User:Armondikov/Now/May12

From RationalWiki
Jump to navigation Jump to search
Icon bullshit.svg
Thought of the Now

...half-baked ideas from the mind of someone who is almost certainly dumber than they think they are.

Archive


26th May 2012[edit]

Philosophy of Chemistry

Resonance structures[edit]

Apparently, "Philosophy of Chemistry" is a thing. Now, I can understand a lot of philosophy of science, in a general sense, and the nature of epistemology. Yet I can't quite figure out what philosophical conundrums arise purely from chemistry alone. So, I read on through both Wikipedia and the Stanford Encyclopedia of Philosophy and... well, frankly I never had much respect for Philosophy as a discipline, and reading through these entries has effectively counteracted the respect it slowly gained from me over the last couple of years. Though a single throwaway line in a Wikipedia article (that I might just delete), there is the example of resonance structures. For the uninitiated, a resonance structure is like this:

Nitrate ion resonance structures.png

Resonance structures allow us to move bonds and charges around using the same formalised framework we use to represent molecules. They let us visualise how charges distribute and so let us overcome some of the restrictions that we're faced with a formalised structure system. Secondly, to a first approximation the more possible resonance structures the more stable a chemical moiety will be. This helps visualise (I won't say "explain", as the relevant WP article does) why certain bonds a stronger, for instance while the amide group is a stronger linkage than an ester group, and why the amide is locked into a planar conformation while the ester has much more rotational freedom.

Philosophy of Chemistry seems to ask if these structures are real. Except any first year will tell you that, emphatically, they're not real. These representations are diagrams, they're there as tools so we can communicate structural ideas with a pen and paper. You may as well philosophically ponder whether bonds are "really" black pencil lines and whether atoms are "really" etched letters! Bonds aren't "real", at least not in the sense that they're solid connections between atomic centres. Indeed, if you want to plug a structure into a quantum mechanical calculation (or a suitable approximate method like DFT) you don't need bonds, only co-ordinates, and the functionals will fill in the gaps. The little balls and sticks are there as aids for us, not true representations. No 21st century chemist will tell you any different; there is nothing to philosophically ponder here.

Semantics[edit]

But that's a throwaway line in Wikipedia that probably won't last to the end of the day if I have anything to do with it, what does the SEP say? Well, going through this is arguably worse than the analogous Wikipedia article because while the WP authors have buggered up chemistry, the SEP authors are living in the 18th and 19th centuries. Seriously, quantum mechanics has basically superseded most of this "philosophical discourse" on the nature of atoms and molecules - something that makes me pretty sure that the "philosophical discourse" on the interpretations of quantum theory will later be superseded too.

Anyway, what else does the SEP claim is philosophy of chemistry? Well, while a lot of it is based on very outdated scientific ideas (nothing new there, Searle's Chinese Room argument is a rehash of the homunculus argument, which I don't think has ever been taken seriously) the rest is based on semantics: what is an element and what is a substance? The problem here is that you can define things however you want, the question is have you actually succeeded in changing anything because of it? The answer in this case is nothing.

Let's take the example of isotopes as given in the SEP. Elements are classified according to their proton (or atomic) number, rather than their mass. Isotopes are elements with a different mass, due to a changing number of neutrons. Generally, isotopes of the same element are considered to be chemically similar - yet there are some exceptions. The "philosophical" question, therefore, is to ask whether these differences make them different "substances"? Yes or no? For the most part, the assertion that isotopes of the same element are chemically identical holds; but no-one educated in the 20th century or beyond says that this is literally true. It would be like saying ethanol and methanol, due to the presence of an OH group, are identical: they're not, though they possess certain qualitative similarities. We don't change anything by classifying isotopes as different and unique substances, we merely end up arguing semantics like a bunch of amateurs on an internet comment board. At worst, we destroy the similarities that make categorisation within the periodic table so powerful by creating redundant groups. Besides, physics already has its own version of the periodic table that does this!

