Hacker News Reader

Comment by jandrese | Hacker News Reader

It's an often repeated fact that steel swords and the like were quite valuable during the middle ages, but it is a revelation that the reason they are so expensive is that they required hundreds or thousands of man hours chopping trees, managing slow burning fires, carting ore and wood around, etc... Modern media always just focuses on the last step (and usually gets it completely wrong to boot) while ignoring the small army of peasants needed to produce the raw materials for that smith to work into the sword/axe/helmet/etc...

View on HN

Sure, but in the same way, what about all the little things that go into making a pencil today? Usually, as long as there is a market, you can break down all those hours of gathering materials into the concept of "buy X at market price". It only starts looking unbelievably complex when you take control of every single stage and step, and are thus doing the equivalent of running multiple enterprises.

That was probably more common in a feudal society, but not a given. I'm sure a lot of producers just bought the materials they needed.

> It only starts looking unbelievably complex when you take control of every single stage and step, and are this doing the equivalent of running multiple enterprises.

Yup, and that's a good part of the problem with central planning and autarky. There is just too much going on for any one decision-maker to effectively handle.

There is a lesson there for large companies, I think. OTOH, go too far with decentralisation and you have Google's 8 or 9 different messaging systems …

That’s not really the issue with planning. Tracking the inputs to a process is fairly simple, it’s the output that’s really the issue. Diminishing returns makes it really difficult to how when to stop. Large companies often waste increasing amounts of money on internal processes because there is no mechanism to moderate the behavior.

Markets can be summed up as discovering the value of the Nth of each and every little thing. And strangely enough you can use them with central planning. NYC could set the number of Taxi medallions based on their market price.

> Usually, as long as there is a market, you can break down all those hours of gathering materials into the concept of "buy X at market price". It only starts looking unbelievably complex when you take control of every single stage and step

But that's not what we're talking about. A medieval sword is expensive because making one requires thousands of hours of labor. Making a modern pencil has a lot of steps, but requires probably less than ten seconds of labor. Pencils are cheap. The number of steps is not relevant to anything. The complexity isn't hugely relevant either, though more so than the number of steps.

Chopping the wood, mining the graphite, drilling for the oil used to make the synthetic rubber for the eraser, you think the time and effort for all that amounts to ten seconds?

The only reason I can come up with that you would think that is that you discount all these products that are easily delivered by the market as having no time and effort that goes into them. The part of the comment I was responding to is "required hundreds or thousands of man hours chopping trees, managing slow burning fires, carting ore and wood around". That's clearly referring to the whole production lifecycle of all the components, as otherwise you can just reduce that to the fixed cost of what you can pay the market to provide it for you.

I didn't use a pencil as an example as a whim, but because this has been studied before[1], economics essays have been written on it[2], and people have actually gone through the steps to make it from scratch which has been covered here[3] (unfortunately the blog that was submitted seems to be gone), and let me tell you, the time to make a single pencil is not ten seconds if you could up all the time going into making the components that are used to create it.

1: https://freakonomics.com/podcast/i-pencil/

2: https://fee.org/resources/i-pencil-audio-pdf-and-html/ (1958)

3: https://news.ycombinator.com/item?id=10631115

He is referring to marginal time for both cases.

How much total labor time do you think goes into a single pencil? And if it's a non-trivial amount, then how in the world are pencils remotely as cheap as they actually are?

Yes all those steps take work but you get a truly ridiculous amount of ore out of one person labor hour, enough for God knows how many thousands of pencils.

I think you're confusing total labor volume involved in an entire supply chain that makes billions of pencils each year, and then what the person is responding to which is that you divide the total amount of labor spent by the total number of swords produced and it still ends up being hundreds to thousands of hours of labor per sword. This isn't remotely true of pencils using modern automated factory production.

And for what it's worth, it seems like you can get bulk pencils for 6 cents on Amazon. It's probably half that by the pallet from a wholesaler. 3 cents means that the labor cost embedded in a single pencil is less than 1 cent. These production lines are entirely automated.

also we need to account for the labour and cost that's gone into the production automation equipment .

Labor and cost are exactly the same thing. If something is expensive it means it took lots of people's time to work on it. If you produce a pencil for 5 cents then you may spend 3 cents on materials and 2 cents on manufacturing equipment. The cost of the materials represent the wages of the people who extracted the graphite and harvested the wood. The cost of the manufacturing equipment represents the wages of the people who built it.

Labor and cost are almost the same thing: cost is labor costs + capital costs (how much the woodland owner and the factory shareholders are making). Usually the labor cost represents most of the cost, but not always (for regular oil, it's not the case). You could also add taxes and other government-related costs, but since most governments run on a deficit, it would account as a negative cost overall.

That's already accounted for in the price of a pencil. They're not being sold for a loss.

