot? Perhaps it's someone's pet and it's chewing on a red plastic toy of some kind?  Just think of the kind of lens/camera you'd need to get that kind of a shot of a wild parrot - it doesn't seem very likely to me.  But this is Wikipedia - we're a collection of real people, not some anonymous corporation.  Hence you can go to the information page for the photo (just click on it) - look at the 'history' tab and see who posted the photo.  Now you can go to that person's talk page and politely ask them all about their photo.  This might not work - but on the whole we're a pretty friendly & helpful bunch here and I'm sure the photographer will be only too happy to talk about their (now very famous) picture. SteveBaker (talk) 12:54, 3 January 2009 (UTC)


 * Done. But again, it really looks like a slice of capsicum to me. Mattopaedia (talk) 14:54, 3 January 2009 (UTC)
 * This picture seems to have a circuitous history. I saw the exchange on the picture's most recent editor's talk page, and then followed the trail back to the original Commons entry, which was created by the Flickr upload bot.  The description's link to Flickr appears to be invalid, but searching Flickr yielded what appears to be the original photo: .  HTH.  --Scray (talk) 18:09, 3 January 2009 (UTC)
 * BTW, it's not obvious to me that the Creative Commons license was applied properly by the Flickr user to that photo, as required for use of that bot, but I am no copyvio expert. --Scray (talk) 18:15, 3 January 2009 (UTC)
 * Hmm why did you need to search flickr? The photo clearly identified the flickr source for a while. I followed the source before I posted my first message and I just checked, it's still there and there's no sign from the commons history it was ever removed Nil Einne (talk) 16:08, 4 January 2009 (UTC)


 * Done? You do realise this photo came from flickr right? I don't see any comments on the flickr page. Or did you PM? P.S. Personally I'm more inclined to believe it's some part of a native plant although I can't say what Nil Einne (talk) 16:08, 4 January 2009 (UTC)


 * Nowadays, I don't think many people have kākā as pets (not counting those in zoos or being looked after by DOC staff of course). I'm not even sure if it's legal. I don't think DOC will take kindly to people catching them and I'm not aware of any population of domesticated kākās. Also bear in mind most NZ birds are fairly tame, one of their problems given the lack of predators during most of their past. This bird was on Rakiura as I mentioned above so it's unlikely to have encountered much to scare it. (And it's also more likely you'll actually spot one). Of course you still need a decent camera and to be a decent photographer but these are becoming more common in this digital age. The photographer seems to be an extremely experience wildlife photographer anyway Nil Einne (talk) 16:01, 4 January 2009 (UTC)


 * AFAIK, it's legal to keep any 'pre ban' parrot (or its captive-bred offspring) as a pet. Even hand-fed Keas (which must be the ultimate 'difficult' pet) are available to buy, though I've only ever personally seen two advertised for sale in 15 years or so... --Kurt Shaped Box (talk) 01:17, 5 January 2009 (UTC)


 * Yes that wouldn't surprise me. Do domesticated/captive bred kākās exist? I guess there's probably some although perhaps they're even more rare then keas (when I searched I couldn't find any references) Nil Einne (talk) 11:54, 8 January 2009 (UTC)

According to the photographer, the picture was shot in the wild on Stewart Island; using, as per exif metadata, a Canon EOS 20D + the Canon 100-400mm USM IS telephoto lens. Here is another picture of possibly the same bird with something else in its beak. (By the way, the image license seems fine to me) Abecedare (talk) 00:13, 4 January 2009 (UTC)


 * Could it be a piece of fruit rind/peel? Perhaps the photographer was feeding it in order to get a good shot? --Kurt Shaped Box (talk) 16:12, 4 January 2009 (UTC)


 * Stick of rhubarb? Size looks about right. XLerate (talk) 21:19, 4 January 2009 (UTC)


 * Could it be part of a fuchsia flower? Tree fuchsias are plentiful on Rakiura. dramatic (talk) 22:12, 4 January 2009 (UTC)


 * I think that rhubarb is supposed to be toxic to parrots. Well, it's one of those 'common knowledge' things that you get told when you first purchase a psittacine, at least. --Kurt Shaped Box (talk) 01:04, 5 January 2009 (UTC)


 * Well Rhubarb is toxic to humans, particularly the leaves so it wouldn't surprise me if toxic enough to parrots that it's a bad idea to give it to them Nil Einne (talk) 11:51, 8 January 2009 (UTC)


 * The photographer says he thinks it was a plum Nil Einne (talk) 11:57, 8 January 2009 (UTC)

Do we have any defense at all against meteors, black holes, etc.?
Like, big ones! In all seriousness, if astronomers detected a continent sized meteor headed our way with say, a year's warning, -couldn't we do something? Or if a black hole was heading our way couldn't we use some kind of antimatter type force to push it towards the moon? Or for that matter, if the moon ever broke free (from a black hole?) and started slowly drifting towards us, could we use some greater force to fastly push it away? When our sun expands during it's death throes, how will life forms of that era deal with it?--Dr. Carefree (talk) 09:49, 3 January 2009 (UTC)


 * No problem. We have loads of asteroid deflection strategies.--Shantavira|feed me 10:15, 3 January 2009 (UTC)


 * Um what? Antimatter type force? Our current ability to produce antimatter is very limited and our ability to use it for practical purposes almost non-existant. And I'm not quite certain what the use of pushing a blackhole to the moon is. Plus the concept of us moving a black hole anytime soon, if ever, seem absurd to me. As to what humans will do, billions of years from now if they still exist and live on earth, I don't think many people have given it that serious thought although I'm sure there's something in science fiction Nil Einne (talk) 10:51, 3 January 2009 (UTC)


 * In practical terms, not only do we have no defense - we don't truly know what form that defense should take (because we don't know enough about the meteors themselves) and we don't have the ability to detect meteors early enough to make a difference.


 * Let's look at those problems individually:


 * Detection: With a large mass on a trajectory that's pointing it at us, the earlier we do something, the easier it is. To take an extreme example: Suppose we only have a few minutes warning...we might have to try to deflect a meteor when it's one earth diameter away from hitting us square-on at the equator...we'd have to bend it's path by 30 degrees to make it miss us instead.  To deflect a mountain moving at a thousand miles a second through an angle of 30 degrees in just a few seconds would probably take more energy than we have on the entire planet - it truly can't be done.  However, if we know that the meteor is a threat (say) 20 years before it hits us - then the amount of deflection we need is a microscopic fraction of a degree and a really gentle nudge would suffice to save our planet.  So early detection - and (most important) accurate orbital parameter determination - is a massive priority both because it gives us time (it might take 5 years to put together a strategy for deflecting this particular rock, building the spacecraft and getting it launched towards it's target) and it reduces the magnitude of our response.
 * Analysis: There are many possible categories of threat. Comets are mostly ice.  Meteors come in several varieties - some are essentially solid chunks of metal, others are solid chunks of rock, still others may be loose collections of small bolders, pebbles or even dust.  Right now, we don't know which is which - which ones are the most common - whether large, dangerous objects are predominantly of one kind or another.  We know (for example) that Comet Shoemaker-Levy broke up into a dozen pieces as it descended towards Jupiter - if we'd had to deflect that comet and we'd sent (say) a single large nuclear bomb then a whole range of disasterous possibilities come to mind: (a) The comet might break up before our rocket gets there and we can now only deflect one out of a dozen large, dangerous chunks. (b) Our bomb might actually do nothing more than break up the comet prematurely without deflecting it's course at all.  So until we "know our enemy" - we're kinda screwed.   We need to send lots of probes out to look in detail at a statistically reasonable collection of comets and meteors - and do lots of science to figure out what's out there.
 * Deflection/Destruction: The problem is that breaking up a large meteor or comet without deflecting it's path doesn't help us. The total damage to the earth from a single rock that's the size of a mountain that weighs a million tonnes is precisely the same as for a million car-sized rocks weighing one tonne each or for a trillion rocks the size of basket balls weighing a few kilograms each.  Simply smashing the meteor into pieces doesn't help at all!  (The number of movies that get this fact wrong is truly astounding!)  So we have to think in terms of deflection - not destruction.  If we have enough time (see "Detection" above) then something as simple as a heavy spacecraft that flies along parallel to the course of the meteor for a few years and provides a REALLY subtle gravitational shift - might be enough.  That's great because it works just as well with a flying rubble heap as it does for a mountain of nickle-iron or a million tonne dirty snowball.  However, getting big, heavy things out of earth orbit and flying as fast as a meteor requires a heck of a lot of fuel and a huge amount of up-front planning.  We certainly don't know about these threats early enough to do that reliably.   So then we're left with hitting the thing hard with a big bomb, hitting it hard with an 'impactor' or nudging it more gently with rocket motors.   None of those things will work for a flying rubble pile.  For solid bodies - that'll work.  We can build a probe with a rocket motor on it.  Make a soft-landing onto the object and start firing our rocket.  So long as the object is strong enough to take that pressure without breaking up - or without our rocket sinking into the surface or tilting sideways and deflecting the rock in the wrong direction - that could work.  But it's more complicated than that if the object is spinning (as many of them are) - because now the rocket has to fire intermittantly when the rock is at the correct orientation or else the miracle of 'spin stabilisation' (which is what makes bullets fly in a nice straight line) will frustrate our efforts.


