Follow the reluctant adventures in the life of a Welsh astrophysicist sent around the world for some reason, wherein I photograph potatoes and destroy galaxies in the name of science. And don't forget about my website, www.rhysy.net



Sunday, 24 August 2014

LEGO FTW


If you're reading this and thinking, should I buy a "Lego Mindstorms kit ?", let me save you the trouble and say yes. Yes you should. Because it's brilliant. Don't bother reading the rest of the post, because that's all you need to know.

I'm a great believer in self-rewarding for significant achievements. Completing my PhD got me a laptop, now 4 years old and still going strong - the second most expensive thing I've ever purchased, but by far and away the best. Publishing my thesis data got me a large glass brick (full of thesis data), which I use in just about every presentation possible. This time I'm celebrating publication of another data set, one which took an inordinately long time.

Lego Mindstorms has always been a case of, "yeah, looks fun, but I don't think I want to spend £200 on Lego." Fortunately, I had a £100 gift voucher from my friends at Arecibo when I left.
"I'll spend it wisely", I said.
"That's a shame, we were hoping you'd spend it foolishly", they said.
That was obviously a far better idea, so that's what I did.

I am not a hardware person. As a child, Meccano was OK, but I didn't have the patience to screw everything together (though for some reason typing in programs on a ZX Spectrum wasn't a problem). Woodwork was just a bit meh. Sewing was tedium itself. As for electronics, circuit diagrams give me a paroxysm of bewilderment while my soldering was the inspiration for the death scene of the T1000 in Terminator 2.


But even I can put Lego together from a kit. Plus I already know how to code in Python (which you can use to control Mindstorms if you're that way inclined, and I am), and the fact the kit comes with the parts for not one but five different robots sealed the deal for me (actually there are now official guides for fifteen !).

Each robot comes with 5 different "missions" which can involve modifying the robot and/or the code to control it. So far I've completed the missions for the most basic robot, the "Tracker". This is a simple tracked robot that demonstrates just about all the major features. It took me about 1-2 hours to construct.


In the box, you get bags and bags of >500 Lego pieces. Sadly there's no plastic container within the box to hold them all, so, having read another review beforehand, I went out and bought some Tupperware boxes. That was extremely sensible, since the raised rims of the lids stop the pieces from falling off when I'm gathering the requisite parts. The parts in each bag were pretty well organised so I kept them in the same system. It's simple to find whichever part I need.


I don't recommend just throwing all the parts in a big bag unless you want to give yourself massive Lego-rage later. Especially since there are a few parts you only get one of. Seriously, no, don't do it. Those boxes were quite a lot more full before the Tracker was built, and there's more in the shoebox.

Apart from the Lego, the box contains the instructions for the most basic robot (with handy parts guide and a ruler), and the cover of the box detaches to form a mission mat (where you can test some of the specially-designed programs). The full manual and instructions for the other robots are available online, unfortunately there's no printed version. But you're going to need to be on a PC to program the robot anyway, so that's not really an issue*. A nice thing about the online guide is that it gives you full 3D, rotatable views of each step of the build process, which can be incredibly helpful when assembling some of the more fiddly bits.

* I think it is possible to program the Lego directly, but I can't imagine why you'd want to.

Batteries aren't included either, so go out and buy a few packs of AA and AAA batteries (requires 6 AA and 2 AAA). You'll also, annoyingly, need a very small crosshead screwdriver to open one of the battery compartments.

The build quality is what you'd expect from Lego - genius. Unlike the cheaper imitations everything fits together pretty easily, but is robust enough to withstand some persuasion, when needed. Things don't fall apart unless you want them to. Assembling the Tracker was about the closest I've felt to being 10 years old again since Jurassic Park was re-released at the cinema. Sheer delight. In about an hour - a very, very happy hour - I had a little tracked robot roving around the house, servo-motors making pleasant whirring sounds.

Next step, add lasers.

The basic idea of Mindstorms is very simple - you have a programmable brick which can take input from different sensors and control different motors accordingly (the brick itself can light up, show simple images on its quaint LCD screen, and make sounds -  "INPUT, STEPHANIE !"). The demo program just makes the robot rotate and move a bit, then stop. Switching it to manual infra-red control was simple and a lot more fun. The robot easily clambers over small obstacles and even quite respectable slopes.
(sometimes the IR stops working and I have to turn the robot off and on again, but I haven't updated the firmware yet)


Things get more interesting with the various missions, which teach you how to program the robot and use the different sensors. Programming is done not by typing out code (that's possible but I haven't tried it yet) but by visually placing different blocks in a sequence. Each block can take input from the sensors (e.g. determine how far away the nearest obstacle is with the infra-red sensor), affect the program itself in some way (e.g., wait for a few seconds, repeat some action), or control a motor.

The instructions on how to create the program blocks were simple, though the interface itself could maybe be a little streamlined. It has a few too many similar buttons that do different things, for example, a + button in one place will start a new project, whereas a + button very close to it will merely start a new program. I'm not really sure what the distinction is for, other than causing confusion when I press the wrong button. However, actually "writing" the programs is very easy indeed - it's an intuitive drag-and-drop interface.


The only difficulty I ran into was enabling the Bluetooth connection to the computer. The programmable block can connect via USB, Bluetooth, or wi-fi (but wi-fi requires a USB dongle, not supplied). Unfortunately the USB slot isn't accessible on the assembled robot, so Bluetooth was the only practical option. It wouldn't have been majorly difficult to take the programmable block out, but it would have been irritating - the cables that connect the sensors and motors to the block are long and inflexible, making things a bit awkward.

Anyway, I found the secret was to enable Bluetooth on the block (duh), but not to use the blocks "connections" setting to find the PC. Instead, use the Mindstorms software on the PC to connect to the block. That seemed to do the trick. So now I can press a button on my PC and a little robot will start shooting plastic balls at some miniature rubber tyres. Which is immensely satisfying.

EDIT : After updating the firmware, the only way I can connect via Bluetooth is to make a connection via the Lego brick itself, then turn it off, then connect using the PC software. That's quite annoying as it takes 1-2 minutes to do this. On many (most) designs, disconnecting the brick would be very difficult or even impossible without disassembling the entire robot.

Things got more interesting with the other Tracker missions. The basic model has a spinning blade, while others have a bazooka (shoots plastic balls a few feet), a gripper (very cool, my favourite - robots with hands are just innately more awesome) and a hammer (classic Robot Wars). My favourite mission was with the hammer - the program makes the robot jerk around like it's developed an unfortunate disorder, hammer twitching spasmodically, until you sneak up on it from behind. Then it turns around and gives you a good thump. Wonderful.

