Tag: science

Collated tweets on science and fiction

- 4 comments

I ranted this weekend on science in fiction on twitter–thought I’d collate those into a blog post. Warning: minimal editing (that’s why every sentence is around 140 characters ^^).

I’m getting a bit tired of “it’s not really hard SF” argument, which–oddly enough– often coincides with “this story is written by a woman/POC/other marginalised POV).  (before anyone asks, I don’t have a specific occurrence to point to; it’s just an accumulation of small things).

Thing is: the discourse about what constitutes hard SF–supposedly “real science! ™”–is very normative. It demands certain narrative forms, certain ways of addressing the reader, certain methods of expositing the science to make it sound plausible. Note that I said “plausible”, which means “what people are ready to believe”, which is different from “what is actually true”.

Even ignoring the problem of evolution of science (I’ll come back to that!), I’ve read hard SF stuff that was… er… out of date/inaccurate (I have a general science background, though my area of speciality is the mathematics of computer algorithms; and the H has a PhD in Quantum Optics, so between both of us we can muster passable science analysis). And that inaccuracy/out-of-date character *absolutely* didn’t prevent me or him from enjoying said books, btw–we don’t want to be snobs, and our enjoyment of books isn’t the cutting edge science. (actually, if I do want cutting edge science, I tend to read journal articles–though of course that’s pretty much restricted to fields of science I’m conversant with, so a pretty limited subset of everything that’s published).

I swore I’d come back to the evolution of science, so here goes: today’s science is likely going to be debunked (aka “evolve”) within a few centuries. 19th century science, pre quantum mechanics and pre general relativity, is vastly different from 21st century science. So any books set in, say, the 24th century that still rely on *today’s* understanding of science are a nice fiction. And, again, that’s OK. We’re writing/reading SF books, not journal articles, and requirements are different (real engineering specs make for bad fiction anyway, a bit dry!). 

There you go, afternoon rant. Would be interested to know what people think?

List of articles hosted in other places

- 0 comments

Articles on Language and Culture:
“On Worldbuilding, Patchwork and Filing Off Serial Numbers”, at Khaalidah’s blog
“Drawing Inspiration from Further Afield: fantasy set in non-Western Cultures”, at Aidan Moher’s blog
“Narrative, Resonance and Genre” at SFnovelists
“Traduttore, Traditore: translations, languages and cultures” at SFnovelists

Articles on Science, Religion and Plausibility
“Scientific plausibility”, guest post at Gareth L. Powell’s blog
“Atheism, Proselytism and other Isms”, guest post at Futurismic

Articles on Writing
“On the Persistence of Rules”, guest post at The Parking Lot Confessional
“Plotting your Short Stories”, guest post at Janice Hardy’s The Other Side of the Story

Science, engineering and large projects: SF in the 19th Century

- 6 comments

And now for something completely different: a few weeks ago, I complained on twitter that the science in SF seemed oddly stuck in the 19th Century, both the actual science research (which seemed composed mainly of individual mad geniuses in their garages having huge conceptual breakthroughs), but also its close siblings, the engineering projects that make up so much of SF (like building space stations, space launchers, etc.), and which seem to bear little relation to anything resembling real life.

I’ve complained about science here, but now for bonus points: engineering projects!

So, exhibit A. This is how a large-scale project looks according to most SF stories I’ve read:
.
Basically, a project manager who is God, or as near to God as matters, with anything from a hundred to thousands of (mostly) nameless, faceless grunts under him doing all the work. The story then tends to be either from the point of view of the project manager as he attempts to solve a pressing technical problem, or, less often, from the point of view of a harried grunt who has to solve a problem before the all-powerful project manager descends on them like the wrath of God (I’ve been nice here and thrown in an Assistant Project Manager, who will provide the necessary dialogue for as-you-know-Bob scientific exposition or provide a sympathetic ear to our grunt’s troubles).

