The difference here is that Dennett is arguing that the thermostat is thinking on account of it's information processing being valued (by human language in particular) as thinking.
...
Second, it doesn't actually explain at all how labeling a thing as conscious causes it to actually be conscious (whether that's between a human being and a thermostat or two human beings).
Let me offer a perspective.
There is such a thing as an Irish Setter. I can point to a Labrador Retriever and say, "that is an Irish Setter", but in doing so, I'm not causing that dog to actually be an Irish Setter.
However, according to wikipedia, a collection of people calling themselves the Irish Red Setter Club, in Dublin, Ireland, set a breed standard for Irish Setters on 29 March, 1886. This group of people chose to call particular kinds of dogs Irish Setters, and in doing so, they made particular dogs "really be" Irish Setters.
The difference between my retriever and the Irish Red Setter Club's setters is the game being played. I'm pointing to a particular dog, and to a label. When I claim this dog is a setter, what I'm trying to do is to claim that the dog qualifies for the title. The requirements for the title were not determined by me, but they are sufficiently clear that we can judge whether or not my retriever meets the breed standard.
When the Irish Red Setter Club was calling dogs Irish Setters, they were not trying to change what dogs were; nor were they trying to change what particular dogs were. There was even a pre-existing vague notion of Irish Setters, and there may have been particular dogs everyone agreed were Irish Setters; there just wasn't a clear cut standard. There wasn't a way to determine which dogs "really were" Irish Setters, not because we couldn't figure out if a given dog "actually was" an Irish Setter, but rather because nobody had established the criteria precisely.
So I'm claiming that it's fair to call the thermostat conscious, and you get to make it conscious, if you're setting the breed standard for consciousness. A thermostat can only possibly not "really" be conscious in two ways:
- You're trying to establish a little bit more of a breed standard (in which case there is no "really conscious" a priori, but you're just establishing that you don't want to talk about thermostats)
- There already is a breed standard (in which case, you get to refer to it when you reject thermostats)
Does that make sense? Beyond this I won't say much, because I'm not quite sure if you're considering yourself as
having the breed standard to reject the thermostat's conscious status. I simply want to raise the point that it only makes sense to ask the question if you
have or are
establishing the breed standard.
The other thing I would mention is that it helps, when establishing a breed standard, for you to study the
things you wish to include in the breed, in order to figure out what
traits you wish to include in the standard.
Now, the problem here is that, applied to non-conscious information processing systems like computers there is absolutely no reason why a computer couldn't apply a shorter run code for a series of functions non-consciously.... (e.g. routine 1, routine 2, routine 3, etc,).
This assumes a breed-standard for consciousness that excludes thermostats.
So, Dennett's theory has two glaring problems. First, it always requires a conscious observer in the first place
"Conscious observer" is more specific than "consciousness". The former implies an agency.
In short, his theory doesn't really explain consciousness at all.
That much I agree with, if you're referring specifically to "Consciousness Explained".
The theory could very well be put forward that, since everything is thought, everything from simple information processing systems (like a thermostat) all the way up to complex human brains are "conscious". Subsequently, a believer in this model might say that when our bodies cease to function, the thought of who we are returns to God and we as individuals (thermostats or computers) cease to exist as distinct individual entities.
True, but a believer in this model doesn't necessarily have to believe that there's a God. It might be that fitting the "breed standard" for thought isn't considered by that person as meeting the "breed standard" for a God. After all, there are a lot of people who profess quite loudly and strongly that they believe in God, and they have particular attributes one might want to include in the "breed standard", such as the ones you're mentioning in this reply:
Personally, I find this too fatalistic. Why bother to create other consciousnesess just to have them disintegrate in the end?
"Why bother to create" implies cohesive agency and purpose. A consciousness with cohesive agency and purpose is a lot more specific than a consciousness per se. But this is, indeed, how many people who do believe in God conceive of him--basically, he's a clone of all of the complicated human social aspects of mind, except that he is purified. And there are a lot of such complicated aspects.
Some might even go so far as to say our individual counsciousnesses are just small pieces of God's mind that eventually return to the source.
Yes, some might say that. But if you're looking at things like thermostats and their behaviors, and you look at our brains in terms of producing conscious behavior, then it would be reasonable to model consciousness as composed of such "thermostats"--that is, functioning parts. In this aspect, once we die, our brains stop functioning--though the mass is still there, there are no more "working thermostats", so there's no return to at least the working thermostat layer.
If you look at the mass itself as conscious pieces, it doesn't quite fit this analysis either, because we are not actually composed of particular matter per se anyway; due to our metabolic nature, matter seems to have to
flow through us to maintain our energy, which is required to survive. In this aspect, if there's a "return to God", it's being done constantly (and some forms of return have some not quite tasteful names).
Either way, as you can see, this is rather different from Dennett's model.
Dennett's not here to defend himself.
I, personally, don't think it's possible to know which theory is correct.
I'd like to offer that it might be worse than our not having an answer--that perhaps we need to work on the question. We need a breed standard before we can classify anything in outside groups with meaning. I don't take it for granted that breed standards exist completely independently of our determination of them, though I welcome a discussion of traits you may wish to add to the standard.