All of these accessories connect interchangeably to the "medium motor" (the large motors are used to drive the caterpillar tracks) and demonstrate ingenious ways that rotational motion can be used to do cool things.


As you can see, a rather nice feature is that there are enough parts to build all the accessories separately - you don't have to take apart the bazooka to build the gripper. Sure, it could use a few more spares, but it's hardly lacking.

With the kit you get a touch sensor, a colour sensor, and an IR detector (which can find the distance to nearby objects). It can also, somehow, sense the direction to the IR controller. You can also buy extra sensors online (the manual even mentions that it's designed for use with third-party sensors), like temperature and ultrasound (a Lego sonogram , anyone ?). The output of the sensors was easy to understand in the program builder - it was trivial to make the robot stop when it detected an obstacle, for instance. You can even control the whole thing from a tablet, though I've yet to try that.

The recommended age for this product is 10+. That's just stupid. If I had a ten year old there's no way I'd share it with them. They won't enjoy it on as many levels as I do.

I'd love to tell you more but I think I desperately need to build the six-legged walking robot now.

Friday, 8 August 2014

If I Wrote The Press Release

#OverlyHonestMethods

Apparently, when scientists talk to the general public they're supposed to sound "excited" and "enthusiastic" about their work. That probably explains why they wouldn't let me write the press release about the latest paper. Here's what I would have written.


Astronomers and students using the Arecibo Telescope have found a bridge of gas 2.6 million light years long between galaxies 500 million light years away. And they seem to be pretty pleased about that, because finding gas is "a jolly good thing", they say, in amusing British accents.
The bridge of gas (shown in green) stretches from the large galaxy at the bottom left to the group of galaxies at the top. A third nearby galaxy to the right also has a shorter stream of gas attached to it. Picture credit: me/Arecibo Galaxy Environment Survey/The Sloan Digital Sky Survey Collaboration, http://www.sdss.org

Admittedly, the stream of gas is probably the largest known, a million light years longer than a gas tail found in the Virgo Cluster by another Arecibo project a few years ago. So that seems to make it fairly interesting. Gesticulating wildly, Dr. Rhys Taylor, a researcher at the Czech Academy of Sciences and lead author of the paper, allegedly made a bold, melodramatic statement that, “This was totally unexpected. We frequently see gas streams in galaxy clusters, where there are lots of galaxies close together, but to find something this long, and not in a cluster, is unprecedented.”

He then briefly passed out from the sheer excitement of it all.

While such allegations remain unsubstantiated, what he definitely said was, "This is a pretty hefty stream of gas. OK, it's really long, but that's no reason to take seven bloody months to referee the paper. I suppose it's kind of neat, maybe even exciting a year ago, but all of that has been ground out of me by the extremely tedious peer review process. Hell, seven months is only slightly shorter than a pregnancy. I think I need a nice cup of tea, or possibly a nice cup of whiskey."

Other researchers on the project remain more upbeat, probably because they didn't face a series of "badly-worded" and "downright strange" concerns to address from the referee. Roberto Rodriguez, a recent graduate from UPR Humacao, was so enthusiastic he worked on AGES data for two REU programs and two research-for-credit modules, and basically just wouldn't quit. "He even shows up now he's graduated," says project leader Robert Minchin, "which is great because now we have a willing vict- volunteer who does the work of a postdoc without any of the cynical ranting or the pesky need to provide funding and travel support."

Roberto enthusiastically explained, “We normally find gas inside galaxies, but here more than half of the gas – over 15 billion times the mass of the Sun – is in the bridge. That’s far more than in the Milky Way and Andromeda galaxies combined ! Also, on an unrelated note, Rhys totally bakes the best brownies ever. Have you seen his website ? I am saying this entirely of my own volition.”

The team doesn't have any convincing explanations for how the stream formed though. "Errr...." said Rhys, looking gormless, "Maybe one galaxy pulled the gas out of the other... because... umm... gravity ? Or maybe it just pushed all the gas out of the way. I dunno. What am I, a magician ?"

Sensibly, the team plan to use computer simulations to try and understand how the bridge formed. They haven't got very far yet, because all of their attempts to even form a stable gas disc have blown up in their faces - literally. "I just don't understand it," said Rhys, despondently, "It just sits there for about 400 million years happily enough and then it explodes for no reason. On the other hand, we've only been at this for three days, so give a guy a break."




The project involved three undergraduate students in the research work. As well as Roridguez, the team included Clarissa Vazquez, now a graduate student somewhere in America-land, and Hanna Herbst, now also a graduate student at the University of Florida. Dr. Robert Minchin, a staff astronomer at Arecibo Observatory, principal investigator on the project, professional beard-wearer and undisputed air hockey champion of Green Bank 2011, said “Student involvement is very important to us. We are proud to be inspiring the next generation of astronomers, and particularly proud of the involvement of Puerto Rican students, because they are the shizzle.”

Rhys adds, "Damn straight. I don't understand why Roberto is still here. If someone doesn't snap him up for graduate school, there's something very wrong with the system. But at least we get to keep him for a while longer."

STOP PRESS : With just hours to go before publication was due, Roberto announced that he had indeed been accepted into graduate school, and there was much rejoicing.

The bridge, as well as a whole bunch of other fairly interesting things, was found in data taken between 2008 and 2011 for the Arecibo Galaxy Environment Survey (AGES), which is using the awesome power of the Arecibo Telescope to survey a large area of sky with a high level of sensitivity. After a long-winded review by other, hopefully cleverer astronomers, the discovery has been accepted for publication in Monthly Notices of the Royal Astronomical Society. A pre-print of the paper can be found at http://www.arxiv.org/abs/1407.0016.

The Arecibo Observatory is operated by someone under a tremendously complicated and boring legal agreement with some other people.

Wednesday, 6 August 2014

Hydrogen, Dinosaurs, And User Support

Now that the bitter cynical ranting is over, it's time for a much more positive note about the research I've been doing for the past year or so. You can, of course, now read the full, magisterial* paper online, if you want the gory** details.

* Read, "fairly interesting".
** Read, "dull".

This is the second part of the trilogy of posts relating to my latest paper. The first was about the peer-review process that is the backbone of modern science, the third (coming soon) will be about what we discovered. In this one, I'll explain a bit about how we detect galaxies, without which we wouldn't be able to do anything at all. Specifically, this is going to be a very pragmatic post about how we look at the data and find what's in it; if you're more interested in how we get the data, you should wait for the third installment. What I'm going to describe here amounts to just three paragraphs in the paper, which should hopefully give you some idea of the work involved in a publication.