Exhibit B: by contrast, this is what a real large-scale engineering project looks like in the 21st Century [1] (click to zoom):

Yes, it’s rather more complicated. There’s also two significant differences worth noting: one, the bottom boxes of the chart are not people, but team leaders, ie every bottom box still unfolds into your actual grunts. Second, I’ve cut at the level of the project manager to keep both graphs at the same scale, but there’s a significant extra layer at the top, which includes our project manager’s immediate hiearchy (his boss), a committee of peers (who follow the project and determine whether to continue funding it or not according to various Go/No Go criteria), and one or several sponsors (who champion the project within the company and to whom the project manager is accountable). Let’s not forget interlocutors outside of the company as well: the actual customer (ie the person paying for the delivery of the project; for instance, in the case of missile systems, the army is paying; in the case of a space station, you can imagine a conglomerate or a government paying…); subcontractors who have to be monitored, other companies working on related segments of the project (for instance, on a space station project, one company does the infrastructure, one company the climate control…).

So, yes, you’ll notice the same thing as with scientists: no project manager exists in a vacuum. They’re always accountable to someone for something (and when I say “accountable”, I mean all important decisions made are scrutinised, not that they’ll be judged solely on whether the project finishes appropriately. Will come back to “appropriately” in a minute).

Another thing is that responsibility is shared and diluted: note that the second org scheme has divided the satellite into different subsystems like the ground portion, the comms system, etc., and assigned different responsibilities within those subsystems. There is no grunt vs project manager system, but a carefully organised hierarchy of decreasing responsibilities fanning out from the system level, which ensures that everyone knows what they’re doing, and most localised problems do NOT make it back to the project manager, who has way more important things to do than concern himself with every little problem. On that same subject, a project manager is very seldom in the field, and most of their day is spent in meetings and in discussions with people (I always feel like laughing when a project manager on a space station spends their time touring the construction site and offering advice on stuff that most workers would take care of on their own…)

Finally, one thing that bugs me in engineering projects in SF is the lack of tradeoffs. Science tends to be “all shiny”, ie when a problem is posed, there is very often a perfect solution, one that meets all the needs and provides all that is expected. In real life, science is *never* shiny, and is almost always about compromises: things can be infeasible simply for technical reasons (for instance, no radio comms will provide the necessary reliability over the necessary distance), they can be infeasible for cost reasons (radio comms can be provided, but not within the allocated budget), and they can be infeasible because of time reasons (radio comms can be provided, however they will take eighteen months to be developed and tested, and we only have twelve months to deliver the system). In my line of work, we call that a QCD triangle (quality, cost, delivery): you simply can’t have all three items at the same time!

Now, coming back to that “appropriately”: a project is of course judged on whether it finishes on time, with the appropriate features and within budget (incidentally, a lot of SF projects never really seem to worry about either delays or costs…). However… you don’t wait until the project is finished to judge this! In addition to regular progress reports, there’ll be regular “milestones” which correspond to important decisions and/or steps in the project’s life. At those points, the project will come under scrutiny more intensely (by the peers, the hierarchy etc.), and will have to provide quite a few elements of justification for said decisions (and the project manager might well be part of a collegial decision process in those stages).

So, there you go, a short Engineering Projects 101–I’ve had quite a few years working on those by now (though admittedly mainly in a European work culture), and quite a few years reading SF, and so far I’ve been very disappointed by the portrayal of these. I might, of course, be picking up the wrong books/short stories/movies… Have I forgotten any gripes people have with engineering in SF? Are there any pieces that do a decent job of getting to grips with this kind of complexity? Feel free to argue/discuss/disagree in comments!


[1] Fake example for a satellite launcher. I copied it from a blog–not saying it’s a typical org, but it’s most certainly one that could exist and apply to a bona fide project.

