« Dragged kicking and screaming ... | Home | Time for a new audio/video pla... »

Titanic Inquiry Project

Sat 5 Jun 2010 by mskala Tags used: ,

While looking for something else I ended up on the site of the Titanic Inquiry Project, and it's fascinating. The Titanic sank in the Spring of 1912 after colliding with an iceberg. Very much can be said about the number of lifeboats there were, whether the procedures for filling them were followed correctly, whether other aspects of preventing and responding to what happened were as they should have been, to what extent the ship's design was or wasn't right, and so on. The British and American governments each launched inquiries into the disaster, and the linked Web site carries transcripts of those inquiries, thousands of pages of them. I forgot the time and spent hours reading the transcripts when I was supposed to be doing other things.

Techie stuff

There's a lot of material up there, and even if I had nothing else to do it would take me a long time to read it all. I would like to do that eventually but it may not happen any time soon. I did, however, read most of the testimony of a naval architect called Mr. Wilding, before the British inquiry.

The chief designers of the Titanic died in the disaster. Wilding was the closest thing the board of inquiry could get to them: another naval architect from the same firm. Most of his testimony is what some people call "engineering porn": detailed examination of intricate technical details of how the ship was built. They wanted to know exactly how the watertight doors between different sections of the ship worked, for instance, so he told them that, and how far apart the stiffening ribs were in the bulkheads. The answer to that latter was 29 and 7/8 inches, so they compared it with the Lloyd's requirements for ships of this size (Titanic was not meant to be built to Lloyd's requirements but to Board of Trade requirements, but they made the comparison anyway). Those standards said 30 inches. So they cross-examined the witness - your construction practice was less than the Lloyd's standard? No, my lord, it was better than the Lloyd's standard; the standard is an upper limit, and putting the ribs closer together is better. And on and on.

Many people today have the idea that the Titanic's construction, in particular its design as a series of tubes watertight compartments, was in some way supposed to be revolutionary or experimental, but it really wasn't. Lots of big ships like the Titanic were built that way, and there were in fact lots of big ships like the Titanic built. The only significant design innovation, it's clear from the transcript, was the ship's size: it was the biggest built to date. But even that was not a dramatic departure; it wasn't all that much bigger than others. There's a very strong sense that the Titanic fit well into the existing engineering culture, and Mr. Wilding goes into much detail of comparing the Titanic's design to that of comparable ships, discussing issues like whether it is or isn't a good idea to use the space between the inner and outer skin for storing coal. Some ships were built that way; others weren't; the Titanic wasn't; there were complicated advantages and disadvantages to both approaches, spanning safety and commercial and other aspects of the design; he favoured not doing it, but other experts might well disagree. The whole thing sounded very much like technological discussions we still have today, even though it was basically a "late steampunk" level of technology.

I was also interested by the infrastructure technology of the inquiry itself. This was never explained directly in the part I read, but the image develops of the board sitting in a room where one wall is covered by a huge diagram of the ship. At one point there's something they talk about that isn't marked on the wall, and they have a brief side discussion of whether they can have the witness mark it for them right then and there. Where's the stepladder? Can we find him a pot of red paint? Do we really need it marked on the wall, or can we just mark it in our own copies? The piles of paper on the table, with everyone having their own copies, must in themselves have represented a huge amount of work for somebody. In 1912 they had typewriters, and some sort of larger-run printing. Xerography, for photocopiers and laser printers, wouldn't be invented until 1938. Fax machines had actually been invented in the 1840s, but were hardly in general use.

And then think about the mathematics. It was a design requirement that the Titanic should remain able to float if any adjacent pair of its watertight compartments filled with water; so before they built her they had to calculate that that would be the case. Think about how to do that: you have to numerically integrate the shape of the hull as a function of depth, so you know how much is displaced when the waterline is at a given point, and then you have to invert the function (again, numerically) to find the waterline for a given displacement. If you want to take list or pitch into account (i.e. if the ship tilts in some way so that the waterline isn't horizontal on the ship's profile) then you end up having to do the whole works in three dimensions. And you have to do it all without a computer and in wacky Imperial units of measure.

The authorities asked for the draught versus displacement curves while the ship was being built, and the designers declined to provide those because it would have taken them three months just to do the calculations. They did, nonetheless, produce a lot of points on the curves, because they had to have such points to be able to build the thing at all. That would mean a skilled expert doing nothing but arithmetic for let's say a working day just to get one number. Wilding produced in the inquiry a set of diagrams showing "What happens if these two compartments flood? What happens if those three compartments flood?" and so on. Those diagrams are a heck of a thing to do with 1912 mathematics, but people were doing such things routinely. I'd have been tempted to just build a model ship and sink it in a bathtub to see which compartment-flooding combinations it could withstand, but that wouldn't have been trivial to do properly, it wouldn't have convinced the Board of Trade to let them build the real one, and in fact they probably did it too anyway.