For the purposes of this post, the one-line summary is that we point our radio telescope at the sky, tell it to look for hydrogen, and then make maps of the hydrogen over a small part of the sky. We can measure not only how bright the emission is and where it is, but also what frequency (or velocity) it's emitting at. So we don't just get a normal 2D map to look at - we get a 3D data cube to search. And that's not easy.


You Could Probably Train A Monkey To Do This, But Not A Computer

In fact, finding hydrogen is very boring. It is. It really, really is. That's because although our data is three-dimensional, there are very few good programs* to let us view it in 3D. Consequently our poor students have to spend weeks looking at things like this :

* The redoubtable ds9 is one, but its features are limited. VisIt is an alternative, but it's so feature-packed it comes with a 30 page "quick" start guide (surely an oxymoron) and I can't get it to do anything.



What you're looking at is our data converted into a series of 2D slices. Each bright blob is a hydrogen detection (we'll get to why they look like that in a moment). What's even worse than having to view it in this unnatural, uninteresting way, is that there's no way to mark the detections when you've found them. You just have to try and remember which one's you've found. And with a particularly galaxy-rich data set like this one, that's simply impossible.

That didn't stop my student from doing an absolutely freakin' awesome job of cataloguing them, but it did take her about six weeks to find them all.

Now you may be wondering, why not just write a program to detect the hydrogen automatically ? Wouldn't that be faster and better ? In fact, no - in this case computers are, for once, slower and worse. The reason is that humans have had millions of years of evolution which give us pattern recognition skills that are, frankly, frickin' awesome. I mean, c'mon, look at this :



... or this...


... or this...


Now, even though I didn't tell you what to look for, you almost certainly spotted the lion and the tiger essentially instantly. Did you spot the velociraptor before reading the caption ? Don't answer, that was a rhetorical question. The point is, even without any instructions, you can identify threats in what are (compared to hydrogen data cubes) incredibly complex scenes with sufficient speed that you could decide to run away, if necessary.

Puerile though the joke may be, the last scene is actually a very good, nay, excellent analogy for hydrogen detections. The velociraptor is quite hard to distinguish from the foliage, just as hydrogen sources can be very faint. Also, the scene is dominated by a young lady afflicted by back problems, who is far more obvious than the raptor. In real data, artificial radio sources from satellites, radar, wi-fi, mobile phones etc. etc. etc., are often much brighter than the galaxies we're trying to detect.

Which is why one night a group of enterprising REU students decided to remove the road signs advertising mobile phone coverage. I mean, c'mon, they were stuck on the Observatory gate, for heaven's sake. I woke up to find one lying outside my front door.
It's obvious to a human observer that the velociraptor is much more threatening than our meditating friend - not so for a computer. You might be able to write a program to detect danger by looking for, say, nasty big pointy teeth, but asking it to be able to classify any possible animal and make a threat assessment is really rather a tall order. Similarly, it's not too difficult to write a program that finds hydrogen, but it's a lot harder to write one that doesn't find a lot of rubbish as well.

There's one other aspect of the meme that makes it extra perfect. Even if you haven't grown up terrified of velociraptors thanks to watching Jurassic Park when you were ten years old, even if you've somehow never heard of a velociraptor (God help you, you don't know what you're missing), you can still identify it as an animal. It just looks like one. Now, since velociraptors are all dead, we could call it a false positive in our search for threats. Similarly, we sometimes find things that look exactly like hydrogen detections, but aren't real. That's why we take extra "follow-up" observations whenever we're unsure.

OK, so our super-smart monkey brains can spot danger really quickly, and don't need anyone to do any astonishingly complicated pattern-recognition programming. Huzzah ! But now imagine that your task was not simply to detect and run away from the predator, but to make a quick sketch of it first (gosh, this is a good metaphor, isn't it ?). That's the crux of the problem - not detecting the hydrogen, but recording it quickly.


"Blender ? That's Your Answer To Everything !"

For some time, I'd been tinkering in my spare time with ways to import astronomical FITS files into 3D modelling/animation software Blender. I knew from the first time I sat down and catalogued a data cube that the software we were using was maddeningly inefficient. You couldn't even copy the coordinates of a detection to a file once you found one - you had to type them out yourself. It was a totally ridiculous waste of time* to spend days or weeks looking at blobs on a screen instead of trying to do actual science. And I knew that if I could only import the data into Blender somehow, all my problems would be solved at a stroke - the tools I needed were already an intrinsic, fundamental part of Blender**.

* A.k.a. "character building", or more accurately, "soul destroying."
** That's NOT why I wanted FITS files in Blender though. I just thought it would look cool.

I would not describe myself as a professional programmer and I tend to think of writing code as an option of last resort. Way back at the start of my PhD, I had no idea how to use the Python scripting language that Blender uses. I started learning it as a side-project to import simulation data (which was relatively easy - we'll get to why in a minute) for a friend. That gave me enough Python knowledge to try importing FITS data in various ways. You can read about my earlier efforts here. Initially I was limited to importing only the brightest pixels in the data - crude, but enough to see the data in 3D.

The same data as in the above movie, but now (crudely) rendered in 3D. Right Ascension and Declination are just position on the sky. Almost all the galaxies here aren't well resolved by Arecibo spatially, so they look like blobs on the sky. But they are very well resolved in the third axis (velocity) since they're rotating, so they look like long, cigar-like blobs in 3D. This is much better in .glass format.

The problem is that it's relatively easy to import and display a few thousand or tens of thousands of discrete data points in Blender, but it's very much harder to import a few tens or hundreds of millions of data points that form a continuous volume - like a FITS file. Anything less than that might be OK for making pretty pictures, but wouldn't be "science quality" - for example if the display was limited to showing only the brightest hydrogen, we'd miss the faint stuff, which can be the most interesting. But if you display the weak emission, you generally have to display the noise in the data as well, and that means you need to display huge numbers of data points.

This is what one million dots look like. To do anything useful with astronomical data, we need some way to visualise at least one hundred million data points.

It took another year or so in Arecibo, playing around in my spare time, before I hit upon a more useful solution than displaying a bunch of dots*. Trying to create a million little dots in Blender would be difficult, but it can easily handle an image of one million little dots. So by slicing the data into a series of images and then mapping each image onto a virtual Blender object (a simple flat plane), it would be possible to display the data without having to remove all the faint, potentially very interesting emission.

*Another problem with this methods is that in Blender, the dots can't have any colour information in the realtime view - so very bright sources look the same as very faint ones.

