First, in some countries this would automatically spark an ethics investigation or be deemed corrupt. It may not be "corrupt" in the US, but I suspect that's more of a relative definition of corrupt than an absolute one.
Second, the US is ranked 24th in the world on corruption. I'd therefore argue that the standards the US government holds itself to is not only nowhere near what it could be, but isn't even anywhere near as good as other nations are managing on a day-to-day basis. This isn't great for smaller nations, though you can understand that they don't have the resources to be equal and of high quality. They also don't have much influence and the impact of corruption is necessarily limited. A fair number are also very new and don't have much experience. A nation like the US is a different matter. They've plenty of resources, they've had three centuries to work out the flaws, and they've far too much power to not be responsible with it.
The fact that New Zealand, Denmark and Finland are first and joint second respectively (none of whom are permanent members of the UN Security Council, hold nuclear weapons, dominate either the IMF or World Bank, or control vast swathes of international trade) is worthy of great respect. They don't have to be as good as they are, they just are because they by-and-large want to be. Not saying they're perfect, this is a ranking system not a measure against a fixed standard, but it is highly commendable none-the-less.
Perhaps, but IE is a major security hole. At the very least, hospitals should be absolutely required to use a secure browser. Secondly, with ERP, etc, being browser based, there's no difference from an operator standpoint between Windows and OpenBSD. You still click links, you still open tabs, you still get to set the wallpaper on the background. Ergo, there's no rational reason to use something that's expensive and insecure over something that's cheap and secure. If there are no platform-specific apps (they're all web-based) then go with the OS that is least likely to endanger service.
A sore throat can be something trivial, but it can also be something major. Going to a GP to have it checked out rather than waiting and seeing is the height of common sense. A hospital, not so much. Hospitals can do nothing a GP can't do, for those sorts of ailments. Hospitals only make sense if you actually need centralized, high-end medical treatment. You can't fit an MRI into a GP's office and a doctor certainly can't take one with them if they're doing house calls, nor will smaller facilities be able to detect everything in-house.
Oh, I thought you were referring to a society with sensible health-care!
The most intelligent health-care systems are ones where the method of delivery is one that suits the complaint. That doesn't necessarily mean the best - a poor but intelligent system will be more effective than a poor but stupid one, and will also be more reliable and more responsive than a rich but stupid one, but the rich but stupid system will still deliver better results in the end. What you want is rich and intelligent, but no country currently does that.
Tell you what, find a good pack of lead-weighted slide rules and we can go over to this education board and educate them on the value of knowing how to do sums.
Your method of solving it was good - and a method I've seen used by a lot of older people -- and computer programmers*. The alternative is to remember that a series of multiples and divides can be done in any order, so (47*75)%25=47*(75%25)=47*3
*Yes, using left/right shifts and add/subtracts, which is all you were doing, is not only very fast in the human brain but it's also very fast on a computer. Much more so than general multiplication/division.
There should be no distinction. Theory is what lets you generalize practice, practice is what lets you understand theory. Having one and not the other is like having paint and no canvas (or canvas and no paint). Back when I were a lad, we'd covered basic graph theory, Venn diagrams, basic addition and multiplication, etc, by second year of infants. I wouldn't say it was perfect, we could have learned a lot more in a lot less time if the teachers hadn't been psychotic, but it was respectable. First year of infants was incredibly stupid - there'd been much more ground covered in each of the two years of preschool before that. First year of secondary school was also stupid, covering much the same ground as primary school had. All in all, I feel sure that even normal, neurotypical students could have been University-grade a good 3-5 years earlier by cutting out wastage and tuning things sensibly.
What kids need later in life is impossible to say - how many people use slide rules today? (As compared to SHOULD be using slide rules...) The fact is, the market changes quickly. Only a mind that is brought up to be agile can hope to both be relevant at the start of their career and at the end. Agility is what matters, facts can always be referenced, although you cannot teach agility without facts.
Not sure that matters. Anyone who believes maths isn't vitally important in every aspect of real-world life should be shot, hung, drawn, quartered and then fired.
Optical interferometry is done, so there's nothing to stop someone setting up an array of optical telescopes in which Hubble was one of the telescopes involved. Ideally, since large optical telescopes are very difficult to launch, future optical space telescopes should be designed to be used in interferometry arrays. The problem is one of synchronizing, since you can't use interference when the signals aren't in phase and relativistic time matters if they're not on a common orbit, but if you record the signal and timestamp points along it, it should be possible to do the interference offline rather than live.
The other thing to remember is that telescope diameter is only one variable. Another is collecting area. You could park two radio telescopes in geostationary orbit such that they were at opposite ends. This would give you a gigantic telescope diameter, but almost no collecting area. SKA, when it is finally built, will have a superb collecting area, vastly greater than what's described here, but a much smaller effective dish diameter than the telescope described in TFA. As such, it'll be far more sensitive but have lower resolution. Combining the two would be awesome.
