Something fun I've been learning. While studying why we should care how we define a "planet", I learned there are 3, not 2, forms of logic we commonly use, both in the scientific method and in our daily lives. Guess which one we use the most? (source: arc.aiaa.org/doi/10.2514/6.…) 
2/ The 3 forms are (1) Deduction, (2) Induction, and (3) Abduction. Deduction is the one they teach in introductory logic class. It goes like this: "All Jedi can use the force. She is a Jedi. Therefore she can use the force." That's deduction. (Source: slideplayer.com/slide/10746918/)
3/ Induction is often wrong, but we use it a lot. It is when we observe just a sample of information and draw general conclusions. "Every Jedi I saw was using the force. Therefore, all Jedi can use the force." (source: danielmiessler.com/blog/the-diffe…)
4/ The mysterious one is Abduction. It's how we create theories & hypotheses to explain things. Abduction is arguably the main type of inference we use, both in daily lives & in science, yet we still haven't explained the logic behind it or how it works. (collegepublications.co.uk/downloads/ifco…)
5/ The previous chart has a simple description of abduction, but in reality it is far more complicated because there are MANY possible explanations for everything we observe. Maybe the beans came from a different bag of white beans. Maybe they were red but someone bleached them.
6/ Part of the problem in abduction is that nobody hands us a pre-made list of all possible explanations to choose from. Abduction is (1) the process that invents the list of possible explanations, AND (2) the process that narrows the list to only the reasonable explanations.
7/ People who study artificial intelligence have made a lot of progress programming the logic of abduction into computer algorithms, but running the computer program might take longer than the age of the universe to sort through ALL the possible explanations it might invent.
8/ Philosophers & scientists who study abduction repeatedly remark how amazing it is that science has made any progress at all! How do human brains make correct abductions (invent explanations) so quickly??!! (I know, it seems easy to YOU, but that's because you are a human.)
9/ Somewhere in the amazing tangle of neural connections, there is a logic that sorts through all the dead end explanations, the lower-importance data, and the seeming contradictions, creating that flash of insight: AHA! I know what causes this! (news.softpedia.com/news/Synapse-N…)
10/ Sherlock was fond of saying he used deduction to solve cases, but what he was really using was abduction. Deduction has far less application in the complexities of real life where we are trapped inside our own limited perspectives. We need to be good at abduction!
11/ Philosophers and cognitive scientists are working hard to unravel the logic of abduction, both so we can understand and improve our own thinking, and so we can program it into artificial intelligences. (Then the robots will put us scientists out of a job 🤖😬)
12/ The earlier philosophers of abduction called it "just a guess", or even a miracle. So far we haven't completely figured it out. One idea is that our brains perform "convolutions" of ideas, mixing ideas with ideas to create the new ideas that eventually solve a problem. But...
13/ If you have N ideas, you can mix them together into N! ("N factorial") different ways. So if you convolve just 20 ideas together, there are over 2,430,000,000,000,000,000 ways to convolve them. So if you think of 1 theory per second it takes 56 times the age of the universe.
14/ The all-important thing that makes science (& daily life) possible is the strategy our brain uses to cut this giant space of possibilities down to manageable size. (Sherlock pretends it is just "elementary", but it's really not.) (source: theguardian.com/books/2014/sep…)
15/ Part of how we do it is by the careful definition of concepts. We can't afford to clutter our minds with non-useful categories, creating more junk our brains need to sort through. (Increasing N from 20 to 21 makes it go from 65 to 1,365 times times the age of the universe.)
16/ Scientific logic is the same as everyday logic except scientists try to use more rigor. So for example, we try to define categories to be **reductively** useful, the type of categories that will enter directly into reductionist theories instead of just adding to the clutter.
17/ This is one reason the IAU's 2006 decision to vote for a non-reductionist but culturally acceptable definition of "planet" (designed to keep the number of "planets" small for the sake of culture) was a huge mistake. (Source: Antonio Ciccolella, en.wikipedia.org/wiki/List_of_n…)
18/ They mixed both roundness and orbital state into the definition, but that mix is non-reductionist. The combination is impossible to fit into any reductionist theories of nature. (It also violates historical use, which had evolved thru 100s of years of abduction to be useful.)
19/19 What I'm finding is there has not been enough research trying to understand the role of categorization in the effectiveness of abduction (and in the scientific method). There's a lot of work for philosophers of science still to do! My manuscript is in work. More to come...