 * So it's safe to say that right now, we're defenseless on all three levels. Our detection ability is getting slowly better - we have surveyed some of the very largest rocks - and we're tracking their orbits.  Perhaps we can now see mountain-sized rocks soon enough - but something a lot smaller than a mountain (like maybe a school-bus-sized rock) can take out a city - and we're nowhere even close to being able to track those soon enough or accurately enough.  NASA have sent out probes to several meteors and comets to take a close look at them - and we've even tried firing an 'impactor' at one them...but we have a long way to go.  A lot of people are thinking about deflection/destruction strategies...but no governments are building rockets and putting them into storage ready for the day when we'll need them - and funding for the entire process is distinctly lacking.


 * At some point we (as a species) need to seriously consider having a colony somewhere away from the Earth. There is always the possibility of the ultimate planet killer coming along that's too fast, too large, too unstable and/or too close to do anything about.  Having a colony of humans living on (say) Mars with a self-sufficient life-style and a large enough gene-pool is the ultimate way to ensure the survival of the species come-what-may.


 * SteveBaker (talk) 12:43, 3 January 2009 (UTC)


 * If you'd like to learn more about it and help create "coursework", there's some activity on Wikiversity at Earth-impact events/Discussion, as well as an opportunity to use your imagination and research skills for colonizing off-planet at Lunar Boom Town. (And SteveBaker: mind if I copy over what you just wrote above? Good detail there!) -- SB_Johnny | talk  12:56, 3 January 2009 (UTC)
 * All contributions to the ref desk fall under the GFDL - so long as you are using them under those same terms, you are welcome to take whatever you need. SteveBaker (talk) 13:45, 3 January 2009 (UTC)
 * I've copied this section to Earth-impact events/Discussion, if anyone wants to explore this topic in more detail. --mikeu talk 16:52, 7 January 2009 (UTC)
 * Surely it would help to break it up as small objects can be burned up in the earths atmosphere, the amount of mass burned off an object will be proportional to the surface area of the object, which dramatically increases if the object is broken up. To take my point to its logical conclusion, an asteroid of a large given mass would cause significant damage, whereas the same mass of dust colliding with the earth would most likely just flouresce in the atmosphere as it "burned". —Preceding unsigned comment added by 84.92.32.38 (talk) 14:12, 3 January 2009 (UTC)
 * The net kinetic energy that has to be absorbed by the Earth system (atmo included) remains exactly the same. While a single large rock would probably cause more damage (at least more localized damage) based on impact, even the distributed vaporization of a massive asteroid would be a catastrophe.  A school bus sized rock, sure -- vaporize it.  A dino killer?  That won't work. &mdash; Lomn 14:52, 3 January 2009 (UTC)
 * Our main capability in asteroid defense is early warning. Here is an incomplete list of early warning systems that I found:
 * Pan-STARRS a (nearly complete) telescope which will continuously survey the sky for asteroid threats
 * Near Earth Asteroid Tracking (NEAT) An organization which has several dedicated telescopes for asteroid detection and tracking.
 * Spaceguard, a loose affiliation of asteroid detection programs/enthusiasts (including a formal program at NASA).
 * Lincoln Near-Earth Asteroid Research (LINEAR) a NASA/USAF/MIT joint project for asteroid detection/tracking
 * Lowell Observatory Near-Earth-Object Search (LONEOS) a completed 15-year project which detected hundreds of asteroids.
 * Minor Planet Center an organization which is the recognized official center for collecting and publishing orbit data about all known asteroids and comets.
 * Catalina Sky Survey another sky survey designed to identify asteroids.
 * Spacewatch an University of Arizona program to discover near earth objects.
 * Near Earth Object Program (NEOP) a NASA effort to study the list of known asteroids and determine the impact risk.
 * We currently know of around 5500 (see the NEOP page) Near Earth Objects, and hundreds more are being discovered every year. So as you can see, we humans have actually put a fair amount of effort into detecting impact threats from space.  No one has ever tested any potential asteroid deflection systems, but as Steve says, early detection is key.  --Bmk (talk) 15:44, 3 January 2009 (UTC)


 * That doesn't seem quite right... if gazillions of little pieces vaporize in the atmosphere or even hit the earth (somewhat more slowed down by atmospheric friction), they wouldn't also vaporize large amounts of actual earth materials as well (which would cause snowstorms in Havana, etc.). Am I missing something? -- SB_Johnny | talk  15:33, 3 January 2009 (UTC)


 * This is little as in much smaller than a mountain, not little as in the size of a pebble. The pieces will still be much too large to burn up in the atmosphere. — DanielLC 19:11, 5 January 2009 (UTC)

(edit conflict)A meteor the size of a continent would be larger than the Moon. The largest asteroid we know of, Ceres, is only about 1/4 the diameter of the moon, and it's big enough to be called a dwarf planet. There are plenty of other potential risks, for example a rouge star passing through the inner solar system would disrupt the trajectory of many asteroids and comets, flinging some toward Earth. The chances for a large black hole to pass through the solar system are very small, since it would be more massive than the sun but smaller than New York. Small black holes could evaporate to Hawking radiation. Remember that the asteroid/comet that probably caused the Cretaceous-Tertiary extinction (the one that killed the dinosaurs) was only about 15 km (9 miles) in diameter. As for the sun expanding, we could maybe move to Mars, but chances are we won't even survive that long. Most estimates predict there is a "high probability" for us to become extinct within the next three million years or so. Anyway, an asteroid bigger than, say, Lake of the Woods, would probably crash through the Earth's crust, exposing its mantle, causing further problems. Anyway, there are plenty of other potential doomsday events that could affect us in the near future (try exitmundi, [warning, popups]), and many of those pose a larger threat to us than the likelihood of an asteroid hitting Earth (which will, with 100% probability, eventually happen). In fact, one potentially catastrophic scenario is already unfolding, and could affect us in our lifetime, yet many are refusing to do anything about it. It's called global warming. ~ A H  1 (TCU) 15:35, 3 January 2009 (UTC)


 * The sun won't just slowly expand to engulf the earth - it'll put out huge pulses of hard radiation and do all sorts of other ill-behaved things along the way. There is probably no place in the solar system where we could survive that event.  However, the sun blowing up is a fairly predictable event.  We can give a fairly accurate prediction as to when that'll happen - and doubtless if our ancestors survive that long - they'll know to a very accurate degree when this is going to happen.   So they'd have time to do something to escape to another solar system.  The problem with meteors and comets is that they are more random - and the best response is to try to deflect them somehow.  We'll certainly have a few thousand or even millions of years notice that the sun is going to give up on us - but we'd be lucky to get 20 years notice of an earth-smasher en route.  Black holes are not worth bothering about - there are no big-but-slow ones nearby - and we don't care too much about small ones.  Fast-but-big ones are impossible to deal with - if one of those comes by, there is nothing we can do.  Meteors and comets are in the middle ground - they DO wipe out entire ecologies (the Dinosaurs - and there was a report out a couple of days ago suggesting that the Clovis people of North America were wiped out by a comet/meteor impact) - and we could do something about it with present-day technology if we put our minds to it.  The odds of you or I personally being wiped out by one of these things is tiny - but it's one of the top risks for humanity as a species - so I think we should spend commensurate effort on solving the problem.  Global warming should be higher on our list - but comet/meteor protection ought to be up there in the top ten goals of humanity over the next 100 years. SteveBaker (talk) 02:31, 4 January 2009 (UTC)