(Dotting a SKA across the entire length of Earth's orbit would be even more awesome, but I don't see that happening for a while.)
You misinterpreted the comma to mean an explanation of the Peter Principle. I was a bit concerned that might happen and thought of moving the phrase to the end. However, there, it was possible that people would take it to mean that everything in the list was summed up by the Peter Principle. In the end, I decided that as long as there was clearly a list that people would recognize that the elements were independent rather than one element explaining another. The problem is that although English allows ownership of subclauses, it does so in exactly the same way as it describes independent elements. I'm trying to construct a formal "natural" language that is free of all ambiguity, but (a) I'll probably fail, and (b) nobody would use it even if I succeeded.
Dunno, which is why I use LaTeX in a text editor, but my point is that WYSIWYG is old and early implementations of it were no worse - and sometimes better - than modern implementations. I'm not keen on WYSIWYM (eg: LyX) either - I don't need to write programs as if they were running and seriously doubt you could write a good program that way anyway, writing a document that way is just as illogical. (Hands up all who think DreamWeaver is Evil Incarnate. Hands up all who use Office and the Microsoft "HTML Printer" to write web pages. My guess is that almost nobody uses WYSIWYG for anything other than cleaning a document up after it has been written, drafted and largely finalized.)
For me, the distinction between the arts and science is a non-starter. I firmly believe in the Neo-Classical notion that they're merely two sides of one coin and that it is the coin that is important and not the side. Thus, all arts may be tested (or enhanced) by science and all sciences may be advanced through the arts. I disagree with Neo-Classicists on some details, especially the idea of there being a few "elect" cultures, but holistic approaches to understanding seem better than deconstructional understanding.
(In fact, because of the complexity of the interactions between different subjects, I'm not convinced it is possible to understand anything substantial without having some grounding in essentially everything. Inter-disciplinary interactions are an oft-neglected field in our modern, segmented society, but it's where the interesting stuff is.)
Leeches turn out to be extremely useful for very specific classes of localized problem, which is basically a form of semi-controlled bloodletting, and those are sometimes used in Western hospitals. What we've done is used history to identify what works and what doesn't ("evidence-based medicine"), which seems to be a very good way to utilize history's lessons.
Not a problem. You'll probably have noticed I can be a bit short with people from time to time, but you'll also have noticed that often those people aren't interested in discussion or understanding but hostility. Nothing wrong with whacking bandits over the head with a big stick. However, people like yourself are interested in understanding and there's everything wrong with whacking people like you over the head with a big stick.
The concept of "Enlightenment" is poorly defined, but ultimately seems to be a permanent shift in the way the brain functions. The structural changes accumulate over time, producing some change in your ability to function, but you appear to eventually reach some sort of tipping point where the function changes dramatically and over a very short time. It's like the dramatic shifts you can get in a chaotic system, when the system leaps from one strange attractor to another. Indeed, that may well be what it is, as Buddhist texts do say that Enlightenment can occur at almost any time under almost any condition. You don't practice X amount and then become Enlightened, it's much less predictable than that.
As to what the change is, that's also poorly defined but appears to be a permanent, unstoppable state that is akin to meditation but may actually be a lot deeper than that. Experts in the field of neurology have not, in my opinion, studied meditation or Enlightenment nearly enough to be able to say. The image I've picked up over time is that meditating is akin to a software algorithm for noise reduction and Enlightenment is akin to starting with error-free, noise-free senses in the first place.
Let's say you've got 12 roads leading out from the centre of a town, where the angles between one and the next are equal, and 12 is the road that points due north. Let's also say that the rings are named alphabetically, with the innermost ring being A. 6 and Q is extremely efficient. Compared to a grid, it's arguably more so because the quadrant is implicit. In a grid, you have to specify.
Ok, what if you want a few roads radiating out from the centre, but also want many more radial streets between only selected rings? Then instead of basing it in a clock, use degrees with 0 as north. 45 and R can refer to one place and one place only. If you're at 57 and S, you know you're one ring too far out and that you have to travel counter clockwise from your current location. Again, quadrant is implicit.
In today's world, being able to read might make you a genius in some people's eyes. However, being well-read and versed in any specific field would appear to have an overwhelming impact on your brain's ability to function in that field, and I could easily see the possibility that anyone who has studied multiple fields in such a way could improve the plasticity of their brain. Indeed, most of my thinking on education is based on that premise. My earlier remarks on what is essentially over-specialization really come down to the idea that it's entirely possible that there's a neurological explanation for "being stuck in a rut".
