(I am not trying to redefine the word, but to retroactively define a fuzzy term that is used often in the literature)
arthurNatureTechnologyWhat2009 shapes its ideas of technology and technological paradigms around exploiting specific phenomena. This framing addresses a lot of the vagueness associated with paradigms that then cascades into any discussion of Technological S curves, Technology readiness level - TRL, the §Tech Tree Model of Heuretics, and basically any vaguely abstract discussion about science and technology.
So what is a phenomenon in the context of science and technology? The dictionary definition is that a phenomenon is ‘an observable fact of event.’ Some dictionaries also associate it with a sense of mystery. There are many phenomena with well-understood mechanisms, but it’s important to consider that connotation. That’s a good input, but doesn’t feel very useful. Thanks to arthurNatureTechnologyWhat2009 we have a ‘boundary condition’ that however we’re thinking about phenomena, they need to be ‘exploitable’ into technology. An important feature of something observable and exploitable is that they are both external to you. In order to ‘exploit’ something, it needs to have something you don’t. It wouldn’t make sense to exploit myself typing these words. That sentence doesn’t even parse.
Like all good definitions, this one is fractal. Every time you zoom in on a fractal, it presents the same pattern on a different scale. Electricity is a phenomenon (things can get charge that flows! Whoah!) but you can also think of the broader phenomenon of electricity as made up of a whole slew of individual phenomena - the Hall effect, resistance converting electricity to heat, etc.
According to wikipedia Emmanual Kant was the first philosopher to think seriously about this. I have not yet read that so I defer to him until I do.