We have one rock-solid defense against asteroids: Bruce Willis. —Preceding unsigned comment added by 79.122.10.173 (talk) 14:12, 4 January 2009 (UTC)


 * There is a new Asteroid Deflection Research Center that has been created to develop asteroid deflection technologies. You might also be interested in reading the news at the PlanetaryDefense blog.  --mikeu talk 16:43, 7 January 2009 (UTC)


 * Shoemaker and Szabo took my idea already of nuclear-powered steam. -lysdexia 00:59, 9 January 2009 (UTC)

Pesticides
I was looking at pesticides recently and one thing which surprised me was the number of different products which are the same brand and seem to be the same thing.

For example there were a number of different products which were 12g/litre of Permethrin in the form of an emulsifiable concentrate, with the same total volume. Prices varied by a few cents in some instances. These were nominally intended for a different purpose, e.g. No Silverfish beetles, No cockroaches, No ants with appropriate instructions as to how much to dilute them (all with water only) and how to apply them although they sometimes gave instructions for some other purposes. Am I right that these are almost definitely the exact same thing but with a different label? Or is it likely they have other ingredients to aide in how they adhere to surfaces or whatever. There was one for spiders which was a higher concentration (50g/litre) which I can see the need for.

There were also a bunch of ready to use sprays for similar purposes most of which were 4g/litre of permethrin in the form of a ready to use liquid (why they recommended different dilutions for the concentrates but the RTU liquids are the same I'm not sure). These were the same quantity and IIRC the bottles were similar, I don't think they sprayed differently or anything. The prices varied more by about $2 or so. Similar to above, there could be differences in surfactants etc and particularly in this case what the containing liquid is although at least two of them said "Will remain active on inert surfaces for up to two months". Or is it likely these are more or less exactly the same product?

I do have photos of some of the products and you can also see them online e.g.    and note they are registered as different products although under the same code HSR000265 (mentioned on the bottle) for the concentrates or HSR000263 for the RTU liquids.

Finally another thing I've noticed is certain wasp powders e.g. have Permethrin in the same concentration (10g/kg or mg/g) as flea powder for carpets. Beyond perhaps difference in particle and in applicators on the bottle, am I right these are more or less the same thing? (Some are in higher concentration but I'm guessing again there's little difference otherwise). Just to be clear I'm not referring to the stuff meant to be applied to pets, which are probably regulated differently (here in NZ they regulated as veterinary medicine as opposed to the carpet powder, wasp powder etc which are regulated as pesticides HSR000262).

P.S. It's possible some of the above have different cis/trans ratios for their Permethrin but any that did mention the ratio were 25:75 Nil Einne (talk) 11:55, 3 January 2009 (UTC)


 * It's certainly possible that the surfactants might be different for products labeled for indoor, horticultural, and veterinary uses. It's also possible that it's simply a marketing decision. I'm not sure how the regulators in NZ define the codes, but I know for the OMRI (wow, no article! OMRI rates products for the National Organic Program in the US) ratings in the US, it's a product-by-product registration. -- SB_Johnny | talk  13:01, 3 January 2009 (UTC)

Kale
Why does kale taste so good? Does it have lots of iron or something? Because personally, I can't get enough of it. I can't understand why anyone wouldn't want to 'eat their greens'. There are people that I used to know who never eat fruit or vegetables of any kind, almost. I can understand fruit, because it's too sweet, but vegetables? No way, there's loads of good vegetables. Something like kale almost tastes better to me than any other food.--Veritable&#39;s Morgans Board (talk) 16:16, 3 January 2009 (UTC)


 * Our taste buds are each a little different; like human fingerprints, it's likely that no two people have the exact same taste for food. Hence, some people love the taste of certain foods more than others, because that particular food..well, arouses the taste buds, I guess you could say.


 * Another reason you might not be able to get enough of it is if it evokes pleasant feelings. If the first time you ate Kale was when a beloved relative served it, it may help; especially if that person is deceased and it helps you keep their memory alive. (And, if it was early enough in your childhood, you may not recall specifically that this is who served it to you first. For instance, I associate spaghetti o's with my geat grandmoth, because she always had some chef boyardee food around when I'd to see her.) It sounds like an unusual reason for something to taste good, but it's all part of how amazingly interconnnected the body is.Somebody or his brother (talk) 16:51, 3 January 2009 (UTC)


 * I should point out that most of what we call 'taste' is really 'smell'. Our taste buds give us only fairly crude information about flavor.  (That's why things taste different - or "like cardboard" when our noses are stuffed up with a heavy cold.) SteveBaker (talk) 18:05, 3 January 2009 (UTC)


 * And overtly linking this to the OP, perception of smell is highly personal. This individuality is likely to be based on both experiential and genetic factors.  --Scray (talk) 19:56, 3 January 2009 (UTC)
 * If I recall right, there's also been some research done that found something specific in dark-green vegetables (broccoli, spinach, kale, etc) that divides people. Some people are genetically far more sensitive to the taste of certain compounds in them and hence find them much more bitter (and hence more disgusting). ~  mazca  t 20:00, 3 January 2009 (UTC)
 * I'm quite convinced that's true of fish. So many people that I know will rave on about how wonderfully 'fresh' a particular lump of fish is - when I can barely taste the stuff at all - let alone determine freshness.  Sushi is just kinda slimy and uninteresting for me - but my wife is a fanatic for the stuff - detecting amazingly subtle differences between one restaurant and the other.  On the other hand - I'm pretty good at identifying types of wine and beer - so my taste mechanism obviously works well at some level.  We know of all sorts of genetic differences that cause 12 different varieties of color-blindness - why should we be surprised at fishyness-tastelessness and veggy-tastelessness? SteveBaker (talk) 02:18, 4 January 2009 (UTC)


 * I'm afflicted with the ability to taste something nasty (metallic) in cilantro. I share this useless superpower with one parent and not the other. —Tamfang (talk) 19:59, 5 January 2009 (UTC)
 * I think the usually difference discussed is the dislike of some food by supertasters, not taste-blindness. But loss of tasting ability occurs with aging or with ageusia (never heard of that problem before). Rmhermen (talk) 00:59, 6 January 2009 (UTC)

"Blind spot" in face recognition?
Is there a name for the phenomenon in which a subject has difficulty telling apart specific pairs of faces (of different individuals) when most people in the general population have no difficulty telling one face from the other in those pairs? (The subject in question doesn't have a general problem recognizing or telling apart faces, only difficulty w.r.t. specific, idiosyncratic pairs.) —Preceding unsigned comment added by 173.49.15.111 (talk) 18:22, 3 January 2009 (UTC)


 * See Prosopagnosia. Of course, you may not recognise it, as it looks like thousand other articles.  --Cookatoo.ergo.ZooM (talk) 19:28, 3 January 2009 (UTC)


 * Thanks for prosopagnosia reference, but the kind of confusion I was talking about is both selective and idiosyncratic, in that the subject generally has no problem recognizing and telling apart faces, except for specific pairs that most people won't find particularly similar or confusing. --173.49.15.111 (talk) 20:58, 3 January 2009 (UTC)


 * It's perfectly possible that this is a mild form of that condition. It seems that there is specific 'circuitry' inside the brain that is specialised for facial recognition.  Prosopagnosia is a complete and utter failure of that circuitry - but it seems reasonable that there might be a partial failure that might make (say) recognising the shape of the mouth and nose work just fine - but eyebrows, eyes and ears fail miserably.  It's tough to know - but 'mild' or 'partial' prosopagnosia would probably be acceptable terminology here. SteveBaker (talk) 02:10, 4 January 2009 (UTC)


