文章探讨了子文化周期是否存在的问题，作者认为，虽然人们常常认为子文化会经历兴衰循环，但实际上，这种循环并非严格意义上的周期性。作者以 disco 和 dance music 的兴衰为例，指出子文化的发展更像是一个不断演化的过程，新的子文化会不断涌现，并取代旧的子文化，而非简单的周期性循环。

🤔 **子文化发展并非严格意义上的周期性循环:** 作者认为，虽然人们常认为子文化会经历兴衰循环，但实际上，这种循环并非严格意义上的周期性。比如，disco 在 70 年代流行，80 年代不再流行，但这并不意味着它会在 90 年代再次流行。子文化的兴衰更像是一个不断演化的过程，新的子文化会不断涌现，并取代旧的子文化。

🧐 **子文化的发展是不断演化，而非简单的周期性循环:** 作者以 disco 和 dance music 的兴衰为例，指出子文化的发展更像是一个不断演化的过程，新的子文化会不断涌现，并取代旧的子文化，而非简单的周期性循环。比如，disco 之后出现了 dance music，dance music 也有自己的兴衰，之后又会产生新的子文化。

🚀 **子文化的发展是一个复杂的过程，受多种因素影响:** 作者认为，子文化的发展受到多种因素的影响，包括社会环境、经济状况、政治形势等等。因此，子文化的发展很难用一个简单的模型来预测。

🐛 **周期性现象存在于自然界，但人类社会中则更复杂:** 作者认为，周期性现象确实存在于自然界，比如蝉的周期性繁殖。但在人类社会中，子文化的发展要复杂得多，受多种因素影响，很难用简单的周期性模型来解释。

🧠 **作者认为，子文化的发展更像是一个不断演化的过程，新的子文化会不断涌现，并取代旧的子文化，而非简单的周期性循环。**

https://astralcodexten.substack.com/p/highlights-from-the-comments-on-subcultures

1: Maximum Limelihood Estimator writes:

I firmly believe that cycles don't exist and never have existed. This is my shitposting way of saying "I have never, once, in my years of experience modeling human behavioral time series, come across an honest-to-god cyclical pattern (excluding time of year/month/week/day effects)." And yet for some reason, every time I show a time series to anyone ever, people swear to god the data looks cyclical.

I called this “a cyclic theory” to acknowledge my debt to Turchin, but you may notice that as written it doesn’t repeat. Just because disco was cool in the 70s and uncool in the 80s doesn’t imply it will be cool in the 90s, uncool in the 00s, and so on forever. It will probably just stay uncool.

The cyclic aspect, if it exists, would involve the constant spawning of new subcultures that rise and fall on their own. So disco begets dance music, dance music has its own golden age and eventual souring, and then it begets something else. The atheist movement begets the feminist movement begets the anti-racist movement begets and so on.

What about the stronger claim - that no (non-calendar-based) cycles exist? I think this is clearly false if you allow cycles like the above - in which case the business cycle is one especially well-established example. But if you mean a cycle that follows a nice sine wave pattern and is pretty predictable, I have trouble thinking of good counterexamples.

Except for cicada population! I think that’s genuinely cyclic! You can argue it ought to count as a calendar-based cycle, but then every cycle that lasted a specific amount of time would be calendar-based and Limelihood’s claim would be true by definition.