Exactly, especially on the scale we're talking here (14-20% growth of the hippocampus). That's major remodeling. Most tests prior to the popularization of MRI would have been subjective simply because patients don't generally like doctors to saw their brains open. Even with MRI becoming widely available, it's not cheap - particularly at adequate resolution - and it's time-intensive, which limits the projects that can use it. The methodology can also be a bit slipshod at times and the popular practice of not actually doing original work but merely doing analysis on a compilation of previous studies means that errors in methods can be magnified.
It doesn't help that in 2009, 33.7% of scientists who responded admitted questionable practices with almost 2% admitting outright fraud and falsification of results. That's a lot, especially as the majority of those carrying out fraud are unlikely to have admitted it. That means that one-off, unrepeated studies should not be trusted over a third of the time merely for scientific malpractice. That's ignoring misinterpreted results, errors in the experiment, errors resulting from the statistical nature of science, or any other such innocent factor. The innocent reasons are significant enough that nobody should trust unrepeated/unrepeatable experiments, the malpractice wallops the number of studies needed and the rigour of those studies right up. Which means even where there appears to be good evidence of brain redesign, you have to be a lot more wary until there's solid confirmation from others.
The study is interesting - I'm a bit bothered by the fact that the only information I can find is just rewordings of the press release, which states that there was "thickening" of the cerebral cortex in areas associated with attention and emotional integration (it later says they found increased grey-matter density in the hippocampus plus decreased grey-matter density in the amygdala). The amounts are never mentioned, beyond it being observable (so it's a gross structural change and not a minor one - the hospital involved doesn't have an MRI powerful enough to see fine detail).
I'm convinced they've measured something, I'm willing to accept that they may have seen what they think they've seen, but I'm not willing to accept that their study proves anything more than a probable correlation, especially as the correlating was done via questionnaires (subjective data) rather than an objective measurement of brain response, participants reported meditation time (subjective data) rather than it being objectively measured, the statistical probability of getting the result by chance is higher than most disciplines would accept due to there being a small sample (the five sigma test applied in hard sciences requires that something could arise by chance 0.0000573303% of the time - here they didn't express their confidence limits at all which usually means they tested at the 5% level), the experiment contained many activities (so you can't isolate what had the effect) and no controls were in place to test similar but non-meditative techniques.
In short, I'd consider what you've linked to to be a superb piece of evidence that activity alters the gross structure, and evidence that follow-up research studying the relationship between mental exercise and neurological structure is needed, but in and of itself, it does not show what activity causes what structural change.
So, no, I would have to say that this study did not show mindful meditation had a physical effect on the brain, but it DID show that it is possible that it did. However, without the additional tests, you can't possibly tell whether it was the meditation, the mindfulness of it, some combination of factors in one of the other activities in the meditating group, or whether any relaxation technique would have worked in the same way. I have respect for Tai Chi, but there have been Buddhist Masters that have achieved Enlightenment by staring at a wall for 7 years. There wasn't even any paint drying. No way was that "mindful" and it certainly had no physical component, but it clearly had a major impact on the brain. Indeed, Zen Master Dogan (I truly recommend you look at the Shobogenzo if meditation and philosophy are your thing) argued that method, philosophy and specifics of practice were all immaterial, that those were essentially window-dressing, that a very few, very tiny core concepts were what mattered and the rest just gave it some form you could make use of but the rest otherwise doesn't matter at all.
You are absolutely right, particularly as high-res MRI is extremely new and very limited. Almost all studies rely on either external observations or gross structural change that can be seen on the cheaper and commonplace MRIs, and of those it is exceedingly difficult to draw any sensible conclusion because of the sheer number of variables and the difficulty of knowing what is a cause and what is an effect - especially when you've a multitude of feedback loops that can turn one into the other.
Tests are also exceedingly hard to design well. The so-called "Mozart Effect" turned out to have an element of truth (listening to music you enjoy and listen to for a particular purpose will typically have that same effect any time you listen to it, even when it's not for that purpose) but not the effect that the original experimenters claimed (Mozart's music has no special properties beyond the one of being composed by Mozart).
IMHO, the best that can be said is that objective experimental data is more trustworthy than subjective experimental data, that multiple experiments to establish what is cause, what is effect and what is independent is absolutely essential, that science isn't about proving as much as falsifying, and that many groups should conduct overlapping but different experiments to eliminate error by flawed design. In this particular case, the first was done, the second can obviously go on forever but has been done to an extent, the third's a bit harder without analyzing the methods in depth, and the fourth hasn't happened but should.
For the plasticity, yes it's been known that plasticity doesn't stop after childhood, but this isn't simply the occasional neural connection forming, we're talking 14-20% growth in the hippocampus in adults. That's a hell of a lot. Yes, it's an area where neurogenesis happens, but that's......a lot. This third study in the series appears to have been aiming to determine if (a) there was growth and not just a lot of X-Men mutants becoming cabbies, and (b) that the growth was a function of the learning and not the experience after.