What of deuterium, though? Though an isotope of hydrogen, its actually chemically different. Why? It's nothing to do with the qualitative nature of the isotope and the bonds it forms, but of the quantitative strengths of those bonds. Bond strengths are related to mass, and the differences show up nicely on spectra that probe the vibrations of these bonds. So, in comparing the vibrational spectra of HCl and DCl we can see two clear absorbances, one due to HCl and one due to DCl. They're widely separated, the mass of the deuterium is significantly different enough to quantitatively alter the bond strength so that we can see it even at low resolution. Yet, look closer and we can actually see four different absorbances - because we also have 35-Cl and 37 Cl present. The difference is quantitative and can be done by considering the reduced mass of the combinations of H, D and 35-Cl and 37-Cl. Dueterium is twice as heavy as protium, whereas 37-Cl is only 5% heavier than 35-Cl.

So, are they "different substances"? Yes, in the sense that 35-Cl and 37-Cl are different and H and D are different. But this difference isn't the same, by any stretch of the imagination, as how H is different to Cl.



15th May 2012[edit]

Philosophy of Science: Error Appreciation

Pretty much every undergraduate lab report will ask for some degree of error appreciation. This is one of the key aspects of science: we're only as accurate as our instruments are, and our answer is only as good as our worst piece of data. So, invariably students are asked to identify where errors can have crept in. So consider the experiment I've been supervising and marking recently, where they have to react two substances and determine the kinetic rate of the reaction. By running the reaction at different temperatures, you can determine the activation energy of the reaction. The grizzly grisly (apparently) details can be found here. So what are the potential errors? Usually, I get a bullet pointed list like this:

  • Accuracy of measurements with pipettes.
  • Accuracy of the spectrometer.
  • Human error.
  • Temperature might not have been constant.
  • Insufficient mixing of compounds.
  • Recording of the reaction didn't start right away.

They're first years, so I'll usually give them a tick for this, it's the sort of ball-park area of what they need to be thinking of, but writing errors down like this doesn't convey much information. It's just a list of keywords and nothing more, as if they're just - to inevitably quote Yudkowsky - trying to guess the teachers password. Effectively, these don't relate to observations. Instead, we need students to answer a far more specific set of questions about error appreciation; how doe these errors manifest in the results, and how do you go about compensating and correcting for them? I'm not going to expand upon all the points above, but I'll rewrite a few key pieces and what I consider to be better answers:

  • Errors in the pipetting of solutions can lead to an inaccurate value of [OH-] (that's "concentration of hydroxide ions", for the uninitiated) which will lead to an incorrect value of k (the reaction rate constant) as k = kobs/ [OH-]. The data point on the Arrhenius plot will then be imprecisely placed, altering the value of EA determined.
  • Non-constant temperature alters the rate of reaction throughout, this will manifest as the ln(kobs) vs time being non-linear. To minimise this, only the linear parts of the data are used and the first 10% of the reaction is ignored to allow the sample to first come to equilibrium.
  • The accuracy of volumetric glassware used was +/- 0.1 cm3, small pipettes were +/- 0.01 cm3. The fluctuations in spectrometer read-outs were +/- 1% due to fluctuating background incidence.

The above is by no means comprehensive, as I'm not going into detail about how to quantify the errors involved, but at least they relate back to observations and, importantly, how to correct for them if possible. Many students correctly identify the need for repeat measurements, but few seem to realise why this is. Last term featured a similar spectroscopic experiment where the report asked them to comment on accuracy and precision. Accuracy is how much you trust the value, precision is how reliably you can spit out the same number of significant figures. Citing a value like 44.5589 is no good if you repeat and it comes out as 42.8841 and another repeat shows 45.3156. Appreciation of errors is of fundamental importance in science, because despite what the science was wrong before crowd will have you believe, 99.999% of all new science is done within these error boundaries.

I'm delivering a feedback session on this in a couple of hours and they're going to get a fucking earful.

11th May 2012[edit]

More God crap...

This is a question that has bugged me for some time; do all monotheistic religions that worship the creator of the universe actually worship the same god?

