You first do the mirror test with many animals that do primarily use vision. Then you rank those using a different test - if you pick your alternate test properly it can act as a proxy to the mirror test.
Then you test your target animal to see how it does on the proxy test.
An elephant passes the mirror test. An elephant can be trained to brush it's own teeth. Over time, a well-trained elephant develops an understanding that the habit of dental care is beneficial and rewarding, and teaches other elephants to do the same.
Thus, we might conclude that all tooth brushing elephants are sentient, self-aware beings.
This would mean that some of the possible pre-requisites for intelligent life are now either eyes OR teeth...
Now, to extend beyond elephants, we'd need to find another behavior, trait or defining aspect that elephants can share with another species that humans cannot. Thus if elephants must be the bridge-species to self-aware-intellect without natural eyes, they must carry three intelligence defining attributes.
1. Eyes, to relate to humans, enabling primary confirmation
2. Teeth, as a shared attribute, for secondary confirmation, especially in elephants stricken blind by randomized misfortune
3. Other: something that does not relate to humans, but can confirm intellect
This means that elephants can see themselves in the mirror, so humans can know that elephants are fellows in individualized personal experience.
A sighted elephant might manage to teach a blind elephant to brush it's own teeth, thus, elephants that fail the mirror test can still be confirmed as intelligent, if they can carry the capacity to brush their own teeth.
Elephants can be encouraged to brush their own teeth, and hypothetically devise opportunities to teach other creatures to do the same, if they recognize that the other creatures have teeth, but no eyes. Emphasis on the HYPOTHETICALLY, since this is a thought experiment.
Elephants, furthermore, can reach beyond the normal capacity for humans to recognize intelligence, in that they can operate their trunks in a manner that confirms intelligence in other creatures with trunk-like appendages, such as ant-eaters, snakes, octopuses, squids, worms and even certain, very long slugs.
Now, given that elephants act as the bridge species to these other creatures, could we extend beyond elephants, and use elephants to confirm intellect in these other creatures that can confirm intellect in creatures humans and elephants, otherwise, cannot?
Does this hold hypothetical water? Possibly, but man, does it ever feel like a long shot. It's like the Monty Python test for a witch.
How about insects? For example, what if we start with bees as a likely candidate? How do we confirm self-awareness in bees? Could we extend this logic from bees, and beyond, over to another type of insect, through a recognition of mutual traits shared with bees?
We are testing theory of mind after all - so perhaps choose a test that speaks to that instead of teeth?
Would we be capable of defining a reliable test for a spontaneously self-aware wisp of cigarette smoke? How intelligent would it need to be, before we notice it? Would it need to start organizing it's shape in front of a mirror?
What about a not-so-spontaneously-arranged swarm of self-aware nanites, which, from certain distance, for all intents and purposes, might bear a striking resemblance to a dust mote?
[1] http://en.wikipedia.org/wiki/Mirror_test#Animals_with_a_non-...
It's a gimmick they used in the early 50's, adding associated smells to movies. I'm not sure about on-the-fly smell creation though.
1: Determine if the subject is capable of detecting the modification 2: Determine if the subject has detected the modification 3: Have the subject acknowledge that they are the subject of the modification by attempting to correct it.
I guess hearing a crow clear his throat would be an indication of sentience.
Take a hypothetical example, say we were presented with a non-visual smell mirror, would we pay it any attention, let alone recognize or identify ourselves in it ? Highly unlikely, on the other hand an animal like scent-hound might find such a thing interesting.