Ok, I can accept that for D. I can probably buy the Occam argument as well, though would contend that there are very few libraries that can provide explicit parallelism, mobile functions and other such nifty features (the main reason Occam's so good for parallelism) in a completely side-effect free environment -- although you are entirely correct that you should be able to do all that at the library level. The lack of such libraries is the main reason I believe that people are finding it just too difficult to reimplement it. eg: OpenMP is a hack that provides a very limited subset of those kinds of facilities and the ATLAS developers recently kicked it out of the code because it's too slow and doesn't work well.
Nonetheless, quibbles aside, it should be possible to extend OCaml to support everything that Occam-Pi has that's useful, whether as syntax updates, libraries or a mix of both. My worry is that over-extending a language can hurt it and I'm not confident that the extensions involved would be minor enough that OCaml ends up trying to do too much and be too flexible. Adding features can break assumptions that are critical to other aspects of a language.
You're probably more experienced in OCaml than I am, so if you're confident that the extensions needed wouldn't break anything, then I'll accept that and go with the idea that that's all you really need. (Despite the occasional claim to the contrary, I listen to people with expertise I don't have.)
So we're down to Haskell, Erlang and OCaml, where one or more of these is extended in the way you've described to handle the extra facilities provided by the languages being dropped.
Wait a minute, your other post claimed that the paradigm didn't matter. Now you're claiming that some paradigms prohibit side-effects and thus compile cleaner to modern hardware. You may have a superior UID, but you need to work on your consistency.
You also need to consider that Fortran compilers have been around a very very long time. I don't know if you can find a Fortran IV compiler and a Fortran 2008 compiler that will compile to the same architecture (that age difference is a bit extreme) but I'm willing to bet that if you could find such compilers where both are written by good quality compiler writers (ie: not by someone high on mushrooms getting bored and writing a compiler for the hell of it), that well-written code that can work on both will compile to smaller, faster code using the more modern compiler.
You could argue that this is because compiler writers are better than they were. Well, define "better". Well, they know more about how to convert the Fortran syntax into machine code, what optimizations are suitable and what aren't, etc. Ok, so we have two variables in this - the variable called "experience in Fortran" and the variable called "experience in compiler writing". My claim is that the first variable is not only non-zero but is significant with respect to the second variable. Doesn't have to be larger, just has to be enough that it's demonstrably greater than noise. Your claim, if taken literally, is that the first variable is absolute zero.
Showing that the first variable is not zero requires that you have two compilers of the same age but where one language is newer than another. Ok, we can do this the easy way. Java and ADA are newer than C, so according to my claim GCC's C frontend should produce better code than GCC's Java and ADA frontends. The compiler is the same generation, so the same compiler experience exists for all three. The backend is the same in all cases. The only difference is in how the instructions are handled. Since ADA is actually very good on limiting side-effects, we can also examine the side-effect issue at the same time.
My experience is that GCC's C frontend is the better of the three. It produces smaller code and faster code, despite C's semantics (which, I agree, are absolutely crud). The only way this can sensibly be explained, given that they're of equal compiler generation, is that people know more about how to process C than they do Java (it's way newer) or ADA (which is middle-aged but not used extensively enough for experience to have built up in the community).
Let's now look at compiler technology itself. Much of the optimization is labeled "back arts" because it's not well-understood. Poorly-understood methods used "because they work" and not "because we know why they work" sounds crappy to me. What about code generation? Well, NASA used G95 for a long time rather than GCC's F90 because GCC's Fortran compiler kept blowing up. Sounds like there were problems. Ok, compiler design? Well, you're way old enough to remember the GCC vs EGCS vs PGCC debacle. Sorry, but I can't excuse you there. The sole reason for that ever happening was because each team made some good design decisions and some poor design decisions. If it had been clear-cut, you'd have had the same thing as you had with XF86 vs the original X Consortium, and XF86 vs the new X.Org - a universal and unambiguous switch with no split in the community, merely a split amongst the developers.
The compiler "crisis" caused a split in the community because each option had benefits the others didn't and the others weren't able/willing to produce something similar on their system. If compilers were truly modular, the PGCC optimizations could EASILY have been merged into EGCS and activated solely for those platforms those optimizations applied to. Didn't happen. The extra facilities being developed for "mainline" GCC could also have been merged into EGCS - some were, but others were rejected. In a modular design, you can't "reject" a module, you can merely not add it into th
Slashdot Mirror
First, in some countries this would automatically spark an ethics investigation or be deemed corrupt. It may not be "corrupt" in the US, but I suspect that's more of a relative definition of corrupt than an absolute one.