The Titanic sideswiped the iceberg while taking evasive action. As a result the ice cut into the side of the ship over a long distance - they said at least two hundred feet - thereby allowing water to enter more of the watertight compartments than the design was intended to secure against. Wilding calculated that based on the rate she took on water, the holes totalled about twelve square feet. Not twelve feet square, but the area of a rectangle twelve feet by one foot. If that were one hole, since it contained points 200 feet apart it would have to have been only about three quarters of an inch wide. He emphasized that he didn't think that was the actual profile of the damage, only an equivalent based on the known information. Much more plausible was that it was really several small holes some distance apart. Nonetheless, the image I get is of a can opener slitting the ship wide open.

Mr. Wilding asserted his strong conviction that if the Titanic had steamed straight into the iceberg head-on instead of what happened, then the disaster would certainly have been much less severe. That's a heck of a thing to say - but he cited the known case of a ship called the Arizona that had done something similar. The fore end crumpled up, killing everyone in that part of the ship, but in so doing it absorbed the ship's momentum, after which the ship remained afloat and made it into port with relatively few casualties. The impact wouldn't even throw people out of their bunks, except for those unfortunates crushed in the crumple zone.

It's a bit hypothetical, but the "better a head-on collision" concept raises some interesting moral questions. Suppose you're the captain, you have heard and believe Mr. Wilding's assessment above, and you find yourself in charge of the ship as it's headed for an iceberg. If you remain on course a collision and some loss of life is inevitable - but probably not a disaster of the scale of Titanic. On the other hand, if you try to evade, you have some chance of escaping a collision entirely; and if the collision occurs, it will be much worse. What do you do?

If you know for certain that the evasive action will be successful, obviously you must take it: better no collision than the less-deadly collision. If you know for certain that the evasive action will not be successful, you had better not take it: better the less-deadly collision than the more-deadly collision. In between those two extremes there must be some level of riskiness for the evasive action at which the best decision shifts from the one to the other.

But it gets even more complicated, depending on your theory of morality, because it could be said that choosing to put lives at risk with the evasive action is qualitatively different from choosing to doom some people (even, people you may be able to specifically identify) to almost-certain death by ramming the iceberg; different in a way that goes beyond the mere number of lives lost. I think a reasonable moral theory can be constructed under which it would be said that you're required to abhor taking direct actions that sacrifice persons' lives, even if as a result you statistically cause a larger number of deaths. If you don't abhor such direct actions, you can quickly be led into the strict-utilitarian "Slaughter one healthy person to provide transplant kidneys for two sick ones!" system, which most of us find distasteful to say the least. If we take the theory that abhors directly killing people, we might be led to take the evasive action even when it's quite unlikely to succeed, because at least it doesn't involve direct sacrifice of human life; better to give everybody at least a chance.

I had a lot more to write here about moral issues, specifically the "women and children" business, but then I mistakenly closed my browser without saving, and the first draft of that was lost. This is already a lengthy posting anyway, so I think I'll cut it here and save the further commentary for a later date.


Steve C
Human nature makes this easy to answer:
The vast majority will pick the action that results in the possibility of a positive outcome, even if it is unlikely.

How many charged the Normandy beaches on D-Day? Of those, how many knew they could have shot themselves in the foot the day before and avoided being sent into that blender of war? Steve C - 2010-06-06 10:00
What people are likely to do is a different question from what they morally should do, though also interesting. On that question I think you're right: people are likely to make decisions assuming the best possible outcome, or at least a fairly good outcome, even if the statistical likelihood of it doesn't support the decision. People buy lottery tickets, for instance. That would suggest the captain in my example would be likely to attempt to avoid the iceberg even if the chances of doing so are slim and a failed attempt would be significantly worse than no attempt. I don't know if that's also the morally right decision. Matt - 2010-06-06 10:31
Your discourse on 1912 computation (I prefer that word to "mathematics", because the maths of 1912 was plenty sophisticated - it's the process of coming up with numerical answers that was onerous) reminds me of how ephemerides were calculated and who were the "computers" of the time. Computers were experts who designed procedures for wielding humongous functions. Now the analytic representation of a planet's motion consists of a polynomial in time (the mean position) followed by sums of hundreds of tiny periodic terms to account for the gravitational tugs from the other planets. Typically these will be coefficient x sin(A*a ± B*b ± ...) where coefficient is a few arc seconds, or a fraction of an arc second, A, B, etc. are integers, and a, b, etc. are the mean-position polynomials of the other planets. My files for the Mars series alone add up to over 350k, to give an idea of how big the sums were. The computer would devise a sequence of operations involving little slips of paper tucked into pigeon holes on the wall. Then they hired (or commandeered) students. Each student would pick up a slip of paper from the assigned pigeon hole, perform the task written thereon - presumably with the help of trig tables and Comptometer adding machines - and place the result in the appropriate other pigeon hole for another student to come along and do the next step. Or so I've read. Sounds like fun.