 * The ability to recognize faces doubtless runs in a continuum, rather than an all or none distribution. Some of us are at the 95th percentile and others at the 5th percentile. Those at the 5th percentile may function pretty well, but be lousy as eyewitnesses, or as a doorman, receptionist, bartender or salesman who is expected to greet "regulars" or club members by name. Likewise a clergyman, teacher, policeman, bounty hunter or politician would benefit from being at a high level of memory for/recognition of faces. A workaround might be to remember verbally that "bushy eyebrows" is Mr. Smith, or similar cues, where someone good at facial recognition would just automatically recognize Mr. Smith. A psych experiment published in a journal a few years ago (I do not have the cite at hand) showed the power orf the normal person to recognize faces. The experimental subjects looked through an unfamiliar high school annual one time, looking at each face once. They then showed a good ability to distinguish faces they had seen from unseen faces from other similar annuals. At the other extreme, a clerk might see a person 10 times over a month and not recognize them the next time. Edison (talk) 05:28, 4 January 2009 (UTC)

Bolbidia
The following is from Aristotle's discussion of cephalopods: ''One of them is nicknamed by some persons the nautilus or the pontilus, or by others the 'polypus' egg'; and the shell of this creature is something like a separate valve of a deep scallop-shell. This polypus lives very often near to the shore, and is apt to be thrown up high and dry on the beach; under these circumstances it is found with its shell detached, and dies by and by on dry land. These polypods are small, and are shaped, as regards the form of their bodies, like the bolbidia.'' I can't seem to find anything about what a "bolbidia" is. Does anyone know? 69.224.37.48 (talk) 20:34, 3 January 2009 (UTC)
 * Most commentators on the Historia animalium seem to assume that a bolbidion is the same beastie as the bolitaina mentioned by Aristotle in the passage immediately preceding the one you quote. In any event, I think the best one can say is that it's some sort of small octopus; trying to identify it with any particular species is probably futile, especially since the word is apparently attested only here and in the Hippocratic corpus. Deor (talk) 00:24, 4 January 2009 (UTC)


 * The ancient Greeks didn't bother much with careful observation of nature or experimentation or things of that ilk. They felt that if you couldn't just think it up out of fresh air and prove it with some kind of math - then it wasn't worth considering.  So it's likely that Aristotle's observations on marine life were sketchy to say the least! SteveBaker (talk) 02:06, 4 January 2009 (UTC)
 * Actually, Aristotle was one who did bother with observation, particularly with regard to marine life. He used to hang out with fishermen to get specimens. The problem is that he, and the fishermen, knew exactly what βολβίδιον denoted and we don't (other than that it's a small octopus), since the art of taxonomic description hadn't been invented yet. Deor (talk) 03:41, 5 January 2009 (UTC)

World steel supply irradiated ?
I saw this statement in the Wiki article Scuttling of the German fleet in Scapa Flow:

"The remaining wrecks lie in deeper waters, in depths up to 47 meters, and there has been no economic incentive to attempt to raise them since. Minor salvage is still carried out to recover small pieces of steel that can be used in radiation sensitive devices, such as Geiger counters, as the ships sank before nuclear weapons and tests irradiated the world's supply of steel."

Regarding the statement about irradiated steel, is this true ?

Thanks,

W. B. Wilson (talk) 20:35, 3 January 2009 (UTC)
 * I've heard it elsewhere, so it is plausible, but it involved very low levels of radiation. I do not recall the source. Edison (talk) 20:52, 3 January 2009 (UTC)
 * Google steel radiation battleship and you get some sources: IEEE mentions pre 1945 battleship steel as a bulk shielding material for delicate experiments detecting "cosmogenic neutron flux." Other materials of interest to such researchers include 400 year old lead. Another reliable source says that at many U.S. Dept of Energy sites, pre WW2 steel is used for shielding. Edison (talk) 20:59, 3 January 2009 (UTC)
 * Thanks! W. B. Wilson (talk) 21:11, 3 January 2009 (UTC)


 * Wow - that sounds AWFULLY bogus. I don't believe this is the reason.  Sure, there obviously IS a reason for using that old steel - but I can't believe it's because of nuclear weapons.


 * Surely the amount of nuclear-weapon-derived radiation irradiating that 60 year old steel between the time you pull it out of the ocean and the time you form it into an instrument is comparable to the amount that modern steel picks up during the brief time between smelting the ore and casting it (remember - that iron ore is millions of years old and has been protected from atom bomb test contamination by being buried under hundreds of feet of dirt - which has got to be as good protection as 40 feet of water). The amount of time that the metal is above ground and exposed to nuclear waste contamination is going to be pretty comparable in either case.  If the metal can pick up contaminants between digging the ore and making it into steel - then the finished instrument is going to be totally useless after just a few weeks because it'll pick up that same contamination during daily use.


 * It just doesn't make logical sense.


 * I'd be more inclined to suggest that steel that's been shielded from naturally occurring radiation by a large amount of water would have reduced amounts of radioactivity simply due to the half-life of whatever radioactive elements are there naturally. This would result in any radioactivity naturally present in the original iron ore having dropped off considerably over the past 60 years.  Meanwhile, iron ore that's been buried in the ground (which produces background radiation from natural occurring uranium, radon gas, etc) could maybe have background levels of radiation that are unacceptably high.


 * I don't really see how atom bomb testing can have very much to do with it...but I could easily be wrong. If I am wrong - then I'd still lay odds that the fallout from Chernobyl was far more significant than those old bomb tests.


 * SteveBaker (talk) 02:02, 4 January 2009 (UTC)
 * Did you take a look at the sources I cited and other reliable sources from the Google search? Apparently fallout or other products of nuclear explosions since 1945 do in fact make their way into modern steel. It is not just the age of the steel per se, or the fact that it was in the ocean. Atmospheric nuclear tests after WW2 and Chernobyl may indeed have done more contamination than the 1945 tests and attacks, but that still leaves modern steel less useful for shielding sensitive detectors than steel from the pre-nuclear age. Edison (talk) 05:25, 4 January 2009 (UTC)


 * Yes - I did look at them - and I agree that this is what they seem to be saying - but if a reference says that 2+2=5 - then I'm going to have to at least stop and question it. This explanation doesn't make any kind of sense at all.  I have no problem believing that modern steel is inferior - I just can't see how it can be artificial 'fallout' contamination that's causing that.  I don't believe that those nuclear tests contaminated iron ore buried in mines that are hundreds of feet below solid rock...that just can't be true.  It's also somewhat hard to believe that none of that fallout wound up floating down through 40' of water and landing on the decks of those sunken ships - or washed down into the ships from nearby rivers and beaches.


 * So if the ore is pristine when it's dug up - and even assuming that the metal from these wrecks is also pristine - we're only left with the time interval between mining our ore - refining it into steel and forging it into whatever the end user needs - during which time it's gonna get contaminated with all of the crud in our atmosphere and lying around on the surface layer of the soil, etc. But that's got to be comparable to the time between raising a lump of 1945 warship to the surface - cleaning off the barnacles - and reforging that into whatever the end user needs...and during that time, it's picking up contaminants at pretty much the same rate as the freshly mined ore.  So it seems like there would be no advantage to using 'old' steel versus 'freshly-dug-up' steel.


 * I could understand not wanting to use steel made from recycled 1960 Cadillacs that have been sitting out in a scrapyard near ground-zero in the Arizona desert.


 * So - yeah - I could be wrong - and your sources certainly suggest that - but I still don't understand how that's possible. I strongly suspect there is more to this than meets the eye.