Second, the US is ranked 24th in the world on corruption. I'd therefore argue that the standards the US government holds itself to is not only nowhere near what it could be, but isn't even anywhere near as good as other nations are managing on a day-to-day basis. This isn't great for smaller nations, though you can understand that they don't have the resources to be equal and of high quality. They also don't have much influence and the impact of corruption is necessarily limited. A fair number are also very new and don't have much experience. A nation like the US is a different matter. They've plenty of resources, they've had three centuries to work out the flaws, and they've far too much power to not be responsible with it.
The fact that New Zealand, Denmark and Finland are first and joint second respectively (none of whom are permanent members of the UN Security Council, hold nuclear weapons, dominate either the IMF or World Bank, or control vast swathes of international trade) is worthy of great respect. They don't have to be as good as they are, they just are because they by-and-large want to be. Not saying they're perfect, this is a ranking system not a measure against a fixed standard, but it is highly commendable none-the-less.
Perhaps, but IE is a major security hole. At the very least, hospitals should be absolutely required to use a secure browser. Secondly, with ERP, etc, being browser based, there's no difference from an operator standpoint between Windows and OpenBSD. You still click links, you still open tabs, you still get to set the wallpaper on the background. Ergo, there's no rational reason to use something that's expensive and insecure over something that's cheap and secure. If there are no platform-specific apps (they're all web-based) then go with the OS that is least likely to endanger service.
Yes, but in Australia you've salt water crocodiles to solve your problems with politicians.
A sore throat can be something trivial, but it can also be something major. Going to a GP to have it checked out rather than waiting and seeing is the height of common sense. A hospital, not so much. Hospitals can do nothing a GP can't do, for those sorts of ailments. Hospitals only make sense if you actually need centralized, high-end medical treatment. You can't fit an MRI into a GP's office and a doctor certainly can't take one with them if they're doing house calls, nor will smaller facilities be able to detect everything in-house.
Oh, I thought you were referring to a society with sensible health-care!
The most intelligent health-care systems are ones where the method of delivery is one that suits the complaint. That doesn't necessarily mean the best - a poor but intelligent system will be more effective than a poor but stupid one, and will also be more reliable and more responsive than a rich but stupid one, but the rich but stupid system will still deliver better results in the end. What you want is rich and intelligent, but no country currently does that.
Tell you what, find a good pack of lead-weighted slide rules and we can go over to this education board and educate them on the value of knowing how to do sums.
Your method of solving it was good - and a method I've seen used by a lot of older people -- and computer programmers*. The alternative is to remember that a series of multiples and divides can be done in any order, so (47*75)%25=47*(75%25)=47*3
*Yes, using left/right shifts and add/subtracts, which is all you were doing, is not only very fast in the human brain but it's also very fast on a computer. Much more so than general multiplication/division.
You have a lawn? In MY day, we had to get out the broch at midnight, walk up hill - both ways - and make our own damn lawn. Otherwise, yes.
There should be no distinction. Theory is what lets you generalize practice, practice is what lets you understand theory. Having one and not the other is like having paint and no canvas (or canvas and no paint). Back when I were a lad, we'd covered basic graph theory, Venn diagrams, basic addition and multiplication, etc, by second year of infants. I wouldn't say it was perfect, we could have learned a lot more in a lot less time if the teachers hadn't been psychotic, but it was respectable. First year of infants was incredibly stupid - there'd been much more ground covered in each of the two years of preschool before that. First year of secondary school was also stupid, covering much the same ground as primary school had. All in all, I feel sure that even normal, neurotypical students could have been University-grade a good 3-5 years earlier by cutting out wastage and tuning things sensibly.
What kids need later in life is impossible to say - how many people use slide rules today? (As compared to SHOULD be using slide rules...) The fact is, the market changes quickly. Only a mind that is brought up to be agile can hope to both be relevant at the start of their career and at the end. Agility is what matters, facts can always be referenced, although you cannot teach agility without facts.
Not sure that matters. Anyone who believes maths isn't vitally important in every aspect of real-world life should be shot, hung, drawn, quartered and then fired.
Optical interferometry is done, so there's nothing to stop someone setting up an array of optical telescopes in which Hubble was one of the telescopes involved. Ideally, since large optical telescopes are very difficult to launch, future optical space telescopes should be designed to be used in interferometry arrays. The problem is one of synchronizing, since you can't use interference when the signals aren't in phase and relativistic time matters if they're not on a common orbit, but if you record the signal and timestamp points along it, it should be possible to do the interference offline rather than live.
The other thing to remember is that telescope diameter is only one variable. Another is collecting area. You could park two radio telescopes in geostationary orbit such that they were at opposite ends. This would give you a gigantic telescope diameter, but almost no collecting area. SKA, when it is finally built, will have a superb collecting area, vastly greater than what's described here, but a much smaller effective dish diameter than the telescope described in TFA. As such, it'll be far more sensitive but have lower resolution. Combining the two would be awesome.