Speaking of Titanic, did you come across the time zone the ship was in when the iceberg was struck? The hour was recorded - I don't have it handy right now but it exists - however the clocks on board were changed regularly to account for the westward travel. When I corresponded with Titanic experts many years ago they couldn't tell me exactly what setting the clocks were at when that fateful hour was recorded. Axel - 2010-06-06 10:53
I suppose I should indeed have said "computation", but I also had in mind the genuinely mathematical issues of the techniques used to do such computation. Computers had to know a whole lot about approximation and how to arrange things to do the minimum amount of work and correct for human errors. Really, we still need to know a lot of that stuff in order to use our digital computing machines effectively - some of it falls within my own field of expertise. But the idea of human labour being heavily involved in the process certainly makes it feel more romantic. My instinctive feeling is that in the days when computers were human beings, those people had to have a much closer connection to and at least some of them a deeper understanding of what they were doing, than users of electronic computers tend to have today.

Any kind of labour-saving device can have that effect, even if it's just a table of precomputed numbers. Aleister Crowley wrote disparagingly of astrologers who would come up with meanings for things like "Venus opposition Sun" without thinking through how likely that would be to ever happen or what it would physically mean. Their ready-made hand computations had already taken them far enough from the natural phenomena under study, that they could make such mistakes.

Note that you can quite easily get Venus opposition Sun in a harmonic chart, even if you hope never to see it in the sky. Is that what technologists call a bug, or a feature? Even if you think harmonic charts are an unworthy modern innovation in the context of Western astrology, I think they have longstanding traditional use in Vedic astrology (though my knowledge of that is very limited).

On timekeeping: we have got testimony from the "Marconi operator" (bear in mind that a lot of basic radio technology was under patent in 1912!) of the Carpathia, one of the ships in the area, discussing time zones here: http://www.titanicinquiry.org/BOTInq/BOTInq08Evans01.php The site's search engine is a little tricky to use (I couldn't figure out how to do a phrase search, and it returns results in an internal form filtered for stopwords, which makes them hard to read); so although I found that quickly, there may be some better material buried in other pages.

The Marconi operator doesn't directly say what time the Titanic was using at the time of the accident, but he says that as of 6:30pm ship's time, at longitude "49.9 West" (I don't know if that's really 49.9 degrees, or 49 degrees 9 minutes, but really 49.9 degrees (i.e. 49 degrees 54 minutes) would make more sense), his own ship was keeping time 1 hour 55 minutes different from New York. The Titanic disaster occurred at longitude 49'57" (by modern measurements from when they sent down the submarines). The difference of three minutes of arc in longitude corresponds to only twelve seconds of time, so I think it's reasonable to assume the Titanic's ship time at the time of the disaster would have been synchronized with the Carpathia's: one hour fifty-five minutes ahead of New York. Matt - 2010-06-06 12:04
Er, make that the Californian throughout... I had the name Carpathia in my head because that was another nearby ship that ended up involved in the rescue. Matt - 2010-06-06 15:33
Steve C
Ok another WWII example is when Churchill had Germany's codes and the Germans decided to bomb Coventry. Churchill chose not to warn Coventry so that the Germans wouldn't know their code was broken. I'd say that's as close to dooming the front passengers while saving the ship as you can possibly get.

But I'd also say that the point you consciously decide to sacrifice "some" to save "many" based on probabilities then it invalidates it as a moral choice. You are now making a rational choice and the "morally right choice" no longer exists.

The act of measuring morality can destroy it. Steve C - 2010-06-06 21:05
I always say computers do for our brains what cars do for our leg muscles (which brings up the distinction - rarely acknowledged - between performance and intrinsic ability). That's "computers" in the modern sense, of course. As for astrologers - tell me about it! Most of them couldn't find the Pole Star on a clear night. Anyway, Matt, thanks for the note about Californian. Axel - 2010-06-08 09:50
"I had a lot more to write here about moral issues, specifically the "women and children" business, but then I mistakenly closed my browser without saving, and the first draft of that was lost. This is already a lengthy posting anyway, so I think I'll cut it here and save the further commentary for a later date."

Christ, it's like your life writes its own punchlines. Maybe Ansuz should incorporate a feature whereby it doesn't let you terminate your browser instance without first confirming that you've saved your changes? Owen - 2010-06-08 14:33
But thanks for taking the time to post more frequent content! Owen - 2010-06-08 14:34

(optional field)
(optional field)
Answer "bonobo" here to fight spam. ここに「bonobo」を答えてください。SPAMを退治しましょう!
I reserve the right to delete or edit comments in any way and for any reason.