I'm sure I've pointed out before that what theological arguments prove is simply "the universe had a cause". Well, they may not necessarily prove it as such, but that's complicated and philosophical and requires a clusterfuck of definitions about what causality is and whether the universe needs a necessary or efficient cause and, while I'm sure that's very interesting, it's not much of a practical help. Most followers of the Abrahamic religions accept God has having the same properties; properties such as creating the universe, knowing everything that happens within the universe, ability to do anything with the universe, and so on. These are agreed upon by those philosophical arguments to prove that there is a God. So, are they the same God, despite the difference in religion and the piddly details to go with it? Ietsism, ignosticism and a few other belief terms would suggest something like this; religions are all worshipping the same god, really, but haven't defined the details yet or worked out the right details, because there must be a generic "God" type thing "out there" or whatever. This generic "God type thing that's out there" is what the majority of cultural Christians seem to believe in; a lot of casual believers who just grew up in the UK with "Church of England" being synonymous with "default" believe in God, but are either unaware or unfussed by the concept of original sin and repenting to Jesus - hence why I distinguish between general Christianity and the Cult of Jesus types I know. These people, who aren't quite in the spiritual but not religious category just yet, seem to have this vague notion of worshipping the creator of the universe. Ietists by another name, perhaps?

Anyway, in the case of Islam and Christianity, the connection to worshipping the same God is historical and pretty self-evident. They both stemmed from the same bronze-age mythology, interpreted differently and rehashed for a suitable location. Yahweh and Allah both mean "God" in their respective languages and both religions give the same definition of the all-powerful, loving, creator of the universe that speaks to people in a way that is completely and utterly indistinguishable from them just being mental. So they're the same God, really. Some people spot the connect and seriously think they're the same. Which raises the question of whether He's happy being worshipped in different ways; one way thinking that a man was His son and spend their weekends drinking his blood while the others follow the OCDs of a desert dweller who married a girl a fraction of his age?

Maybe, who the fuck knows.

But that's just the Abrahamic cluster. What about Brahman, the supreme universal spirit of Hinduism? This doesn't have the same roots as Judeo-Christian and Islamic sects. So, is Brahman the same entity by definition? What about the analogous force in ancient religions like Atum? Because they created the world and are all-powerful transcendent entities, they must be the same thing by definition. As always, it's the "by definition" part that bugs me the most. It enables religions to all feel that they're right because by definition they're worshipping the right thing - even if they're drinking blood when they should be washing their toes five times.

So all these entities have the same properties; if any theologian from these respective faiths/religions was to argue the existence of their particular god they'd mention the exact same observational properties. God is all-powerful. God created the universe. God knows everything. That's how this thing interacts with the universe. If they're different entities, and the worshippers don't think the other's exist, what is the difference? How can we tell the difference between Yahweh, Brahman, Atum and Allah?

7th May 2012[edit]

Philosophy of Science: Naive Views

One of things I'd certainly want to address if communicating philosophy of science would be a lot of the more naive views of the scientific method. Can science say anything for certain? Are our current theories right? Will our current understanding the universe stay constant and unchanging forever? The answer to all of these questions is a resounding NO... but they come with some caveats. For instance, we've done the double-blind tests required to show homeopathy is complete cobblers - that's sufficient because that evidence isn't going to change. Consider the following comments I've recently seen:

...you will be faced with new discoveries, ones that may contradict what you thought was the case before. I think if your scoffing at any idea brought to the table, you are showing yourself in a bad light, not in the spirit of things at all.

Like fuck. New discoveries can change where theories enter the unknown and the predictive. They can't change existing knowledge or evidence. No single observation is going to prove that the world is flat, rather than spherical.

...the clever scientist is the one that doesnt say he knows something, the idea of empirical certainty in a system without a unified set of rules is bogus, untill we get a complete picture, the best that can be said is 'probably'.

True, but only in the naive sense of how science works. When we say "probably", we often quantify it - and well-calibrated uncertainty is far more useful than absolute certainty (which is arguably impossible). But this in no way implies the converse that the remaining 0.001% we have, after saying something is 99.999% likely, is a valid and likely option!

Obviously, this must stem from some kind of faux open-mindedness, the desire to look good and open-minded for its own sake. But, really, this ends up being synonymous with "gullible", and we shouldn't treat that as a virtue.