Now, at this point I could cut a long story short, but I'm not going to. Why ? Because it was bloody difficult, that's why ! With still very limited knowledge of Python, I needed a proof-of-concept that this method could work. I used a program called kvis (which we normally use for viewing the data in 2D) to create images of slices of the data. It only outputs in the obscure .ppm format, so then I had to convert them into something Blender can read like .png. Then, as an initial test, I manually loaded a bunch of these images into Blender, using each one to control the transparency of the plane it was mapped on to. This is massively impractical - if your cube has 100 pixels on a side, it needs 100 slices to show everything, and loading each one is boring. But it worked.

Pretty convincing for a bunch on planes.
Now I was really getting somewhere - I could view volumetric data in realtime without needing to cut any of the faint emission. The concept was proven, so I began working on this in earnest in between supervising a summer student, who was usually busy searching the cube the hard way with kvis, and finally publishing the data from my thesis. First, I had to find a workaround for a major problem with Blender : viewing the data from behind. Somehow, what looks great from one angle just looks mwwurrrgh (technical term) from the reverse angle.

Bugger.
After a couple of weeks wrestling with this one, I eventually consulted the Blender forums and found that the workaround was to create a copy of the image planes and put them somewhere else. Making a copy somehow "resorts" the textures so they look fine from the opposite direction. So then I wrote a little script to automatically change which images should be displayed depending on the orientation of the viewpoint. Swivel around too far and it automatically changes the view to that of the image copies instead of the originals.

That made things almost useful. A great deal of struggling with Blender's Python reference eventually allowed me to load the images automatically, which was infinitely more practical than loading them one at a time. The reference guide is, unfortunately, fantastic if you basically know what you're doing but not a great source for tutorials, but after a lot of trial and error I was able to make it work.

"Finally", I also had to teach myself matplotlib to avoid having to use two other programs to convert the data. That was relatively simple since the documentation was much better. This still only gave me the very very basics of what I wanted, but it was working : I could load any FITS cube I damn well wanted and look at it as nature intended, in 3D, with a few mouse clicks.


That wasn't the end though. Looking at it in 3D is nice, but not very useful by itself - the axes of the data need labels ! For that, I had to learn how to convert the pixel values into coordinates, something I'd been perfectly happy to let other programs do by magic. I never wanted to go into the details of world coordinate systems or write a routine to decide where it would be best to place the tick marks. But without this, no-one would ever take it seriously, me included. It would be a cute little gimmick, nothing more. Eventually I figured that one out too.


That meant I could also click on a source and know exactly where it was on the sky. So now instead of manually writing the coordinates down, you just click on a source and Blender calculates the position for you. Even better, you can add an object to hide the source, so that you never forget which sources you'd detected. You can turn the "masks" on and off if they're causing problems. Although you can do all this with other programs, they're not interactive - you have to enter the box coordinates manually and then re-load the data, which is much, much slower.



The upshot of this is that instead of having to spend weeks tediously poring over a data cube finding galaxies, you can get that stage done in about a day. It's about 50 times faster than using kvis. Conclusion ? Rhys wins. I didn't put that in the paper though.


Epilogue

Fast-forward about a year and the prototype had been developed into an all-singing, all-dancing FITS viewer now called FRELLED. Now it could load files with a user non-hostile interface very reliably. Another year or so and it could load simulations, cross-reference NED and the SDSS, plot contours and automatically create animations to impress people. Tasks that previously took minutes now take seconds, tasks that took weeks now take hours. If I had a TARDIS, I'd go back in time to PhD-me and say, "Rhys ! Forget the data analysis ! Learn Python, you dolt !".

Worryingly, that probably is the first thing I'd do if given access to a time machine.

Poster for the 2013 AAS in Long Beach, California.
Currently, after about two years of on-off work, FRELLED is more or less "core complete". It does everything I need it to, plus a few things I need it to do now that I didn't need it to do before. It's also evolved to the point where (I hope) the user really doesn't need to know anything at all about Blender. It's gotten pretty complex, so it still throws a wobbly from time to time when someone does something unexpected, but generally, through much toil, it's pretty stable.


You may be wondering, well, was it worth it ? YES OF COURSE IT WAS YOU STUPID PETTY FOOL ! HOW DARE YOU QUESTION - err, by which I mean, yes definitely, because it's not just useful for the incredibly niche aspect of looking at hydrogen data cubes. That's certainly what it's best at, of course.



But it can do a lot more than that. In principle it can load in any volumetric data set. Here, for example, is an MRI scan of a banana flower that someone converted into a slice-by-slice 2D GIF, which, with a small amount of tweaking, I was able to reconstruct this back into 3D in FRELLED :



Of course, simulations are even more fun because we can watch things evolve with time. Here's one that went wrong because the galaxy melted :


So, FRELLED has made my day-to-day life a lot easier, giving me more time to watch YouTube (don't worry, that's just astronomer-speak for "do ground-breaking science") and consume copious amounts of tea. In the future, it may become even more useful. Current surveys have at most a few thousand hydrogen detections... pretty soon, thanks to new telescopes and instrumentation, we'll be in the era of hundreds of thousands of detections. Rather than rely on those slow, unreliable programs to detect the hydrogen, with FRELLED (or something like it) and maybe a bit of crowd-sourcing, potentially we could let humans do the pattern recognition they're so good at without eating up years of astronomer's valuable time. I call that a success.

Tuesday, 15 July 2014

Review : God's Philosophers


Science in the medieval period was virtually non-existent. The Catholic Church did not take at all kindly to anything that contradicted its rigid, inflexible teachings, and the few people brave enough to speak out were summarily dismissed as heretics and burned at the stake. At least, that's what we're all taught. And that's what many people today would like to believe. But it simply isn't true, says Dr James Hannam, graduate of both Oxford and Cambridge, in his fascinating 2009 book God's Philosophers. You can read the first two chapters of the book for free on his website.

Rather than a detailed review of the style of the book, instead I'll give a short summary of some of the aspects I found more interesting. I'm going to concentrate on the philosophical and moral aspects of the relationship between science and religion; if you want more direct examples of how highly devout medieval Christians contributed to modern science, you should read the book itself.

As far as style goes, suffice to say it's very accessible, with a good balance between science, philosophy and theology. You don't need to be a scientist, philosopher or theologian to understand it. However, to get anything out of the book you'll have to be prepared to surrender many of the impressions of the medieval Church you've probably grown up with. If you're one of those people convinced that only science has led us out of the darkness of religion and into the light of reason, well now's a perfect time to put your money where your mouth is. You're not allowed to extol the virtues of science unless you're prepared to question your own viewpoints - otherwise you ain't no scientist at all, boyo.