 * SteveBaker (talk) 06:30, 4 January 2009 (UTC)
 * It is possible that steel which was melted and re-forged would have the trace amounts of radiation throughout, but old steel which has not been molten may be immune to this radiation. Since steel does not occur naturally, the only way to get "clean" steel is that which was forged in the past before this "contamination" took place.
 * I have also heard this in the past, and I am unsure of its basis in truth, but that is the theory that popped into my head. Running On  Brains  09:33, 4 January 2009 (UTC)


 * The reliable sources (specifically, the IEEE paper), do not cite reasons why the "pre-1945 battleship steel" is preferable or different - so there's a leap of logic to assume that this material has in any way been affected by nuclear testing. It may be different for a huge variety of reasons - different composition, different testing standards, something to do with aging process, ... etc.  Nimur (talk) 14:32, 4 January 2009 (UTC)


 * The Health Physics journal does however. How reliable it is, I don't know Nil Einne (talk) 15:35, 4 January 2009 (UTC)


 * Not sure about the reliability of statements about radioactivity from atmospheric contamination, but I'm willing to say that they're plausible. Iron ore is usually refined in a blast furnace, which strikes me as a very efficient way to move very large volumes of modern, radioactives-contaminated air through the ore.
 * Meanwhile, the level of (radioactive) atmospheric carbon-14 just about doubled in the mid-1960s due to aboveground nuclear weapons testing. (Levels are still about 10% above 'natural' background carbon-14, which is generated mostly by cosmic rays.)  Measuring the abundance of (excess) carbon-14 in tooth enamel (which lasts essentially for a lifetime, once laid down in one's adult teeth) and comparing to atmospheric abundance of carbon-14 is a recognized forensic technique  for determining the age of human remains.  It's accurate to better than plus or minus 2 years.
 * With a half-life of about six thousand years, the level of carbon-14 in the pre-WWII steel won't have been significantly reduced by radioactive decay. That said, there are no doubt other radioisotopes released by the aboveground testing.  I don't know what they are, or what their relative abundance is, or whether they would also find their way into steel.  Just food for thought. TenOfAllTrades(talk) 16:08, 4 January 2009 (UTC)


 * Right, the problem isn't whether the iron ore is contaminated. It's whether the residual contamination in the atmosphere is enough to make its way into any steel processed in our modern, somewhat-contaminated atmosphere. And nobody is claiming the steel is glowing green—it's very minor contamination that only comes into play involving instruments that are sensitive enough to detect such contamination. And Steve, I think you underestimate the amount of residual, long-term fallout produced by the US and Soviet atmospheric testing programs, which detonated probably around 400-500 megatons worth of weapons from 1945-1963. It is significant and measurable amount. --98.217.8.46 (talk) 19:51, 4 January 2009 (UTC)


 * There was a time when children enjoyed "snow cream" made of snow, sugar, milk, eggs and vanilla. Then in the late 1950's or early 1960's parents stopped making it because they read about the danger of strontium 90 in the snow from atmospheric nuclear testing . If steel was recycled, air used in the process would have brought these same nucleotides into the steel, certtainly in greater amounts than in 1944. If new steel is made from iron ore, there is no way to exclude the contamination laying on the ground when the ore is mined by blasting away the earth and digging up the ore. The difference is probably between a very low level in modern steel and a vanishingly small level in pre-1945 steel. Surface contamination would not put the radioactive particles in the bulk of the steel. Edison (talk) 23:04, 4 January 2009 (UTC)

sexual reproduction would never work.
God and I disagree about whether human reproduction works. The idea that you can combine two people's DNA to get a third working human is ridiculous. Would you combine two pieces of software, taking half the bits from one, half the bits from another, to get working offspring software? No... It would segfault as soon as you ran it. I estimate that fewer than 1 out of 100 humans would be born alive and well if DNA were really being 'combined' from the mother and the father. —Preceding unsigned comment added by 79.122.54.197 (talk) 23:37, 3 January 2009 (UTC)
 * Do you have a question? Deor (talk) 23:48, 3 January 2009 (UTC)
 * Yes. I'd like to clear up my confusion, hence sharing my arguments with you.
 * (To original Anon poster): Reality would disagree with you. -- Flyguy649 talk 23:50, 3 January 2009 (UTC)
 * I disgree strongly with reality on this point. The fact that it happens doesn't imply that it's possible.
 * But that's the definition of possible. When something can and does happen. --Russoc4 (talk) 00:07, 4 January 2009 (UTC)
 * then I suppose a woman u.s. president is not possible, since you said possible is what can and does happen. —Preceding unsigned comment added by 79.122.54.197 (talk) 00:52, 4 January 2009 (UTC)
 * I guess that wasn't what I meant. I mean that action implies definite possibility. It is possible for there to be an African American president of the US... it happened. --Russoc4 (talk) 01:23, 4 January 2009 (UTC)
 * it will happen in about 17 days. Didn't happen yet. In any case, you need to specify that anything that is, is possible, but that just because something isn't, doesn't mean it's impossible. - Nunh-huh 01:29, 4 January 2009 (UTC)
 * Comparing DNA to software is a false analogy. This is where your logic is going wrong. Read up on meiosis. Also, sometimes mixing DNA can go wrong, see: nondisjunction. --Mark PEA (talk) 00:02, 4 January 2009 (UTC)
 * So then if the current understanding of reproduction is wrong, what do you suggest is going on? --Russoc4 (talk) 00:06, 4 January 2009 (UTC)
 * To clarify Mark's answer, reproduction is not like cutting half out of each parent and sticking it together. Firstly, DNA replicates. The replicated versions of DNA are what go on to form a new human. I'm not convinced you'd understand if we went into detail here since you don't understand the basic concept, but there are plenty of good explanations out there if you search Google. —Cyclonenim (talk · contribs · email) 00:30, 4 January 2009 (UTC)
 * thanks for watching out for my interests, but why not try me. anyway DNA is just like software, the genetic code is completely equivalent to 2 bits per base pair -- there is NO OTHER INFORMATION it contains.  Anyway you're technically right that replication happens before combination, since the reason those haploid cells exist in my testicles or your fallopian tubes at all is because they have reproduced from other cells.  But then, when my sperm hits your egg, God would have these two haploid cells combine into a duploid cell (before starting to split etc), through sexual combination.  Which is patently ridiculous -- otherwise you could just take a haploid version of two pieces of software, combine them willy-nilly, and get a new piece of software.  Nice try God.  —Preceding unsigned comment added by 79.122.54.197 (talk) 00:42, 4 January 2009 (UTC)
 * as an aside, your statement "there is NO OTHER INFORMATION it contains" is only approximately true. see epigenetics. - Nunh-huh 01:33, 4 January 2009 (UTC)
 * Don't tell God what to do with his software.  bibliomaniac 1  5  00:49, 4 January 2009 (UTC)
 * (edit conflict)Another analogy: science has trouble explaining how bees can fly. ~ A H  1 (TCU) 00:51, 4 January 2009 (UTC)
 * are you saying you disagree with God that bees can fly? What does that have to do with my question?  Start your own thread!
 * Most software is a waveform, DNA is matter. They are not "just like" each other.--OMCV (talk) 01:12, 4 January 2009 (UTC)
 * DNA encodes genes. the code is almost binary.  —Preceding unsigned comment added by 79.122.54.197 (talk) 01:15, 4 January 2009 (UTC)
 * DNA can exist without hardware, computer software can not. There is a ton more INFORMATION in a DNA base pair than two bits.--OMCV (talk) 01:21, 4 January 2009 (UTC)
 * I strongly disagree. There is NO more information in a base pair than two bits because physically, chemically and in every other way two atoms of the same isotope of an element are identical.  Two adenine bases are quite utterly indistinguishable from each other (unless there is some weird isotope present in one or more of the atoms - and I REALLY don't think that codes for anything special).  Hence there is no place for more than two bits of information to reside.  Your claim is bogus - wrong, wrong, WRONG!  Furthermore - the DNA *IS* the hardware - no different from a punched card, a flash memory chip or a hard drive in that regard.  It's meaningless to say that the hardware can exist without hardware.  With appropriate DNA synthesis techniques, we could store Wikipedia on a DNA molecule with two bits per base-pair.
 * The information contained in a DNA strand is independent of the 'hardware' DNA molecule (we did the human genome project thing and stored the information from the DNA strand onto a bunch of CD-ROMs). That's no different from the information contained in a flash-memory chip.  The distinction (and it's a pretty subtle one) is that "software" is the information that's stored on a chip or a disk drive someplace and "DNA" is the storage mechanism itself.  In the case of computers, we can copy the software easily from one piece of hardware to another - and in the case of the information on the DNA strand, we can copy it (painfully) by gene-sequencing onto different hardware (a CD-ROM maybe) - or copy it easily by the rather brute-force approach of making an exact copy of the "disk drive" that the information is stored upon.  Sure, we slightly blur the terms "DNA" and "Genome" where we rarely blur "RAM Chip" and "Software".
 * Can DNA exist without hardware? No - because the DNA *IS* the hardware.  Can we store the information that's stored on the DNA without the DNA hardware?  Yes!  We already did that with many species when we gene-sequenced them - but no information can exist without being stored SOMEHOW in either matter or energy.  Can we store software without hardware?  No - just like the information on the DNA, it needs some sort of hardware to hold it.  But that could be handwriting one a piece of paper - or photons shooting down an optical fibre.  Your distinction is bogus - and grossly misleading. SteveBaker (talk) 03:44, 4 January 2009 (UTC)