7th May 2012[edit]

Philosophy of Science

For some time I've been wanting to assemble a philosophy of science course for undergrads. However, I'm often mindful of the following phrase, attributed to Feynman:

Philosophy of science is about as useful to scientists as ornithology is to birds.

Arm-chair philosophers have apparently said birds would find ornithology actually quite useful, but to be honest I don't think they quite get what Feynman is on about here. To teach philosophy of science to scientists requires some knowledge of what actually goes on in a lab. What use is the difference between induction and deduction when you're looking an an IR spectrum? How does knowing the Popperian definition of "falsifiability" actually help in experimental design? Is the history of paradigm shifts going to help you bring about the next one? All this is usually overlooked by those discussing philosophy of science, who tend to be philosophisers first, and researchers second - if ever. So, instead, I'd be tempted to call this: "How To Be A Proper Fucking Scientist"

So, how do you act as a Proper Fucking Scientist? I think this would be a far better question to answer than merely "how do you teach philosophy of science?" Indeed, I'd be tempted to rip up the usual ways of communicating philosophy of science and even exorcise the P-word entirely from it! This is about making better scientists, not better essay writers, and so a focus on errors that are made time and again by almost every student is a vital component of assembling this How To Be A Proper Fucking Scientist module:

  1. Good and Bad science - this would effectively let us discuss demarcation without being too obscurely intellectual about it. But it also allows subtlety as real science doesn't fit neatly into the "real" and "pseudo" classes that non-scientist skeptics like to use. This is really the core topic.
  2. Experimental design - how the principles we can learn from understanding heuristics and biases, as well as more arm-chair-philosophy concepts like "falsifiability" apply in an experiment. How to go about formulating an hypothesis and then testing it rigourously - but importantly, how to structure it so it can be tested.
  3. Error recognition - knowing what constitutes a proper appreciation of experimental error and how to correct for it. This is often so badly done, even by second and third year students who still think "human error" is a useful term.
  4. Science communication - understanding how your ideas are transferred accurately from one person to another. How we present and how we review.
These are four core branching points that form the interface between practical science and philosophy of science. Arguably the first two are very conventional, and it's the latter two that really focus in on the practicality that is almost never taught properly. Combined, this should enable students to be taught it in a way that will actually benefit their scientific endeavours.

4th May 2012[edit]

Atheist or Agnostic?

Here is Neil deGrasse Tyson expressing with extreme elegance a lot of things that I think and believe:

Although we do part ways somewhere in the middle. I agree on everything about the labelling and "isms" and "ists" so much I want to stamp his words all over my life, yet unlike Neil deGrasse Tyson (who should only ever be referred to by his full, unabridged, awesomsonic name like that) I actually find "agnostic" to be the far less useful term than "atheist". If I had to pick between them, I'd go with "atheist" as the most accurate and less douche-baggy of the two.

Why do we have a term for non-belief and people who lack a religion? I see the point about how we don't have words for, to go back to the tired and pathetic cliché, non stamp collectors.There's no need for that term, and herein lies the problem with that cliché; stamp collectors make up a minority of the population. Having a word to describe the majority of the population who don't engage in that activity would be somewhat pointless and a waste of good syllables that could be used for something more productive. To view why we have the word "atheist" we need to step back and look at the world from the viewpoint of a religious believer, a world that evolved to entangle itself with the concept of God. So, perhaps a better comparison would be to ask why we have a word for amputees; they simply lack a limb or two, why have a word for it? It's because historically - and still today - religion is all encompassing and ubiquitous. For many people, vocal and passive believers alike, religion is a necessary component of ones personality, life and identity. Someone missing an arm or a leg stands out from the crowd as incomplete, and in a world wrapped up in religion so strongly, an atheist stands out just the same.

If you fill in a form, you'll likely spot a box marked "religion". Create an online profile, there'll be a section for "religion". "What religion are you?" is a more common question that "are you religious?" - and even then, it's "are you religious?" in the context of "are you evangelical and fervent about it, or just a simple believer?" We don't, to quote Neil deGrasse Tyson's examples, see the same thing with skiing and golfing. At worst we see someone holding up a sign saying "Golf Sale" on a street corner, we don't see every town and village with a monument dedicated to golf, with people singing the praises of golf and wondering why you don't play golf and why you hate golf. So yes, there is a reason, and a fairly good one, to have atheistic terminology. It's strongly context driven. As much as it might boggle the mind, it's useful; after all, we still have "dark" as a secondary quality to describe the simple absence of light.