(Dotting a SKA across the entire length of Earth's orbit would be even more awesome, but I don't see that happening for a while.)
Feel free.
You misinterpreted the comma to mean an explanation of the Peter Principle. I was a bit concerned that might happen and thought of moving the phrase to the end. However, there, it was possible that people would take it to mean that everything in the list was summed up by the Peter Principle. In the end, I decided that as long as there was clearly a list that people would recognize that the elements were independent rather than one element explaining another. The problem is that although English allows ownership of subclauses, it does so in exactly the same way as it describes independent elements. I'm trying to construct a formal "natural" language that is free of all ambiguity, but (a) I'll probably fail, and (b) nobody would use it even if I succeeded.
Do you get off on being offensive, stupid or wrong? Since your post is all three, it's hard to tell.
Dunno, which is why I use LaTeX in a text editor, but my point is that WYSIWYG is old and early implementations of it were no worse - and sometimes better - than modern implementations. I'm not keen on WYSIWYM (eg: LyX) either - I don't need to write programs as if they were running and seriously doubt you could write a good program that way anyway, writing a document that way is just as illogical. (Hands up all who think DreamWeaver is Evil Incarnate. Hands up all who use Office and the Microsoft "HTML Printer" to write web pages. My guess is that almost nobody uses WYSIWYG for anything other than cleaning a document up after it has been written, drafted and largely finalized.)
For me, the distinction between the arts and science is a non-starter. I firmly believe in the Neo-Classical notion that they're merely two sides of one coin and that it is the coin that is important and not the side. Thus, all arts may be tested (or enhanced) by science and all sciences may be advanced through the arts. I disagree with Neo-Classicists on some details, especially the idea of there being a few "elect" cultures, but holistic approaches to understanding seem better than deconstructional understanding.
(In fact, because of the complexity of the interactions between different subjects, I'm not convinced it is possible to understand anything substantial without having some grounding in essentially everything. Inter-disciplinary interactions are an oft-neglected field in our modern, segmented society, but it's where the interesting stuff is.)
Leeches turn out to be extremely useful for very specific classes of localized problem, which is basically a form of semi-controlled bloodletting, and those are sometimes used in Western hospitals. What we've done is used history to identify what works and what doesn't ("evidence-based medicine"), which seems to be a very good way to utilize history's lessons.
Not a problem. You'll probably have noticed I can be a bit short with people from time to time, but you'll also have noticed that often those people aren't interested in discussion or understanding but hostility. Nothing wrong with whacking bandits over the head with a big stick. However, people like yourself are interested in understanding and there's everything wrong with whacking people like you over the head with a big stick.
The concept of "Enlightenment" is poorly defined, but ultimately seems to be a permanent shift in the way the brain functions. The structural changes accumulate over time, producing some change in your ability to function, but you appear to eventually reach some sort of tipping point where the function changes dramatically and over a very short time. It's like the dramatic shifts you can get in a chaotic system, when the system leaps from one strange attractor to another. Indeed, that may well be what it is, as Buddhist texts do say that Enlightenment can occur at almost any time under almost any condition. You don't practice X amount and then become Enlightened, it's much less predictable than that.
As to what the change is, that's also poorly defined but appears to be a permanent, unstoppable state that is akin to meditation but may actually be a lot deeper than that. Experts in the field of neurology have not, in my opinion, studied meditation or Enlightenment nearly enough to be able to say. The image I've picked up over time is that meditating is akin to a software algorithm for noise reduction and Enlightenment is akin to starting with error-free, noise-free senses in the first place.
Let's say you've got 12 roads leading out from the centre of a town, where the angles between one and the next are equal, and 12 is the road that points due north. Let's also say that the rings are named alphabetically, with the innermost ring being A. 6 and Q is extremely efficient. Compared to a grid, it's arguably more so because the quadrant is implicit. In a grid, you have to specify.
Ok, what if you want a few roads radiating out from the centre, but also want many more radial streets between only selected rings? Then instead of basing it in a clock, use degrees with 0 as north. 45 and R can refer to one place and one place only. If you're at 57 and S, you know you're one ring too far out and that you have to travel counter clockwise from your current location. Again, quadrant is implicit.
In today's world, being able to read might make you a genius in some people's eyes. However, being well-read and versed in any specific field would appear to have an overwhelming impact on your brain's ability to function in that field, and I could easily see the possibility that anyone who has studied multiple fields in such a way could improve the plasticity of their brain. Indeed, most of my thinking on education is based on that premise. My earlier remarks on what is essentially over-specialization really come down to the idea that it's entirely possible that there's a neurological explanation for "being stuck in a rut".