Hannam is under no illusions that science (or natural philosophy as it was then) and religion did not sometimes come into direct conflict, and that when push came to shove, the Church had the upper hand. But the extent to which that meant the Church actively suppressed freedom of thought has been vastly exaggerated, and the achievements of medieval theologians (yes, theologians) that were crucial for the breakthroughs of Copernicus, Kepler and Galileo - who were all devout Christians - have been unfairly airbrushed from history. It's not an attempt to convert anyone, only to point out the debt modern science owes to medieval Christian thinkers. They not only rediscovered the works of ancient philosophers, but surpassed them.


Theology, Doubt, and Natural Philosophy

As a fervent and unswerving agnostic, the concept I found most interesting from the book was the idea that theology and science were closely linked. Although I was aware that medieval universities existed, I've normally dismissed them as being irrelevant theological schools, producing nothing of any real consequence. This was a huge mistake. Theological training, says Hannam, not only included the study of mathematics and natural philosophy, but was virtually obsessed with logic. The underlying idea, it seems, was that by studying the natural world one could understand God. Natural philosophy was seen as, "the handmaiden of theology" - not perhaps a very flattering label, but completely at odds with the (mis)conception that science and the Church were mutually exclusive.

The obsession with logic produced some genuinely very interesting consequences. Could God create a weight so heavy he couldn't lift it ? No, said the theologians - that would be a logical contradiction, and even God isn't that omnipotent. Moreover, not even the most devout scholar (and remember that these people were members of the Church) believed in the literal truth of the entire Bible. They couldn't, because that is simply impossible. If you took the Biblical phrase, "four corners of the Earth" literally, you'd believe it was flat*. The idea of interpreting the scriptures figuratively goes at least as far back as St (emphasis : Saint) Augustine, 354 - 430 AD.

* Which they didn't, even by 1000 AD - this is a more recent myth about medieval scholarship. Other examples included various passages which state that the Earth does not move, the interpretation being here that it meant "from the perspective of someone standing on the Earth".

In a somewhat roundabout and almost perverse way, the Church actually encouraged freedom of thought. While initially the teachings of Aristotle were almost treated as Gospel, the omnipotence of God would allow him to create anything he liked (so long as it wasn't a logical contradiction). So it was perfectly fine to contemplate vacuums (which Aristotle didn't believe in) since, even though they might not exist naturally, God could potentially create one if he wanted. By today's ideologies this is a truly bizarre way of overcoming difficulties with mainstream theories, but it was certainly better than assuming the ancients had already got natural philosophy licked - most of Aristotle's ideas were complete gibberish. His prestige kept his ramblings at the forefront of scientific theory long after they should have been rejected, but medieval inquiry was (albeit very, very slowly) able to come up with better ideas.

Interestingly, God's ability to do as he pleased in no way hindered investigations into the workings of the natural world. God was seen as the primary cause of all things, but he usually operated by invoking secondary, natural causes that proceeded to operate with strict rules. Moreover, if God didn't like what you were up to, he was free to cause a miracle to stop it. If you were sick, you could try using magical remedies to get better, but if God wanted you to be sick, then sick you would damn well stay*.

* At least one popular image of the medieval world appears to be entirely correct : doctors were something to be avoided like the... err, plague. Prayer actually did have a much better chance of success - supernatural deities aside, at least priests weren't going to bleed you half to death as a "cure".


The Burden of Proof

When something was found that disproved a statement in the Bible, or a decree by the Church, this was accepted. The flat Earth is one example, the fact that the antipodes are inhabited (a notion condemned by the Pope sometime in the 8th century) is another*. When the level of proof was 100% certain - as in finding people living in the antipodes - then even matters of faith had to give way to rational science.

* The Church was, however, quite right that the hot, sweaty tropics are uninhabitable. The fact that people stubbornly continue to live there anyway is beside the point.

A running theme of the book is that (for instance) the image of the Earth being at the center of the Universe was the most rational, logical viewpoint given the evidence available at the time*. Proving the Earth is round is easy; proving it goes round the Sun with only the evidence of your eyes and nothing else is very, very hard. It's extremely difficult to see the world as medieval thinkers would have, though Hannam makes a valiant effort. Of course, we know today that the acceleration of a rotating Earth isn't sufficient that we can all feel ourselves whirling round at tremendous speed, but at the time, that was a logical, sensible reason to reject the notion of a rotating Earth.

* Interestingly, Earth wasn't placed at the center to reflect its importance - quite the opposite. Heaven was the most important place (the medieval Universe believed to be about 90 million miles across), so anything further away from the surface of the Earth was closer to God. Conversely, anything deep in the ground was closer to Hell and therefore worse. We were literally living on the surface of a "Middle Earth", if you like.

Hannam makes the important point that science cannot function where every idea requires absolute, irrefutable proof. For instance, I wasn't around when the Earth started forming, so I cannot proove God didn't do it. But to make any further progress, I must assume (based on other lines of inquiry and well-tested theories) that this is the case, and proceed from there. That is the essence of a scientific principle, something which unfortunately seems to have largely escaped the medieval mind. Having speculative models was fine, as we'll see, but being allowed to assume they were true was quite another. When it came to theorising, religion definitely had the upper hand over natural philosophy. And that caused some very acute unpleasantness when Copernicus's model of the Solar System was found to be much more accurate than the then-mainstream Earth-centered view.


The Limits of Freedom

There were most certainly limits on what ideas the Church would put up with - cross them, and you would indeed meet with a very nasty fate indeed. But those limits were very much larger than I realised.

The Inquisition wasn't a good thing, but don't confuse the Papal and Spanish Inquisitions. The latter (see Toby Green's Inquisition) was a brutal, oppressive system that was largely politically motivated, and did indeed place very tight restrictions on what you could and could not say. The former, however (as Green also says) was quite different. If you admitted your "crime" (and by today's standards of course they would not be crimes at all) to a Papal inquisitor, you'd be set free (though woe betide repeat offenders). Not so with the Spanish Inquisition, where torture was common and confession still meant death more often than not. The Inquisition of the Papacy was not a nice thing, but the Spanish Inquisition was far, far worse.

So what would it take to be declared a heretic ? Well, quite a lot, actually. You could speculate about almost anything, so long as you were clear to state, "this is just an idea, I don't know if it's really true." Take Cardinal Nicholas of Cusa (1401 - 1464), who had the notion that maybe the Universe was limitless and the Earth was just another star moving through space, and not in the center at all. He even postulated the existence of aliens. He was a Cardinal, for crying out loud, and no-one thought anything amiss with this. It's worth noting that this idea was based on the idea that the Universe would have to be limitless to "reflect God's majesty", not for any scientific reasoning. There was absolutely no scientific observation (or even discussion - Olber's Paradox being centuries in the future) at the time that gave any evidence of an infinite Universe.

