You know you're awake because reality has continuity and makes sense.

Conversely, dreams are fragmented and implausible. You can train yourself to check this to lucid dream (become aware that you are dreaming).

I've had vivid dreams where I was completely convinced I was in reality. Continuity was maintained well enough to fool me. How could I know I wasn't in a dream? How can I truly know beyond doubt that I am not in a dream?

Yes you can sometimes do that (I've lucid dreamed quite a few times before and it's pretty wild), but it's nowhere near 100% reliable. you can get pretty good with practice, but even then it's not a scientific test. It's more of a process of reasoning and gut checking. The human mind can fill in an incredible amount of detail when needed. In one lucid dream I was playing a song on the guitar that I had (as far as I know) never heard before (myself in the dream was inventing the song). I realized I was dreaming before waking up and practiced the song over and over so it was stuck in my head, and then immediately after waking I wrote down the tablature. It ended up being a pretty good song, and (at least so far) I haven't heard anything else like it.

Obviously this is heavy anecdata, but I've never heard a scientific explanation of how to be certain whether or not you're dreaming (or as Descartes describes, being deceived by a demon).

>it's not a scientific test.

Isn't every subjective experience, almost by definition, unable to be tested scientifically?

We can objectively measure someone's brain activity during sleep, but we can't really measure their subjective experience. Implying that everything made up of subjective experience (like consciousness) can't be measured. Which also implies it can't be proven by scientific means. So in response to the GP comment, I'm not sure we can "know you're awake" (in the scientific sense, at least).

Yes, exactly, that's the point I'm getting at. We experience "life" through our senses, and our senses can be fooled. Even measuring brain activity requires a subjective level of "trust" that we're seeing a real machine that's taking real measurements, rather than a simulation of a machine. We can look for glitches in the matrix, but absence of evidence isn't evidence of absence.

It's wildly impractical to live your life like that, but at the end of the day we just don't know and probably can't know.

I'm pretty sure what you're describing is the hard problem of consciousness.

There are tried and true techniques, like checking your watch. When you're dreaming, it will show different times when you check it.

So if you train yourself to check your watch for consistency (in your normal everyday life), your unconscious will continue the practice in your dreams, and you'll notice the inconsistency.

Sorry if it came across that way, it definitely wasn't the intent! I think Sam Harris has some pretty fascinating and accessible discussions on the topic. There's also some interesting neuroscience work that seems to conflict with the common notions of free-will.

oh yeah, totally agree Sam Harris on this is really awesome. I also love Robert Sapolsky's stuff. He's got several great videos on Youtube and his books are awesome as well.