Exactly, especially on the scale we're talking here (14-20% growth of the hippocampus). That's major remodeling. Most tests prior to the popularization of MRI would have been subjective simply because patients don't generally like doctors to saw their brains open. Even with MRI becoming widely available, it's not cheap - particularly at adequate resolution - and it's time-intensive, which limits the projects that can use it. The methodology can also be a bit slipshod at times and the popular practice of not actually doing original work but merely doing analysis on a compilation of previous studies means that errors in methods can be magnified.
It doesn't help that in 2009, 33.7% of scientists who responded admitted questionable practices with almost 2% admitting outright fraud and falsification of results. That's a lot, especially as the majority of those carrying out fraud are unlikely to have admitted it. That means that one-off, unrepeated studies should not be trusted over a third of the time merely for scientific malpractice. That's ignoring misinterpreted results, errors in the experiment, errors resulting from the statistical nature of science, or any other such innocent factor. The innocent reasons are significant enough that nobody should trust unrepeated/unrepeatable experiments, the malpractice wallops the number of studies needed and the rigour of those studies right up. Which means even where there appears to be good evidence of brain redesign, you have to be a lot more wary until there's solid confirmation from others.
The study is interesting - I'm a bit bothered by the fact that the only information I can find is just rewordings of the press release, which states that there was "thickening" of the cerebral cortex in areas associated with attention and emotional integration (it later says they found increased grey-matter density in the hippocampus plus decreased grey-matter density in the amygdala). The amounts are never mentioned, beyond it being observable (so it's a gross structural change and not a minor one - the hospital involved doesn't have an MRI powerful enough to see fine detail).
I'm convinced they've measured something, I'm willing to accept that they may have seen what they think they've seen, but I'm not willing to accept that their study proves anything more than a probable correlation, especially as the correlating was done via questionnaires (subjective data) rather than an objective measurement of brain response, participants reported meditation time (subjective data) rather than it being objectively measured, the statistical probability of getting the result by chance is higher than most disciplines would accept due to there being a small sample (the five sigma test applied in hard sciences requires that something could arise by chance 0.0000573303% of the time - here they didn't express their confidence limits at all which usually means they tested at the 5% level), the experiment contained many activities (so you can't isolate what had the effect) and no controls were in place to test similar but non-meditative techniques.
In short, I'd consider what you've linked to to be a superb piece of evidence that activity alters the gross structure, and evidence that follow-up research studying the relationship between mental exercise and neurological structure is needed, but in and of itself, it does not show what activity causes what structural change.
So, no, I would have to say that this study did not show mindful meditation had a physical effect on the brain, but it DID show that it is possible that it did. However, without the additional tests, you can't possibly tell whether it was the meditation, the mindfulness of it, some combination of factors in one of the other activities in the meditating group, or whether any relaxation technique would have worked in the same way. I have respect for Tai Chi, but there have been Buddhist Masters that have achieved Enlightenment by staring at a wall for 7 years. There wasn't even any paint drying. No way was that "mindful" and it certainly had no physical component, but it clearly had a major impact on the brain. Indeed, Zen Master Dogan (I truly recommend you look at the Shobogenzo if meditation and philosophy are your thing) argued that method, philosophy and specifics of practice were all immaterial, that those were essentially window-dressing, that a very few, very tiny core concepts were what mattered and the rest just gave it some form you could make use of but the rest otherwise doesn't matter at all.
You are absolutely right, particularly as high-res MRI is extremely new and very limited. Almost all studies rely on either external observations or gross structural change that can be seen on the cheaper and commonplace MRIs, and of those it is exceedingly difficult to draw any sensible conclusion because of the sheer number of variables and the difficulty of knowing what is a cause and what is an effect - especially when you've a multitude of feedback loops that can turn one into the other.
Tests are also exceedingly hard to design well. The so-called "Mozart Effect" turned out to have an element of truth (listening to music you enjoy and listen to for a particular purpose will typically have that same effect any time you listen to it, even when it's not for that purpose) but not the effect that the original experimenters claimed (Mozart's music has no special properties beyond the one of being composed by Mozart).
IMHO, the best that can be said is that objective experimental data is more trustworthy than subjective experimental data, that multiple experiments to establish what is cause, what is effect and what is independent is absolutely essential, that science isn't about proving as much as falsifying, and that many groups should conduct overlapping but different experiments to eliminate error by flawed design. In this particular case, the first was done, the second can obviously go on forever but has been done to an extent, the third's a bit harder without analyzing the methods in depth, and the fourth hasn't happened but should.
For the plasticity, yes it's been known that plasticity doesn't stop after childhood, but this isn't simply the occasional neural connection forming, we're talking 14-20% growth in the hippocampus in adults. That's a hell of a lot. Yes, it's an area where neurogenesis happens, but that's... ...a lot. This third study in the series appears to have been aiming to determine if (a) there was growth and not just a lot of X-Men mutants becoming cabbies, and (b) that the growth was a function of the learning and not the experience after.