What about Giodarno Bruno (1548 - 1600), that noted visionary burned at the stake for believing that the Earth wasn't at the center of the Universe ? Nope. Bruno was, quite simply, a nutter.  He seems to have tried to come up with an entirely new religion based on magic, plagiarised more competent thinkers (though Galileo also did this) whose mathematics he simply couldn't understand, and had the annoying habit of loudly telling everyone he was a genius. Basically, he would be this guy :


Now, being a bit of a jerk and believing in magic are pretty stupid reasons to convict someone, but, although he did believe the Earth went around the Sun, that's not what he was investigated for - it wasn't declared a heresy until 16 years after his death. Unfortunately, the list of charges has not survived, so we can't be sure what the real charges were. Certainly, Bruno was an example of the Church behaving at its worst and this is clearly an example of suppression of freedom of thought, but Bruno wasn't a martyr for science by any stretch of the imagination. He was a magician and a mystic, with no more claims to scientific genius than Cardinal Nicholas.

Not that even heresy was always a guaranteed immolation, mind you. Virgil of Salzburg (700 - 784) not only got away scot-free for teaching the antipodes were inhabited by people not descended from Adam, but was later even canonized. William of Ockham (he of the razor) also escaped, although in this case quite literally by hiding under the protection of the Holy Roman Emperor. And Galileo (more on him in a minute) was, for a while, allowed considerable leniency by a somewhat corrupt and capricious pope.

Before returning to Galileo, it's worth noting one very clear example of religious doctrine impeding scientific inquiry. Nicholas of Autrecourt (1300-1369 AD) attempted to claim that everything was made of atoms. Unfortunately, this appeared to make the transubstantiation in the Eucharist (bread into flesh and wine into blood) not merely miraculous, but a logical contradiction - something cannot appear to be made of bread but really made of atoms of flesh. Nicholas was forced to recant and the offending document burned, but he himself gained a cushy job as a dean and lived more or less happily ever after. Hannam contends that this is a rare example, and if he'd only not made the link to the Eucharist so explicit and just been a little more careful to emphasise that it was "only an idea" (which would have been sensible given the lack of definitive evidence at the time), everything would have been tickety-boo. He doesn't use the words "tickety-boo" though, which is a shame.


And Yet It Moves ?


Galileo is the archetypal heretical scientist. Unlike Bruno, he was certainly no mystic, and his ideas were based on solid observational evidence. Ultimately, this one does boil down to science versus religious ideology... well, sort of. Maybe. It's far from as open-and-shut as you might think.

Many discoveries essential for Galileo's work had already been made. Copernicus had published his Sun-centered Solar System model (dedicated to the Pope) in 1543, with a careful foreword by his friend Osiander stating that it was only a hypothesis. Interestingly, this seems to have been more to ward off scholarly skepticism than Church wrath. At the time, the model seemed at odds with the observation that the constellations do not change throughout the year. Since the medieval Universe was only 90 million miles across, if the Earth moved through a significant fraction of that, the constellations would appear distorted as it approached the sphere of the fixed stars.

Copernicus's solution was the correct one, but so dramatic it was very hard for everyone else to accept. To reconcile the theory and observations, he increased the distance to the stars by a factor of a billion. Changing theory has never been a popular move - the idea you need to create multiple Universes to kill a cat still causes problems for quantum theorists today, as does the idea of changing Newtonian gravity to fit observations of galaxies. We seem to have an innate tendency to prefer to think that we're basically on the right lines, unless the evidence becomes overwhelming. And in 1543, Copernicus' case was not overwhelming.

But it worked. There seems to have been little or no ecclesiastical backlash (perhaps because the book was so mathematical, suggests Hannam) and by the 1570's a Papal commission was using the Copernican model to develop our modern calendar. There was no getting around the fact that it produced much better results than the alternatives, though it seems that people were happy to accept it as "only a theory". The trouble only began when people began to believe it might also be a true description of reality.

By 1588, Tycho Brahe had demonstrated that the planets were not moving in giant crystal spheres. His model of the Solar System still had the Earth at the center, but the planets orbited the Sun rather than the Earth. In 1596, his protégé Johannes Kepler published a model with the Sun at the center, and no-one seemed to mind. Very few people believed the idea, but some priests stated that it while it was religious unobjectionable, it was scientifically unsound.

So three great scientists had already postulated alternative models of the Solar System that directly contradicted scripture and no-one had even got so much as lightly singed. For a long time, Galileo also got along just fine, conducting scientific observations and publishing them in what amount to 17th-century popular science books. He was also on good terms with the Pope. As we've seen, no-one at the time was particularly bothered by Copernicus or Kepler's theories.

That all changed though with a particularly strict Inquisitor, Cardinal Bellarmine. Galileo held the opinion (as did many others, theologians included) that the Bible should be taken figuratively except in matters of morals - leading to the famous quote, "The intention of the Holy Spirit is to teach us how one goes to heaven, not how the heavens go.". The official Council of Trent had said much the same thing.

Bellarmine was having none of it. His view was, unusually, that the Bible was the literal word of God and should be taken as such unless absolutely irrefutable proof was given. The Italian friar Foscarini had recently forced the issue by writing a letter tackling the problems with scripture and the movement of the Earth head-on. It failed utterly. His letter was banned and Copernicus' work suspended pending corrections. Galileo got off with a warning not to teach Copernicus' theory. Hannam's implication (he does not say this explicitly) is that Bellarmine was the main proponent responsible for turning the previously unobjectionable Copernican model into a heresy, but I would have liked a lot more detail on this point.

Four years later, in 1620, the Church released the corrected version of Copernicus' work. It was hardly a ruthless censorship (as the Spanish Inquisition would have done) - ten corrections, released as a special insert, in a book hundreds of pages long. Hannam gives the example that "admitting the Earth moves" became "assuming the Earth moves", emphasising that it was just an idea, not a proven fact. Which, after all, it wasn't - this is scarcely worse than having a paper or thesis go through peer review today (albeit with potentially more extreme consequences in the case of failure to comply).

Bellarmine died a year later, and soon after Galileo began work on a a popularisation of the Sun-centered Solar System, confident that as long as he was careful, his friend the Pope would support him. And perhaps he would have done, had he not (according to Hannam) made a single catastrophic error of judgement, by what amounted to mocking the Pope. It's worth noting, however, that initially it did pass muster from the Papal censors and it was allowed to be published. Only when the Pope read it did things get ugly.