What Stargate isn’t telling you about science…

- 4 comments

aka Common Myths about Science That Bug The Hell Out Of Me:

A scientist develops a great new invention in his lab on his own.
It used to be possible, but the days of great geniuses and polymaths like Newton and Descartes are past. Nowadays, it needs a team to develop anything. Most scientists work in teams, and so do most engineers. Someone might still have this amazing idea and make a breakthrough, but a complete prototype on his own? Not possible, as this involves several different areas of competence (see the point just below, too). Also, having people with similar competences to check what you do is usually a Good Idea, if only to make sure you’re not making any mistakes.
Just for the record, a simple prototype for a demonstration, built from scratch, should require at least a dozen people to handle the various aspects of the job. And that’s a bare minimum–like your student association building a robot for a competition. A company would have far, far more people designing the thing.

On a related subject: the scientist who knows everything about every field. He/she was trained in aeronautics, but is also a dab hand at biology, and chemistry too, when needed (I’m looking at you, Sam Carter).
Again, the days of polymaths is past. It’s possible to have vague knowledge of a lot of subjects, but to be able to make deep and sophisticated calculations in various divergent fields… You can’t be proficient in more than 2-3 connected areas (the BF had a wider education than most, and is still only proficient in physics and somewhat knowledgeable in computer science. He sucks in biology or mechanics. I’m good at computer science, reasonable in applied maths electronics, and suck at everything else).

The scientist(s) who has this great and amazing idea, and builds a prototype in a few days or a few weeks. Frequent bonus: the prototype survives field use and turns out to be perfectly operational.
Here’s the deal: developing anything is a long and drawn-out process, and field conditions are not a joke (sand that gets everywhere, weird temperatures… Your average materials are often going to take it badly). Building a prototype, even as part of a team, is more likely to need a year than a few weeks. And I’ll eat my hat if that hastily-conceived prototype is actually up to field conditions unless God takes a personal hand in the matter…

The aforementioned prototype is taken for an experiment, and no one keeps any backup anywhere. When it’s destroyed, people complain that they won’t be able to rebuild it.
Er, yes, OK. Sometimes it has happened. But this is BAD planning. Most companies/army research centres have backups and document every step of the prototype production. Not being able to rebuild it at all smacks of incompetence.

The scientist is setting up an experiment in field conditions, make modifications to the setup, and takes ages to relaunch the experiment (this usually happens when the bad guys are firing on the scientist’s position)
If you have, say, an electrical circuit and you’ve just rewired it with a few components, you don’t actually need to spend ages typing on the computer to make it work. It should be the equivalent of flicking a switch, and if you need more than a few seconds, I’ll start wondering about your actual skillset…

On a related subject, the scientist sets up an experiment in field conditions, and appears to have no idea what they’re going to do when the experiment fails.
Experiments have a protocol. They are actually prepared. You just don’t show up with your new shiny equipment and start fiddling with it in the thick of the action. Otherwise, the likelihood is that it won’t work, or worse, that you’ll fry something. And if you’re a good scientist/engineer, you’ve considered the fact that it might not work and have thought of one if not several backup solutions. Fiddling is all well and good, but the sad fact is that it’s seldom effective.

The scientist solves what looks like an amazingly complicated problem to the profane, but is actually quite a basic problem.
This is obviously a case of writerly misdirection and/or lack of research, but it’s really annoying when you happen to know a little about the subject matter (and nothing kills my trust in an SF show faster than this). The most recent example of this was Sam Carter bragging about writing an algorithm that searched through a database for words composed of a particular set of 18 symbols. This is a trivial problem (all the more so if the database is sorted, which they’ll have done if they have any brains).
A bonus case of this is when people start misusing scientific concepts. If I hear “logarithmic decrease” again, I’ll scream (a logarithm is a function that actually increases, and it’s also the one that has the slowest possible growth, so I’m not quite sure what a logarithmic decrease is supposed to be except obfuscation).

See, this is why I can’t watch most shows that are all about the cool science. I need characters to distract me 🙂

What about you? Any common misconceptions that drive you up the wall when you watch movies/series?