The radial system is also efficient - and, if correctly designed, should actually be better than a grid in some cases.
Like I said, being stuck was supposition. Of course, you might be one of those rare mega-geniuses for whom no normal rules apply.
Ok, I can accept that for D. I can probably buy the Occam argument as well, though would contend that there are very few libraries that can provide explicit parallelism, mobile functions and other such nifty features (the main reason Occam's so good for parallelism) in a completely side-effect free environment -- although you are entirely correct that you should be able to do all that at the library level. The lack of such libraries is the main reason I believe that people are finding it just too difficult to reimplement it. eg: OpenMP is a hack that provides a very limited subset of those kinds of facilities and the ATLAS developers recently kicked it out of the code because it's too slow and doesn't work well.
Nonetheless, quibbles aside, it should be possible to extend OCaml to support everything that Occam-Pi has that's useful, whether as syntax updates, libraries or a mix of both. My worry is that over-extending a language can hurt it and I'm not confident that the extensions involved would be minor enough that OCaml ends up trying to do too much and be too flexible. Adding features can break assumptions that are critical to other aspects of a language.
You're probably more experienced in OCaml than I am, so if you're confident that the extensions needed wouldn't break anything, then I'll accept that and go with the idea that that's all you really need. (Despite the occasional claim to the contrary, I listen to people with expertise I don't have.)
So we're down to Haskell, Erlang and OCaml, where one or more of these is extended in the way you've described to handle the extra facilities provided by the languages being dropped.
Wait a minute, your other post claimed that the paradigm didn't matter. Now you're claiming that some paradigms prohibit side-effects and thus compile cleaner to modern hardware. You may have a superior UID, but you need to work on your consistency.
You also need to consider that Fortran compilers have been around a very very long time. I don't know if you can find a Fortran IV compiler and a Fortran 2008 compiler that will compile to the same architecture (that age difference is a bit extreme) but I'm willing to bet that if you could find such compilers where both are written by good quality compiler writers (ie: not by someone high on mushrooms getting bored and writing a compiler for the hell of it), that well-written code that can work on both will compile to smaller, faster code using the more modern compiler.
You could argue that this is because compiler writers are better than they were. Well, define "better". Well, they know more about how to convert the Fortran syntax into machine code, what optimizations are suitable and what aren't, etc. Ok, so we have two variables in this - the variable called "experience in Fortran" and the variable called "experience in compiler writing". My claim is that the first variable is not only non-zero but is significant with respect to the second variable. Doesn't have to be larger, just has to be enough that it's demonstrably greater than noise. Your claim, if taken literally, is that the first variable is absolute zero.
Showing that the first variable is not zero requires that you have two compilers of the same age but where one language is newer than another. Ok, we can do this the easy way. Java and ADA are newer than C, so according to my claim GCC's C frontend should produce better code than GCC's Java and ADA frontends. The compiler is the same generation, so the same compiler experience exists for all three. The backend is the same in all cases. The only difference is in how the instructions are handled. Since ADA is actually very good on limiting side-effects, we can also examine the side-effect issue at the same time.
My experience is that GCC's C frontend is the better of the three. It produces smaller code and faster code, despite C's semantics (which, I agree, are absolutely crud). The only way this can sensibly be explained, given that they're of equal compiler generation, is that people know more about how to process C than they do Java (it's way newer) or ADA (which is middle-aged but not used extensively enough for experience to have built up in the community).
Let's now look at compiler technology itself. Much of the optimization is labeled "back arts" because it's not well-understood. Poorly-understood methods used "because they work" and not "because we know why they work" sounds crappy to me. What about code generation? Well, NASA used G95 for a long time rather than GCC's F90 because GCC's Fortran compiler kept blowing up. Sounds like there were problems. Ok, compiler design? Well, you're way old enough to remember the GCC vs EGCS vs PGCC debacle. Sorry, but I can't excuse you there. The sole reason for that ever happening was because each team made some good design decisions and some poor design decisions. If it had been clear-cut, you'd have had the same thing as you had with XF86 vs the original X Consortium, and XF86 vs the new X.Org - a universal and unambiguous switch with no split in the community, merely a split amongst the developers.
The compiler "crisis" caused a split in the community because each option had benefits the others didn't and the others weren't able/willing to produce something similar on their system. If compilers were truly modular, the PGCC optimizations could EASILY have been merged into EGCS and activated solely for those platforms those optimizations applied to. Didn't happen. The extra facilities being developed for "mainline" GCC could also have been merged into EGCS - some were, but others were rejected. In a modular design, you can't "reject" a module, you can merely not add it into th