If you went through the process yourself by hand each pencil is hilariously time consuming. But with giant machines mining the graphite, logging machines cutting down trees, trucks transporting both, going to a mill and then a factory to combine them, the amount of labor per pencil is minuscule because it gets spread out across millions of them. In medieval times you don't get the benefits of automation and mechanization, it's grunt work all the way down.

> But with giant machines mining the graphite, logging machines cutting down trees, trucks transporting both, going to a mill and then a factory to combine them,

Okay, so how much time does it take to make a steel sword with modern techniques and machines to do all the work mentioned? If we're going to compare, we have to compare while matching as many variables as possible. At a time when almost all processing was by a person doing it, possibly with the help of a domesticated animal, how much time does it take for people to but down trees, mine graphite (or lead), harvest rubber or process petroleum into synthetic rubber, and transport all those materials to a location so they can make them into pencils.

> the amount of labor per pencil is minuscule because it gets spread out across millions of them.

And the amount of labor per steel sword is also reduced because you're making tens or hundreds of them.

A pencil still isn't ten seconds by this metric, but that doesn't matter, as the point wasn't that a pencil is as hard or harder than a steel sword to make, but that if you want to look at all those inputs going into making a sword and count the time of people cutting down trees as part of it, then it makes sense to look at that part of it for modern processes as well.

The only reason a pencil as we recognize it today (that is, not a twig with graphits, but the six sided with rubber item you imagine as the ideal of a U.S. style pencil) doesn't cost $5 or $10 or even more is because we've found ways to vastly reduce the costs of harvesting and processing those inputs.

The point of my original comment wasn't to say the person I was responding to was wrong (they were entirely correct about the reason for the cost is time and effort), but to expand on their comment and note that the complexity can also be abstracted away the same as it was back then. A sword is cheaper today than it was then for the same reason a pencil is cheaper today than it would have been then, the input components are vastly cheaper because we've found ways to harvest and process them more efficiently.

> Okay, so how much time does it take to make a steel sword with modern techniques and machines to do all the work mentioned?

A random Cold Steel longsword was about five hundred bucks, after a quick search.

Figuring 100% markup and the rest labor averaging $15/hr, about 17 hours of labor.

That's an extremely lazy calculation, but it has to be in the right ballpark.

It's also worth pointing out that swords haven't remotely benefited from large economies of scale in the same way that other modern manufactured goods have simply because swords are a niche good in the modern day and aren't manufactured in the millions annually like other things are. If you want to compare like-for-like, look at goods that are as important in the modern day as swords were in medieval times. One similar enough modern example (because it's also worked metal) would be silverware; you can easily get stainless forks, knives, and spoons in bulk for less than a buck nowadays. Metal silverware in the medieval era was so expensive and thus such a luxury good that only the elites had it. Common people ate with their hands or roughly whittled wooden cutlery. There's also the problem that stainless steel wasn't invented yet, and other metal substitutes for it in cutlery applications are either much more expensive (silver/gold), or rust.

Definitely, contemporary swords are boutique luxury items.

A good machete is about 30 bucks, and a sword is fancier, but not much: $60 is a decent guess at the price of a quality sword, if they were an important sidearm that most people needed to have.

Yeah, I was thinking of machetes too. There are of course trash novelty swords at a similar price range, but those aren't so great at the actual job of cutting things. Machetes are.

Also, $30 gets you a good machete (as distinct from fancy/hand-made). $5 gets you a cheap Harbor Freight machete that nevertheless does seem like it does the job.

If we manufactured swords in the same volume we do chef's knifes, machetes, hammers, screwdrivers, prybars, etc., they'd be super cheap.

I meant that to be a but more rhetorical, but it does express my point. In the past it was far more expensive, and other than the assumption of how much of the cost is the final stage labor and how much is materials cost (and to some degree labor cost, which also has changed over time), the same calculations can be used roughly to approximate relative cost of items in other time periods.

In the past, a vastly higher proportion of the cost would have been base materials, but also they might take some of that in-house, if there aren't industries to support manufacture of materials to a form better suited to the final stage of manufacture.

What if the cost of the steel going into a modern sword was something like $10k-$20k because it was so costly to harvest and process because of labor? That then makes you wonder why steel is so expensive, and now you're using economics as a window to peer into into how ancient economies worked. :)

> Chopping the wood, mining the graphite, drilling for the oil used to make the synthetic rubber for the eraser, you think the time and effort for all that amounts to ten seconds?

Looking at a single pencil, 10 seconds is well in excess of what is actually spent.

Yes, if you are only making a single pencil, then it would take much, much more, but the labor savings from mass production is precisely what makes the modern society possible, and this is precisely how the situation today is very different from how things worked in the past. No-one has spent more than a few seconds in labor in that single pencil of yours, simply because they spent their thousands of hours of labor on producing very large quantities of varying commodities, vanishingly small portion of which were spent on making that single pencil. And this is exactly the key difference, because the woodcutters and colliers who did all those thousands of hours of labor to produce that charcoal did very emphatically not produce a large quantities of product, only some of which was actually spent on smelting that iron. All of the products of their labor was consumed for just a relatively small amount of iron.