But what about "agnostic"? Am I agnostic? Perhaps, if you want to call it that - I don't particularly care. However, here's the thing about the way agnostic is used; it's used to say "I'm not 100% certain about it". At worst, it's used to say "I'm not 100% certain and never can be". It's true that when it comes to professions of religious belief that we cannot be 100% certain of them, so why is it absurd to state it openly? It's absurd precisely because it's true - it doesn't convey any useful information. It's true about all beliefs, not just ones we arbitrarily label as religious. 100% certainty, certainty that we've excluded every conceivable and inconceivable alternative hypothesis isn't possible. This is true for everything, so why the need to state it? I can be agnostic about the Judeo-Christian God in the exact same way that I'm agnostic about Allah and Brahman, and equally about the existence of ghosts and ghouls, the Invisible Pink Unicorn and Carl Sagan's garage-dwelling dragon. We can make up entities and rattle them off one-by-one and we'd all be agnostic about them and their true existence.

In fact, with agnosticism of the strongest variety we specifically put these beliefs outside the realms of confirmation, and disconfirmation, by making them float in their own special region that isn't causally connected to us. The absurdity here is that "not causally connected to us" is pretty isomorphic with "does not exist", and is the logical conclusion that we generate - it's what we actually expect to see - when we say something cannot be proved or disproved in both practice and principle. So declaring agnosticism is far more absurd that declaring a mere secondary property like "atheism". Why do it? Because it maintains some faux open-mindedness, it distances from the associations that an atheist has professed certainty in their belief (confusing degrees of belief with degrees of certainty should be a deadly sin) but effectively allows you to have this perceived middle ground between believing and not-believing - even though declared agnostics emphatically don't believe in higher powers and religions. We don't see self-described agnostics splitting their time evenly between synagogues, mosques, churches, meditation and the Richard Dawkins forum - see, degrees of belief and degrees of certainty aren't the same thing!

Either way, agnosticism and atheism are both redundant terms that don't convey much information. Yet they are still used because religion is given such a prominent place. Even so, I said I agree completely with Neil deGrass Tyson; I'd prefer not to be descried as either.

3rd May 2012[edit]

Fun times.

This is one of those rare "I'm having a sandwich" type blog updates. Come the end of the year, my research group is getting its own shiny new building. That means first moving out of our current one ready for it to be demolished. Of course, it's not that easy because we're stocked not just full of chemicals but also full of very expensive NMR instruments; namely a 9.4 Tesla cryostat which needs to be shifted into storage.

But you can't just move these things, they're a superconducting magnet sat in a vat of liquid helium only a few degrees above absolute zero (which in turn is in a vat of liquid nitrogen to insulate it against the ravages of an air-conditioned room!). Superconducting means that the coil offers no electrical resistance, so won't heat up as any other electrical device does - from phone chargers to computer processors. But what if it warms slightly? In that case, it may lose superconductivity, which means resistance and heat, which means more warming, which means more resistance, which means more heat... This reaction runs away with itself and will quickly boil off the helium and, eventually, the nitrogen too. So, about 30 litres of liquid helium and nearly 100 litres of nitrogen will expand into a gas and race out of the vent towers on the top of the magnet cryostat housing. These vent towers are pretty small (they have to be, to keep the helium in) and so the pressure release creates quite a sight (if you're into that sort of thing). This is known as a quench, and is exactly what we will be doing - deliberately and in controlled conditions, of course - tomorrow.

Anyway, here's what happens if you don't do it properly. The silver foil is the insulation foam from the vacuum layers basically being pulverised by the sudden release of gas, while the smog is the sudden condensation of water vapour in the air as it's rapidly chilled by the helium:

Oh, incidentally, remember when the LHC was knocked offline couple of years (time flies!) back, almost just after it was switched on, and it was out of action for months? This is pretty much what happened, except instead of one unit it was a third of the entire LHC's ring that did it.