 * Every base pair has vastly more information attached to it than the most complex computer subroutine. Lets remember that its not possible to simulate a single atom in a system containing more than one electron.  There are thirty or so atoms and hundreds of electrons included in every nucleotides containing information about potential bonding and countless other physical properties.  The fact that every one of these subroutines is identical and can be grossly simplified to a one of four bits does not negate any of this information.  Its similar to claiming a stick figure or perhaps a nine digit number can accurately represent a person.  At times such a simplification can be useful, for example in XKCD, but its still a simplification and limitations have clearly not been appreciated in this situations.  The question posed here did not concern the ability of DNA, software, or a computers to store data but the functionality of the systems.  Specifically the successful breeding to render functional offspring.  The functionality of data is directly related to the hardware its stored in.  In this case there is DNA, a type of hardware (I agree) with embedded data, versus computer software, which is comprised of bits independent of its hardware.  It is not surprising that a system with an interrelated data/hardware system displays greater functionality, such as self replication, than a software system that simply resides in its hardware.  To look at it from another angle we could also ask why two paper dolls (mostly hardware) can't breed to form a new paper doll while people can.  As I mentioned there are similarities between people and dolls (stick figures) but for some reason most people think that sounds dumb.  That was my point.--OMCV (talk) 05:07, 4 January 2009 (UTC)


 * It's very clear that you don't understand the first thing about information theory (and I happen to have a degree in the subject). And it's too hard and too late for me to explain it to you.  You have not understood the distinction between the "information" and the "substrate" in these two cases.  That blurring of boundaries is throwing off your thinking and causing you (honestly) to talk gibberish. Suffice to say that it's as irrelevant that a DNA base-pair has all of that internal 'state' (due to atoms and electrons and 'stuff') as it is irrelevant that the transistors that make up one 'bit' in your RAM chip is made of atoms and electrons and stuff.  It's not the amount of stuff inside that matters - it's the amount that's stored and reproduced when the information is 'expressed' (copied, transferred, used).  In the case of DNA, it doesn't matter what the spin on the 9th electron of that 3rd carbon atom inside that base pair is because that information doesn't get passed on to the other strand of the DNA as it's copied - and it doesn't cause any difference in how the gene that it's a part of forms a protein molecule.  So just as the precise state of the atomic structure of one bit of your RAM chip doesn't affect how Internet Explorer runs - so the internal state of that base-pair doesn't affect how the DNA works.  Hence two bits per base pair - period.  This is a VITAL property of systems like software and DNA - and it's the reason computers use digital circuits and not analog circuits for running software...if every teeny-tiny quantum fluctuation affected how your software ran - it wouldn't work repeatably and it would be useless.  Similarly, if the precise electron configuration of a DNA base-pair mattered when copying it or expressing a gene as a protein - then the DNA would make different proteins all the time and we'd die within a few milliseconds!   Accurate reproduction and expression is a property of purely DIGITAL systems - and DNA is exactly that - a quaternary-digital system.  Two bits per base pair - period - nothing else matters.


 * As for self-replication. We most certainly can (and do) make self-replicating software (we call them "viruses"...and there is a reason for that!).  It turns out not to be very useful because computers can run multiple instances of a single copy of the code - but that's "an implementation detail" that doesn't affect any of the arguments...if it did, we'd make our software replicate - it's trivial to make it do that.


 * If you still think otherwise then I'm sorry but you simply don't know enough science to comment intelligably on such a technical matter.
 * SteveBaker (talk) 06:07, 4 January 2009 (UTC)


 * I don't have a problem with self-replication. It's the idea of two pieces of software replicating with EACH OTHER that I find ridiculous.  And no, we don't have examples of two different viruses of the same "species" (close enough to reproduce 'sexually') combining with each other willy-nilly, so that if they have fifteen offsprings, each will be a healthy, working piece of software, and also subtely different from its siblings.  It's a ridiculous thought, and I can't believe God would have people combine in such a way.  It's absurd.  As I said, I estimate fewer than 1 out of 100 offspring combined in this way from two people's DNA would be a healthy, functioning human. 79.122.10.173 (talk) 13:49, 4 January 2009 (UTC)


 * It's irrelevant that the code is almost binary, that doesn't make it software. DNA is not made of electrical signals, ones and zeroes and on and off, but chemicals. Not only that, but if we have to compare it to numbers, it's not two different digits it's four (for the four bases). You can't say DNA is like binary code when it simply isn't. —Cyclonenim (talk · contribs · email) 01:24, 4 January 2009 (UTC)
 * Software isn't "made of electrical signals" either - I can store software in my brain (I can recite the quicksort algorithm from memory in 10 different computer languages - so it's not even in 1's and 0's until after I type it into a computer and have the compiler compile it). I can store software as magnetic signals (most of my software is on a magnetic disk at this very moment) - Computer software is transmitted to the Mars rovers every day using a radio link - the software is 'made of' photons for the many minutes it takes to get to Mars...and I could go to any one of a hundred companies around the world and have them store a short piece of software as a DNA molecule (See: Gene synthesis) - although at about $0.25 per bit, I'm not sure I'd be storing anything very large that way!  Then we can go the other way - we can gene-sequence a DNA molecule - get the long list of A's, G's, T's and C's - and replace each one of those letters with '00', '01', '10' or '11' and our genome information is just 1's and 0's.  We can send it to Mars on a radio beam, copy it to our hard-drives - and (interestingly) pay a company $0.25c per bit to turn it BACK into DNA again.  There quite simply ISN'T a distinction between software and the bits that make up a DNA molecule.  You are confusing the information with the thing that holds the information. SteveBaker (talk) 03:58, 4 January 2009 (UTC)


 * Yes, DNA codons consist of three nucleotides, where four different nucleotides are possible. --Russoc4 (talk) 01:29, 4 January 2009 (UTC)
 * So it's a base-4 coding system - quaternary - not binary. However, that's a trivial distinction - they are both 'digital' codes and that's what matters.  In the early days of computers, there used to be computers that worked in base-10.  We software engineers habitually 'pretend' that binary digits are grouped into three or four so we talk about base 8 (octal) and base 16 (hex) numbers without caring too much how they are represented 'under the hood'.  Many communication systems use base-4 coding schemes.  That DNA happens to use base-4 is quite utterly irrelevent in terms of this analogy.  You can analogize the DNA 'genetic code' with a piece of software code and the comparison stands up quite well:  Nucleotides are like bits (except they are in base-4), Codons are like machine-code instructions, Genes are like subroutines, Chromosomes are like compilation units, the entire Genome is like a computer program (and a surprisingly small one IMHO).  The analogies are very close (and I don't think that's an accident).  However, the OP's analogy fails quite utterly - but not for that reason.  My complete answer (which explains this point) is below. SteveBaker (talk) 03:07, 4 January 2009 (UTC)