Note that no-one is talking about complexity here. Modern steel manufacture is definitely more complex than it was in the early iron age. We are talking about marginal cost. And that is where all the difference is.

> Yes, if you are only making a single pencil, then it would take much, much more, but the labor savings from mass production is precisely what makes the modern society possible

Yes. And what I'm saying is that you can look at the base cost of inputs back then compared to now. A pencil today is cheap, but there are a lot of items that go into them, and all of those have seen a drastic drop in price as we've developed new techniques for harvesting those components.

A pencil created in the same time period would have a vastly increased price too (obviously not as large as a sword). What I was originally trying to express, and it got kind of lost in my second comment, is that the original comment is right, but you can basically boil down the "hundreds or thousands of man hours chopping trees, managing slow burning fires, carting ore and wood around, etc" to "base component cost was exponentially higher".

That concept was somewhat lost in my second comment, when I was going more into what inputs of a pencil, but didn't continue along with how those inputs have gotten cheaper over time. The cost to make one manually doesn't exactly show the cost of doing it before modern practices, but it's much closer than modern materials harvesting practices, so gives a better idea of what the cost would be back then.

I wasn't trying to contradict with my first comment, but to expand and note there are ways of looking at this that scale along well and help you think along the lines of why things cost what they do at certain times. If you think of a sword as a set of inputs, and then ask yourself what those cost in the 1300's, you'll quickly come to the realization that swords were very expensive because all the things that went into making them were very expensive by the nature of harvesting and/or processing them.

A modern graphite pencil is indeed something of a miracle, and needs some civilization to support its construction.

Interestingly, fabricating other writing and drawing implements requires less-than-iron-age sophistication. Charcoal, referenced in the article and used prior to graphite for many/most artistic drawings, is considerably easier to make than steel swords or graphite pencils, and has been used for drawing since the Paleolithic (and still is). Ink made from a suspension of carbon particles is only slightly less crude (needs an aqueous binder).

If you were an artist thrown into pre- or post-civilization Earth and lacked a stash of graphite pencils (but could make fire), baking some vine sticks in a small stone oven for a few hours would enable you to write or draw... if you could find something to draw on.

Back to the original article -- what I found most interesting was that my intuition about the complexity of making iron (to say nothing of steel) was way off. Some things, like charcoal, ink, and even oil paint require surprisingly small operations to fabricate[1]. I incorrectly assumed iron was in that category.

[1] Watrous, "The Craft of Old Master Drawings" https://uwpress.wisc.edu/books/1945.htm has a good discussion of how traditional drawing media were created.

I remember an article on HN once about medieval clothes. The reason people only usually owned one or two complete sets was that a single shirt could be the equivalent of thousands of dollars worth of labor in today's money.

Perhaps this: https://www.sleuthsayers.org/2013/06/the-3500-shirt-history-...

And yup, it took around 600 hours of labor to spin the thread and weave the cloth needed for a single shirt.

Doesn't history generally ignore the contributions of legions of lower class to any of man's great achievements?

The kind of history you see on television and in general textbooks usually does. The omission has been recognized for a long time though, hence "social histories".

The goal of this whole series (and the last one, about bread) is to in some part rectify that demonstrating and explaining the economics of the process from a more labor-oriented perspective.

To expand on others’ comments, popular history may still do this, but professional academic history got over that a long time ago. The subfield that deals with the experiences of ordinary people and “bottom-to-top” complete pictures of a society is called Social History. Social History is now over 60 years old (one of the most brilliant and influential books in this tradition, E.P. Thompson’s Making of the English Working Class, is from 1963), and Social History has shaped the perspective of pretty much every professional historian writing in the last half century, whether they were for or against the key ideas. The intellectual debt is to (non-doctrinaire) Marxism, later supplemented with a health dose of anthropology (which led to the related field of cultural history). Basically the Social Historians more or less won, although the naive “great men great events” approach remained strong in popular history outside the academy.

Social History was and is enormously influential among professional historians, probably the most influential intellectual shift in the history business in the last 100 years. (All people, everywhere, make history! A powerful message!) The author of this blog is a great practitioner of the social history tradition but it’s not a novel approach per se. That said, there should be more social history blogs! This blog regularly hits the front page of hacker news; people love this stuff!

I am loving the series and yeah this does kind of sum it up. Based on the numbers given in this installment it seems like it would take a couple of months for one person to cut enough wood, and collect enough ore, to create a sword. (assuming they had access to the natural resources, and they had the physicality to do the labor, and the skill in smithing to craft the sword. Big assumption that)

Using comparative economics, you could ask what would you have to sell something for to cover all of your expenses for two months? (food, housing, everything) Noting that if you're doing that much manual labor you are probably eating like an athlete in training (i.e. a LOT).

Pretty sobering.

Even beer required a large amount of energy to make, usually from trees or coal.

Reply on news.ycombinator.com