Were these arguments in reference to Dwarf Fortress? If they were, maybe those threads are relevant and worth linking.
Sadly, I doubt it, it was on IRC, and the argument was over whether or not intuitive knowledge even existed, and ended with the other side essentially just not believing me and refusing to listen to anything else I said.
What tasks? And how do you think the knowledge was gained in the first place?
The tasks that most require academic knowledge are those that relate to utterly non-intuitive actions, such as basically anything related to chemistry, as the tanning example above points out.
You can read that bit on how it was done
here, and it's far more disgusting than you think. There's a reason tanners were "untouchables" in many Asian cultures.
You can almost say the same about things like brewing, but at the same time, people really knew nothing about brewing, and drinking alcohol was a horrifyingly dangerous practice. (The Irish concept of a "Wake" was created because of the alcohol-making process. Before burying someone, you left them on a table for a couple days, just to make sure he was really dead, and not just unconscious from some sort of poison in the booze caused by letting the wrong kinds of microorganisms into the fermentation process. Oh, and did I mention that they used lead as a sweetener for centuries?)
So, yes, you can become a doctor the Dr. Zoidburg way, by just cutting people open and being surpised by the way that you didn't expect
that organ to be over
there, but it's probably best they understand some of the principles of medicine, first. (Of course, real medieval medicine involved asking the patient what his zodiac sign was, and then doing a horoscope to figure out which humor was out of line... Basically, Zoidburg would be an improvement.)
People did eventually learn how chemical reactions worked, and really had science took off when they managed to put together enough experiments to create an accurate enough model to generate the Scientific Method for really understanding the world in academic terms, but that's outside the scope of the game. Mostly, chemistry was discovered in ancient times through accidents and sudden sparks of insight. Cheese was just milk that happened to get the right kind of culture in it by mistake, and someone had the bright idea to try it again and again until he figured out tricks to replicate it reliably. Yeast was much the same. It's slow, painful, and probably involves killing a few people, but you can do it.
Keep in mind, though, that trying to learn herbalism by eating every mushroom you find to see if they are poisonous or not is likely to get you killed before you master the knowledge.
Try to separate in your minds what parts of knowledge are academic and what parts are intuitive, though. "The
fly amantia mushroom is poisonous/hallucinogenic" and "Fly amantia is red with white spots on top" are academic knowledge, but remembering what it looks like in terms of being able to mentally visualize it, instead of just remembering a text description, and knowing what sort of plants it tends to grow near all tend to be intuitive forms of knowledge.
Academic knowledge, the kind you just store as data or learn from a book, is usually easy to forget. Intuitive knowledge is something that comes from literally laying down new pathways in your mind to perform certain repeated functions better, and are things that are very difficult to forget.
You can forget the exact component serial numbers of your car very easily, but the skill on how to drive it is intuitive, and to be a good driver, you have to be able to know how much to push down on a gas pedal or turn a wheel without having to stop, hesitate, do a conscoius calculation, figure you need to turn the wheel 10 degrees right to merge into the next lane, and then suddenly remember you need to then turn 10 degrees back to the left once you get there to correct your course. Correcting the steering wheel for the purpose of staying in your lane or how much you turn the wheel to merge lanes, or how much pressure you put on the gas are things you don't need to conscoiusly think about because practice has built neural pathways that let these functions become performed only semi-consciously. Your conscious mind only needs to be there for ensuring that the subconscious is aware of the right stimuli (looking at the other cars around you), and the higher-level functions, like "where am I going, anyway?"
Academic knowledge can guide intuitive knowledge, certainly. Especially in something you could call "philosophical understanding", which was why medieval medicine was so horrifying - it was built around a horribly mistaken Socratic philosophical framework that refused to believe in the operation of the organs, or even trial and error, and instead worked on some hypothetical humors system. In a sense, your philosophy on a subject is the way in which you even approach learning that subject, and it is therefore the most important academic knowledge you can learn.
Again, this is kind of going on a really massive tangent to prove a point, but setting up the... well, actually, it's basically a philosophical viewpoint on how people learn, if I were to introspect a little... is the most important starting point for discussions of how to model the learning process.
If you wanted to discern what skills are things you could never do without some sort of academic knowledge, or at least sitting down and pondering and experimentation, then just look for things that are totally unintuitive. The construction of and use of some tools is unituitive, as are some fabrication techniques. Wooden cabinets are fairly difficult to make intuitively, especially if you start out with nothing but a boulder, and are told to make a carpentry shop and all your tools from it. Anything involving chemical reactions or biology is not necessarily intuitive, either.
Still, I'm sure there's not much better method to learning glass blowing given the technology they had than to just start grabbing sand, and testing what things made good glass.
You might want to take it from the other direction - what things would you just have to sit and watch and learn from seeing or doing, instead of reading it in a book to understand its principles, knowing how little principles were understood at that time?
Really, Dwarf Fortress isn't a difficult game. The difficulty is in learning the interface and how to play. After that, it's often too easy.
Choosing skills for starting dwarves should be assisted with an interface to see what tasks your dwarves can succeed at, and warnings when you try to embark without crucial skills.
Incidentally, also something I've gotten into pretty heated debates over.
Perhaps it would be better described as "front-loaded complexity", then. The game essentially has no real logical progression in complexity, there are just a pile of things you have to do to get everything set up, and nothing really flows from one to the other, they just all have to get done, and the best you can do is prioritize them.
It's only "hard" in the sense that the game comes with no tutorial to make the absolute flood of information and options the game sends the player discernable, but it's "easy" as soon as you understand what you are doing.
Still, unless our starting 7 have every bit of knowledge to start all the industries they'll need, then what you're doing is chucking one more thing in the pile of things that dwarves need to do the instant they start a fortress if they need to not only get picks, start mining, find ores, find fuel, set up smelter, set up furnace, set up smithy,
learn how to make every piece of equipment by reading books, and then start moving raw materials through the assembly line to start getting the first pieces of metal whatever produced. When you're trying to get a military started from scratch, that's one more step in a chain that already has multiple components, all of which you have to perform the instant you hit the pavement. Then, once you have it built, you never have to worry about it again, really.
I'm not saying it's insurmountable, but I always try to work towards distributing the complexity of the game. Right now, the game has you build everything immediately, and then basically just lets you sit back and watch your fort run itself after the first two years are over, since your job is basically done. It's what I've been trying to do with my
two major suggestion threads, anyway.