The problem, it seems, was that Galileo did not have a good enough explanation of tides. His idea that they were caused by the rotation of the Earth, with the water being left behind and so sloshing around. He not only tried to use this to prove the Earth rotates (even though he'd been warned not to) but for the counter-argument in the dialogue of the book he used an argument by the Pope that God can create any circumstances he wanted (basically saying that, "no, the tides are caused by the will of God"). Worse, he put this toward the end of the book, which the Pope took as adding insult to injury.

The resulting trial didn't exactly see Galileo stand up for his ideas, and I can't say I blame him. Initially claiming he didn't support Copernicus at all, when pushed he admitted that, "well I suppose some moron might read my book that way" (not a direct quote, unfortunately). No-one but believed him, but, threatened with torture, he stood his ground and called their bluff. Satisfied by his apparent conviction, the threat was withdrawn. Though he avoided the rack or a burning, he was condemned to life imprisonment under house arrest.



Conclusion

Hannam does  an admirable job of demonstrating the debt modern science owes to medieval theologians. Within this very limited scope, the book is excellent. It would have been nice to have added a least a few related points though - in particular, it's clear that Universities offered protected intellectual havens, but almost no mention is made of how academic ideas were regarded in society. Similarly, there's no discussion on what the academics thought about more broad issues : morality, the Crusades, etc. Although this lack is understandable, given the obvious subtext of the book it would have been nice to include something.

There are certainly examples from history of the Church suppressing freedom of thought, sometimes brutally. But these are rare, and in the case of scientific inquiry, practically non-existent. The ultimate weapon, the auto de fe, potentially could have been used against troublesome scientists, but it doesn't seem to ever actually have been. The fact that the Church was ever allowed to have any say over what people could publish and could punish (or even execute them, if only in principle) for stepping over the mark was not, of course, a good thing. In practise, this was extremely rare. Medieval theologians contributed far more to modern science than the Church ever held it back. At least, Hannam does a good job convincing me of this.

Maybe it's possible to argue science would have been better off without the Church, I don't know. But at the very least, the popular idea that the Church only worked to hinder progress is now simply untenable as far as I'm concerned. It's also worth remembering that even in the modern age, espousing viewpoints too far from the mainstream may not get you burned alive, but it will certainly get you laughed at and ejected from academia. Given the intensely rational nature of medieval theology, the line between heresy against scripture and unconventional ideas is, perhaps, thinner than we might like to think.

Perhaps the most glaring omission from the book, however, is any discussion relating to what happens next. Hannam concludes with the trial of Galileo. What a future version of the book is badly in need of is an epilogue to summarise the next few centuries. What I would especially like - and this would probably need a whole other book - is some discussion on how we went from being able to say, "well, obviously the Bible isn't meant to be taken literally about everything" in mainstream medieval theology, to the current view of a disturbing number of people that the Earth is only a few thousand years old.

Far from condemning modern scientists as heretics, a medieval theologian - if properly briefed on the current evidence - would almost certainly have little good to say about modern creationists. These ideas do not take us back to a medieval world view - they're much worse than that. Even Cardinal Bellarmine would be forced to concede to modern evidence. Small wonder that Frank Herbet once quipped that in religion there is "always the unspoken commandment, thou shalt not question !". Ironically, such a view is a far better description of modern extremists than it is about the logical, questioning mind of the medieval theologian.

Wednesday, 2 July 2014

Referees Who Move Goalposts Make Lousy Peers

My latest paper, The Arecibo Galaxy Environment Survey VII : A Dense Filament With Extremely Long HI Streams, has finally been accepted for publication. This has consumed, in one way or another, about the last year or so of my research time. So I'm going to indulge myself with not one, not two, but three entire blog posts. In this first one I'll look at the peer review process, which for this particular paper was slightly less fun than being given an enema by a bear. In the next post I'll describe how we went about searching for galaxies using Blender, and in the third, thrilling conclusion, I'll say something about what it is we actually learned.

I imagine that giving an angry bear an enema would also be unpleasant.

Peer review does not mean writing articles in the Guardian about what you think of Jeffrey Archer, unless you're a journalist. In science, the idea behind peer review is simple : someone writes a paper, and someone else (usually anonymous and selected by a respected academic journal) checks to make sure it makes sense. In this way, scientific literature is pretty well-defended against giant space bananas and moon landing conspiracy theorists. No-one think's it's perfect - a perfectly objective system is probably impossible - but it's the only system we've got for checking that everyone else is basically sane.

In this case it took over seventh months from first pushing the big scary, "WARNING : THIS WILL SUBMIT THE ARTICLE ! ARE YOU ABSOLUTELY, POSITIVELY, DAMN WELL BLEEDIN' SURE YOU REALLY REALLY WANT TO DO THIS ?!?!?!" button to having someone say, "This looks decent, let's publish it." Almost as long as creating a baby, though less icky, and as you can imagine I'm none too happy about that.

Although now I come to think about it, if it's a choice between one or the other, I'll take the publication, thanks.


I've never had to referee a paper, mercifully. I don't imagine it would be much fun. Several weeks trying to make sense of other people's wranglings with no time for your own research can't be much of a hoot. But I have received several referee's reports on my own work. Typically, this results in my mental processes going something like this :

First reading
"Add error bars in figure 6 ? What a JERK ! No-one else puts error bars on this plot. This guy is clearly a complete amateur."
Two hours later
"Huh, errors aren't so hard to calculate after all, maybe I should add them in instead of insulting the referee's mother at the next submission..."
The next day
"Hey ! With error bars this figure looks waaay better !"

Which is precisely why this post has been in draft form for several months, but I'm not going to even try claiming an unbiased viewpoint.

Referee's comments come in all varieties, from the extremely helpful (which are probably the majority, if I'm honest) to the extremely unhelpful (more on them later). A few are just downright strange. One I've seen every single time is, "shorten the text." This is like saying, "There are too many notes, that's all. Just cut a few and it'll be perfect.". It doesn't make any sense. Tell me which bits you think are too long, or I'm just not going to do anything*.

* Actually what I'm going to do is cut out one paragraph just so I can say I've done something. What I'd like to do is remove the final word from every sentence.

What's almost worse and just as common is that the referee will simultaneously state that the paper is too long and require major additions. Referees, please don't do this. It makes you sound like a crazy person. You have to at least suggest where the cuts should be made, or it makes no sense. I can't possibly make it any shorter if you're telling me to add stuff.

