Comment by jasode | Hacker News Reader

Comment by jasode | Hacker News Reader

When looking at human-generated categories or labels, it's useful to think about the meta layer of "lumpers vs splitters": https://en.wikipedia.org/wiki/Lumpers_and_splitters

This thread's article divides drama situations into 37 buckets. And another person divides stories into 7 types: https://en.wikipedia.org/wiki/The_Seven_Basic_Plots

In music... a Lumper might think of only 2 types of music -- instrumentals vs songs with vocals. A Splitter says there are 1000s of genres because even within sub-category of rap music, East coast Miami rap is totally different and 180 degrees opposite from West coast Los Angeles rap.

Lumping vs Splitting type of thinking happens in programming topics too. E.g. A lumper has no problem saying, "C/C++ are low-level languages with manual memory management" ; but a splitter will object with, "I hate it when C and C++ are lumped together! They are TOTALLY different languages!"

Depending on what you focus on, everybody's customized taxonomy is "correct".

Lumpers vs Splitters is a fun categorization. When I see a group of things categorized I view it as a set partition and I wonder what equivalence relation is being used to distinguish the partitions (categories). The relation is usually unstated so it becomes a mental exercise to steelman the categorization by trying to think of one.

“Lumpers vs Splitters” is itself a splitting or a lumping?

Lumping, for sure. There's people in the middle who recognize the difference between rap and jazz, but not the different variants within.

This is a special case of the observation that no discretization of a continuous space is "correct". They are inherently subjective, and some may be more useful for any given purpose than others.

Reminds me of a common review trope: "When it comes to X, people either love it or hate it, I'm somewhere in the middle"

to mangle Jim Barksdale: there are only 2 ways to create insights - bundling ideas or unbundling them.

Those are seven plots. The book mentioned in the article lists dramatic situations. A story would in all likelihood involve more than one dramatic situation. The article (frankly, quite blogspammy-looking) makes the same mistake right in its title.

A quick search reveals a better resource with extra information about each situation: http://juliahwest.com/prompts/37_dramatic_situations.html

>A story would in all likelihood involve more than one dramatic situation. The article [...] makes the same mistake right in its title.

I'm not making the same mistake and I wasn't equating "situations==stories". Yes, there exists a many-to-1 relationship of situations nested inside of a story ... but that is not relevant to my point.

I'm emphasizing the meta layer of creating <n_count_of_buckets> for <any_topic> and how it's arbitrary and subjective instead of being definitive.

So your link with extra information of the 37 types is diving deeper in the wrong direction for my purposes. Instead of reading more about 37 situations, we ask where does the number _37_ itself come from? Why not 34 situations? Why not 42?!? Because it's arbitrary and it depends on the human dividing up the mental buckets.

E.g. the first paragraph[1] in the article of this thread sort of hints at this "meta" layer arbitrary taxonomy. A bullet summary of differences:

- 2 types of stories -- Leo Tolstoy

- 3 basic plots -- William Foster-Harris

- 20 -- Ronald Tobias

- 36 -- George Polti

[1] excerpt: "Great literature is one of two stories," we often quote Leo Tolstoy as saying: "a man goes on a journey or a stranger comes to town." That's all well and good for the author of War and Peace, but what about the thousands of screenwriters struggling to come up with the next hit movie, the next hit television series, the next hit platform-specific web and/or mobile series? Some, of course, have found in that aphorism a fruitful starting point, but others opt for different premises that number the basic plots at three (William Foster-Harris), six (researchers at the University of Vermont’s Computational Story Lab), twenty (Ronald Tobias), 36 (George Polti) — or, as some struggling screenwriters of a century ago read, 37.

You could have supported your point better by offering an alternative taxonomy of possible dramatic situations. Maybe you could invent one?

Either way, the author of the website I have linked concisely sums it up as:

> Georges Polti says that all stories boil down to just 36 dramatic situations and takeoffs of those situations. Somebody else out there added #37. If you're stuck for a situation, try this.

That’s it. The number is arbitrary to a degree. It’s a tool for writing.

It is obvious how situations listed can be segregated more coarsely (just “rivalry” instead of one for equals and non-equals) or finely (see subplots on that resource). Being less specific (fewer categories) ensures relevance over longer period of time and/but leaves more to writer’s interpretation. Looking at successful stories from another angle, one can likely identify a different set of dramatic situations.

(edited)

"There is two kinds of music, the good, and the bad. I play the good kind." — Louis Armstrong

"Oh, we got both kinds. We got country and western." — The Blues Brothers.

IME this is situational and when people make a point of increasing the degree of distinction like this it's either because the particular situation demands more nuance, or they're doing it as a competetive excercise to show they can see more categories. The level of granularity should be appropriate for the situation (and it's more important that whatever categories you end up choosing are actually correct) and I think most people understand that.

In general distinctions between people like this smell off to me, especially when they present one category as better and allow the reader to pick.

Reply on news.ycombinator.com