 * While my background is in biologyu

I once found that if I added subroutines to a program from a library of Fortran subroutines, each did what it was supposed to do and the overall program worked. This might be a better analogy to the genetic combinations of sexual reproduction than your false analogy of taking a few bits from one program and a few bits from another. Each of the combined units has to have a certain degree of completeness, like a gene, and not just random base pairs like your analogy would imply. You need complete modules. Edison (talk) 01:31, 4 January 2009 (UTC)
 * (EC)I'll bite a bit. Meiosis is not like trying to put together two random programs; it's more like taking two versions of a program that were once the same but have been further modified or tweaked by different developers and then put together. However, the coding for each module within the software would have to be about the same size and in roughly the same place within the overall program. (I'm not a programmer; I'm a biologist. Pardon my gross simplification/poor analogies of programming. And I'm thinking Windows and its various components in my analogy.) Human chromosomes are by and large extremely similar, each is around 96-99% identical to its sister chromosome. (This is from the recent sequencing of the haploid human genome reported in Nature earlier in 2008; I've got to run, but I'll get the exact numbers and ref later.) So during meiosis, the two chromosomes swap genetic material but the genes are essentially in the same order, with only minor differences between them. You aren't trying to put together a fly with a human; that wouldn't work. Or Windows crossed to Mac OSX. I'll post more later, but I have to run now. -- Flyguy649 talk 01:38, 4 January 2009 (UTC)


 * The thing our OP is missing is that the two humans are ALMOST identical - they differ only by the tiniest percentage of their DNA (remember - human DNA differs from chimpanzee DNA by less than 1% - so a 1% difference is a LOT!)  But if you attempt to cross-breed (say) an Aardvark with an Apple Tree - then the odds of getting a successful living plantoid-creature from the random gene shuffling are indeed almost exactly zero.  I don't know where this 1 in 100 estimate comes from - but I can't believe there is any science whatever behind it - so that's what we call "a wild guess" and not "an estimate"!


 * Our OP's computer program analogy is also faulty. If you took two functioning computer programs that were both derived from the same original source code and did almost exactly the same thing - and were identical except for a few different lines of code and a couple of different constants - then you could indeed take random subroutines from one and put them into the other and the resulting code would stand a very good chance of working indeed.  (In the software business, we do this all the time when we do a 'merge' using a version control system such as subversion - and indeed the resulting hybrid of my latest version and my co-worker's latest version does indeed work 99% of the time.)


 * More importantly for the survival of our offspring - there is a lot of redundancy in the genome - there are many genes that are duplicated on two different strands - so if one gene fails, the other takes over.  That's how someone can be a 'carrier' of something like sickle-cell anaemia without suffering any consequences from it.  It's only when both parents have one defective copy of the gene that their offspring stands a 1 in 4 chance of getting two bad copies and the disease manifests itself.


 * So you may rest easy - this is no problem at all to understand - no deep mysteries - just simple genetics. SteveBaker (talk) 01:41, 4 January 2009 (UTC)


 * I thought the Windows analogy was good. People can have a random assortment of different versions of different DLLs and it all still works. Sort of. By the way there are genetic algorithms implementing something like sex used in software and they are quite good at evolving useful designs. Dmcq (talk) 12:14, 4 January 2009 (UTC)
 * Nice one! Genetic algorithms are an excellent response.  Remove the "reproductive" component and the algorithm will not improve.  --Scray (talk) 15:55, 4 January 2009 (UTC)

In the computer code analogy, mixing bits or machine language instructions from two different programs is likely to end badly. But if the programs were Basic, and one was displaying a picture while the other was playing music, or printing text, mixing the instructions would often yield a program which would execute, occasionally with interesting results. There could still be structural errors which would prevent execution, but sometimes the result would be something like the combined execution of the two programs. If one programs said print "hello" and the other said to sound a "beep" these two events could happen in turn. Loops or "goto" statements could produce random results when intermixed. Edison (talk) 22:49, 4 January 2009 (UTC)


 * So here's where genetics and programming as we do it differ; especially as regards the poster's presumed dilemma. In order for there to be two humans creating a third human, they must be functional "programs." That is, PRINT "HELLO WORLD" and PRINT "HELLO World" are a priori valid programs, otherwise there would not be the issue at hand. Additionally, DNA when it is copied has a number of "protections" "built-in" such that it is very unlikely that copying PRINT -> PRINT you'll yield anything other then PRINT. It's still possible, and something along those lines may contribute to stillbirth, cancer, and spontaneous abortions. But the data isn't just held in quadrinary form - every value is duplicated, so not only does a mutation have to spontaneous occur AND sneak past the error checker (which was called p13 when I was in school, although apparently they went around and named these things since then) it also must occur at two sites! Even giving a 50-50 probability (which is surely outrageously in the arguments' favor) for all that, it's still less then 1/8 possibility for a single point mutation (contrast with 100:1, the OP's claim). Also in our favor is that a lot of our genetic material is junk DNA, or functionally irrelevent - say, my eye color. Even if I mutate a brand new spanking eye pigment, I'm still a functional human being. Finally, DNA - as best I understand it - doesn't seem to be a blueprint in the terms we'd think of it - "Make a five foot long spine, see page 15 for details," but rather, it's more like a progressive algorhythm (although I seem to have that name wrong - if anyone remembers the term, by all means, edit and claim) where "any" set of data can fit in. That is, if there was a point mutation in "five foot long spine", it would of necessity be a valid product (as opposed to, for example's set, a cluster resulting in "dive tool rong line") - although perhaps not necessarily "human".

Or, in short, while you're entirely likely to get PRINT "HELLO WoRlD" from the two parents, it's not terribly likely you'll just hop on over to "MOV AX,GGGG" It can and does happen, but consider Levenshtein distance as applied and given two functional inputs. 98.169.163.20 (talk) 23:55, 4 January 2009 (UTC)
 * While SB and 98 have given good responses and others have briefly mentioned this a key point is that you don't just randomly get bits and pieces of DNA from you parents. You get whole chromosomes. When we're talking humans your parents by and large don't really have different genes. They have different alleles, in other words different versions of the same gene. Most of these differences have little effect. To use a programming example, let's say you have a subroutine which in response to a signal adds questions marks to the end of sentences (that signal is sent by other subroutines which only send the signal when the sentence is a question). Parent A has two functioning copy of this subroutine, Parent B has two defective one. If Parent A and B mate, the resulting program could have either two copies of one gene, or one of both. If it has at least one copy of the subroutine, it will add the questionmarks. If it has none it won't. The later will be a little annoying but is not critical. We know that because parent B had it and parent B wouldn't exist if the problem was critical. And that's a key point. It's not possible that parents are missing key subroutines (or genes) because they wouldn't themselves exist if they were (it is possible they only have one copy of a key subroutine or gene). P.S. One clear example of the flaw in the OPs thinking is this "otherwise you could just take a haploid version of two pieces of software, combine them willy-nilly, and get a new piece of software". Except that a mother and father aren't two pieces of different software. They are both humans. We're not talking about a Dreamfall and NOD32 'mating' but two different versions of NOD32 with very minor differences mating. Nil Einne (talk) 10:49, 5 January 2009 (UTC)
 * Yes - exactly. And (as I pointed out) - software engineers do PRECISELY that as a routine part of their jobs.  Already this morning I took a version of the software that I've been working on for a couple of weeks - and a version of the same software that someone else has been working on for several weeks and used the 'merge' feature of the subversion version control software to merge them together.  The resulting program runs like a charm.  Sometimes it doesn't come out like that - but generally, it does.  This is precisely analogous to some set of DNA data that started off as a common ancestor of husband and wife - which has been altered (very slightly) by evolution over the past dozen generations so that the husband's version of those DNA strands is a little different from his wife's.  The resulting "merge" of those two sets of information is (just like my merged software) very likely to work just fine.  More importantly (as others have pointed out) - the way cellular biology works - there are two copies of most of the information in the DNA - and you get one from each parent - so there is a fair bit of redundancy that make the system much more fault-tolerant than my software merge analogy.  That's reflected in the fact that perhaps one in fifty of my software merges results in a program that either won't compile or crashes in some way - even though both 'parent' versions of that software compiled and ran just fine...where much more than 1 in 50 children grow up without significant DNA defects.  SteveBaker (talk) 15:18, 5 January 2009 (UTC)