In like vein, the weirdest, most incomprehensible suggestion I've received so far was that I should add "supporting figures to the paper" (there wasn't any context to this). What, you mean like a picture of a giraffe or something ? A self-portrait in crayon ? Maybe a drawing of Atlas, he's a supporting figure, after all. Love to help, but I'm not psychic ! Of course, the assume-everyone-is-psychic factor affects us all, but not usually to that degree.



Fortunately, this particular mystery was solved in consultation with a friend, who brilliantly realised :
"It means there aren't enough characters. You need to flesh out your story with supporting figures that the referees can relate with. Maybe add a downtrodden scientific paper reviewer, who started out with the bright-eyed idea of changing the world through peer reviewed science, but instead met paper after paper of nonsense and flim-flam, and now has no joy left when a paper of some real worth comes through."

Much more annoying are criticisms of the second draft that could have been made at the first submission. Now obviously there's nothing wrong with saying, :
"Oh and I forgot, you need to correct the spelling of GALAXIE on page 2 and make the caption bigger."
No-one can spot every typo at the first reading. But if you're going to suggest :
"You should replot all your figures so that they actually show something completely different, because although I didn't bother to mention it last time, I think they're all wrong. And also I don't think you wrote your code in C, so please do that."
... then that's fine too - it's the referee's job to do this - but don't suggest this at the second draft when you could do so at the first. This is unfair and unprofessional - it's moving the goalposts, which is traditionally frowned upon. Especially if it's done by a referee. Also, it really wastes a huge amount of time - I could have made the changes months ago and have the paper ready for publication already.



Perhaps the most annoying of all is to repeat verbatim criticisms of the first draft that the author quite clearly, unmistakably and - above all - directly addressed in the second draft. As in :
Referee : "You need to say how many galaxies are in your sample."
Author : "There are 5 galaxies in this sample."
Referee  : "You need to say how many galaxies are in your sample."
That's not a real example, but some comments really are every bit as inane as that. They're even worse when they're longer comments that took correspondingly longer to address. I can only assume in these cases that the referee didn't actually read the author's response - which makes it a no-win situation. If you're not going to even read my response, why should I bother making corrections at all ?

The worst example of this was a referee who insisted that there were better data processing techniques we could use. Great ! It was obvious to anyone that the method we were using, while an accepted and well-tested approach, might not be ideal. So in my response I asked the referee to provide a reference so that I could try this new approach for myself. Unfortunately in the second report the reviewer didn't answer my question at all, but instead did a copy+paste job of their previous comment.

Sigh.

But we all have bad days. In my response to the second report, I again asked (twice this time, just to make sure) for a reference to these shiny new methods. In the third report, the referee flat-out ignored this altogether, merely stating simply that our measurements must be wrong - and worse, that not only were the numbers wrong, but all of the major results were also wrong. Even though they hadn't raised any previous objections, and simultaneously stated that the paper was now "much improved".



In this case this rather brusque dismissal, at the third iteration of the paper, was particularly galling since I'd spent a very, very long time making sure those measurements were as good as I could damn well make them. The end result was quite nice, agreed with previous measurements (where available) and wasn't particularly controversial. All lines of inquiry pointed to the same result. To have the referee then suddenly dismiss them - after not raising any previous objections - without giving a clear reason why was simply too much, so we asked for a second referee.

Now, as an aside on the much-vaunted objectivity of science, it may interest readers that referee 2.0 had no qualms whatever with the data analysis, which was vindicating. Whereas referee 1.0 was perfectly happy with the introduction, referee 2.0 thought parts of it were incomplete and misleading, and provided a list of some genuinely interesting papers in support of this. Which is great, and I happily made the requested changes (although I'm not sure I agree with the referee's sentiments, a happy compromise was easily reached)... but, it does, of course, expose a basic flawed truth of peer review : not all scientists agree with each other. But referee 2.0 provided justification for their assertions, whereas referee 1.0 didn't.


Make no mistake : I support the peer-review process. And philosophically, I don't like the notion that if you disagree with the reviewer, you should find another reviewer - but let's not go nuts. Sometimes, just like everyone else, reviewers are just jerks. Peer review may not be the perfect, objective solution we'd all like it to be - but it's a damn sight more objective than not doing it at all.

In short, reviewers :
  • Specify what it is you want the authors to do. Authors aren't psychic. They'd like to be, but they're not. This means vague instructions really aren't helpful. Asking the author to "do this bit again, but better" (that's a real example, albeit paraphrased) won't work !
  • Don't make non-constructive criticisms. Ask questions rather than stating that the authors have missed something obvious - it at least gives the impression of benefit of the doubt. There's no point whinging that the paper has lots of typos - send them a list instead*. As above, vague, hand-waving instructions help no-one.
  • Don't move the goalposts. Spend longer on the first draft so that the authors will be able to address as much as possible at stage one. Telling them that they need minor revisions at the first draft, then saying major revisions at the second draft (after the previous comments have been addressed) is misleading, unprofessional and delays publication. And if you do find a major error at the second draft, have the common courtesy to apologise for not spotting it sooner !
  • Demand rigour, but don't be anal about it. Life's too short to worry about half-spaces, or, more importantly, the precise meaning of a word. Insisting that the author uses a long-winded phrase (e.g. "galaxies in the overdensity between redshift 0.02 and 0.03") just in case a commonly-used word (e.g. "population" - seriously, referee 1.0, WTF have you got against the word "population" ?) is misunderstood helps no-one.
  • Check that the author hasn't already addressed your point in the paper or their response. There's really no reason for the author to bother if you're not going to read what they write - it makes the whole process meaningless and unworkable.
  • Remember that not everyone is a native English speaker. Many of my non-Anglo/American colleagues report that referees can be very unhelpful when it should be obvious enough that the author might not speak English as their first language !** If you suspect this is the case, then criticising their writing flaws is simply rude. Don't do it. Instead, give them a list of corrections.
  • If you can't justify your assertions, don't make them. The same applies to the authors, of course. Unless you can point out why something is wrong - and, if at all possible, how to correct it - don't say what you think is wrong. If the authors have justified what they've done, you may disagree with them - but unless you can prove that they're wrong, that should be the end of it. You've got to allow people to publish things you disagree with.
* And no, they can't be "easily avoided." If you've read your own work a dozen times it becomes literally impossible to spot any more typos. Fresh eyes are needed, and it makes a lot of sense to me that those eyes should be provided by the referee/editor. To say, "please check more carefully in future" is condescending.
**Helpful hint : if the author's name is Pierre von Hindenberg, chances are they're not a native English speaker - and with a name like that, they're probably way more awesome than you will ever be.