 * It's an invalid analogy. DNA is more akin to data, not executable code.  If I have two well-formed XML files, both of which conformed to the same set of strict XSD rules, you absolutely can combine the two XML files to produce a third valid XML file. 216.239.234.196 (talk) 13:36, 5 January 2009 (UTC)


 * No - I can't agree with that. There is VERY little difference between 'code' and 'data'.  XML is virtually a programming language.  Very often, what's data for one program (eg a '.html' file for Apache) is also code for another (JavaScript for Firefox).  Sufficiently complex data file formats (like XML) can have bugs in them that are similar in nature to coding errors.  I can't accept that analogy. SteveBaker (talk) 15:18, 5 January 2009 (UTC)


 * Executable code contains machine level instructions that a CPU will execute. Data does not.  Think of code as action; data as information.  216.239.234.196 (talk) 17:50, 5 January 2009 (UTC)


 * Actually, code and data are indistinguishable. If you change a file extension from .exe to .bmp, what was once a program is now a weird-looking picture. That's how buffer overflows work, if you enter the right data, it becomes code and gets exectued. Franamax (talk) 18:39, 5 January 2009 (UTC)


 * No, just because both code and data can be stored in the same place doesn't mean they're the same thing or that they're indistinguishable. It's been a long time since I worked in machine language, but if you gave me binary code and binary data for a CPU I know, I could easily distinguish which was data and which was code based on the op codes, registers, values, etc.  I could also write a program that could distinguish between the two. 216.239.234.196 (talk) 17:27, 6 January 2009 (UTC)


 * So I repeat - is JavaScript code or data? It doesn't contain "machine level instructions" because it's interpreted, not compiled so by your first criteria, it's data (and Apache would agree with you - but Firefox would not).  It does cause 'action' - so by your second critera it's 'code' - but an XML file can also be "action" if it's interpreted as (say) a route for a robot to follow.  You could have a tag in the XML that tells the robot to repeat the previous route - or to repeat that route until it hits a wall...before you know it, your XML "data" has become "code".  I suppose one could resort to the Church-Turing thesis and say that anything that acts equivalently to a Turing machine is "code" and everything else is "data" - but the BIOS ROM in your PC (because it cannot modify itself) is not code by that definition.  Think about programming in Logo.  You say things like "left 90 ; forwards 10 ; right 90" to make a robot move around - is that a route like our robot's XML route-description-file?  Is it code?  Sorry - but your distinction is blurry as all hell - and there is a very good reason for that - some things are both code and data - some things are clearly just code and others just data - and yet other things change their nature from code to data and back again depending on the context.


 * In a sense, DNA is code that is "executed" by the RNA molecule 'computer' - in another sense it's data that the RNA "program" processes. It's a lot like JavaScript in that some processes are treating it as data (the process that duplicates a DNA molecule for example doesn't care what 'instructions' the DNA strand contains - it's just data) - while other processes treat it as code - with a proper instruction set (such at the process that creates proteins).  Read our article genetic code - it describes DNA as a set of 'rules' - which is pretty close to saying 'code'.  A codon is made of three base-pairs.  Each base-pair represents a 2 bit value - so there are 4x4x4=64 possible codons. The analogy with bits and machine-code instructions is completely compelling. The 64 codons contain instructions to the RNA to grab a particular amino acid and hook it onto the end of the protein - but also has instructions that stand for "START" and "STOP".  The analogy is even more complete than that - some organisms have different RNA "computers" that use different machine-code sequences from ours - just as the old Z80 computer had a slightly different instruction set from the (broadly compatible) Intel 8080.  The RNA "computer" even has things like exception handling to deal with illegal instructions and prevent the system from "crashing" if a bad DNA sequence is encountered.  We implemented all of these things into our computers before we understood how DNA worked - it's an amazing case of "convergent evolution" where cellular processes and electronics have converged on almost identical solutions. SteveBaker (talk) 18:39, 5 January 2009 (UTC)

Congrats to the OP for a provocative question topic... I presume from the way the question was posed that the intent was to stir up a great argument and it has certainly been fun to read. However, as stated by several others already, the analogy is false. While it may be true that DNA is "information" and that some aspects of the DNA code are similar (and in many ways analogous) to computer code, we're missing a huge point here. Humans are diploid organisms, meaning that there are two nearly exact copies of each chromosome, and two nearly identical copies of (almost) every gene being expressed at the same time. It isn't as though fertilization slaps together random halves of two parental genomes to get a new whole (it was already suggested that the OP review meiosis for further clarification). I assume (as a non-expert in computer programming) that it would be a lot more difficult to do that way. THAT is why the initial question is a false analogy. Computer systems are the equivalent of a haploid genome. To pose the original question, you would need to have a computing environment in which every computer was a diploid system (new DOS = Diploid Operating System?) running two versions of each program (gene) simultaneously. Allow the user of each computer system to make small alterations (mutations) in the programs to optimize them for their own uses. THEN, select two computers (perhaps via an online dating service?), randomly choose one version of each program running on those computers (the haploid gamete), compile the two sets of programs together into a new, unique diploid software combination (the diploid offspring), and determine whether the resulting software could operate a new computer. Keep in mind that you'd have to limit users to tweaking the existing programs rather than writing new programs within a given operating system, unless you wanted to create a new species of computer systems. I expect there would be changes made in programs that would be incompatible when both programs were being run simultaneously, but you might do better than 1:100 if you designed the system robustly enough. --- Medical geneticist (talk) 21:56, 5 January 2009 (UTC)
 * My question back to the OP is, Why don't computer systems work in such a way as to execute two nearly identical copies of each program simultaneously?


 * Actually some fault-tolerant systems do work that way, e.g. for the space shuttle. It's expensive though. Dmcq (talk) 15:06, 6 January 2009 (UTC)


 * (@Dmcq) Yes - although two copies isn't enough - if the two program produce different results - which one do you believe? This is a problem with DNA too - the two copies may produce subtly different enzymes for some key body function (for example) and if one of them works great but the other doesn't then you may simply be mildly deficient in some important respect - or you might die of it - or you might function 100% OK - except that one of your children gets two copies of the defective gene and dies.  That kind of uncertainty isn't going to help the space shuttle (Shall we fire the rocket motors now?  Dunno - one computer says "Yes" and the other "No" - so let's fire half of them?!)...so the shuttle has (IIRC) seven computers - some of which run different software written by different programmers - and they vote on what to do.  If one computer consistently votes against the others - it gets turned off.  But I've worked in fault-tolerant systems (telephone exchanges) that employed multiple computers and voting logic - and you end up wondering what happens if the voting logic fails.  It's a pretty subtle business to make a truly fault-tolerant system.


 * (@Medical geneticist) what you are describing is more or less a 'genetic algorithm' - and we certainly have those and use them fairly often in some areas of computer science. We'll take a bunch of (literally) random computer instructions and call that a 'genome' - we'll make a bunch of random organisms and run them to see which comes closest to producing the output we want.  The one that comes closest is then 'mutated' by changing a few instructions at random to make a bunch more programs - which we again run to see which comes closest to what we want.  Do this (automatically) for a few million generations and the program to do what you need literally writes itself!  It's not a diploid system - and it's typically an asexual process (we don't mix the genes from two parents - we just mutate one parent to make a LOT of children).  This is OK for small, specialised cases (I used it for recognising and identifying the outlines of buildings in satellite photography) - but in general, it's a tough way to make software.  I did once make a system for generating the parameters to make 3D models of trees for drawing a large forest in my graphics system.  That allowed both random mutation and sexual reproduction which worked very well - the user looks at a bunch of 3D tree models and clicks on the one (or several) that they like the most.  The screen then clears and the choices are cross-bred - or if there is just one, it's randomly mutated - and the descendents are used to generate more tree models.  After just a handful of generations, you start to get really nice-looking trees that fit your personal idea of what a tree should look like.   If you want things that look like mature oaks - that's what you get.  If you want fir trees - you can make those.  Skinny immature apple trees - no problem!  If you'd like to play with a system like that, do a Google search for Biomorphs and find an online biomorph player to dink around with - it's fun! SteveBaker (talk) 16:33, 6 January 2009 (UTC)