To Be Conscious in a Body, Frozen|
May 16, 2009, 11:21a - Consciousness
It's hard to tell if a thing is conscious. You know, if there is something that it feels like to be that thing. I know it feels like something to be a person, and I think it doesn't feel like anything to be a shoe (unless perhaps I've been smoking some salvia), or to be a dead person. But what if I was paralyzed, or in a coma? Is it possible that I would still be alive up there, just unable to communicate with the rest of the world?
Today, I ran across a very interesting paper in the journal Neurology about how some researchers were able to communicate with a person who had "locked-in syndrome". This has been done many times before, but what caught me off-guard was the way that they communicated. In fact, it struck me as particularly ingenious.
Locked-in syndrome is a condition where, through brain damage or neurodegeneration, a patient can no longer communicate with the outside world. The neurons that control their muscles just don't work properly anymore, so they're paralyzed. In some cases the rest of the brain seems just fine, so there is reason to believe that there might still be someone upstairs.
"Classic" locked-in patients lose most, but not all, muscle control. The most famous locked-in person might be the main character of the autobiographical book/movie, "The Diving Bell and the Butterfly". That person composed his entire memoir by communicating via eyeblinks.
The subject of this story is a bit different, though, as she was a "complete" locked-in patient, one who had no muscle control at all. Being in this state begs the question, "How do we know that she's even conscious and not just in a coma?" To answer this question, researchers usually try to show that they can communicate with the person. If they can, and the person can communicate back, it's reasonable to conclude that the person is still conscious, so it probably isn't time to pull the plug.
To start, a bit of background. Previously, it had been shown that some completely locked-in patients could be trained to control EEG activity (this alone is amazing). EEG (short for electroencephalography) is a technique for recording electrical signals coming from the brain, by placing wires on your head at various places. These wires make contact with your skin, and can pick up the net electrical activity coming from chunks of brain. It's a very coarse technique and doesn't really tell us much about how the brain works, but it's handy because it doesn't require any surgery and can be used to correlate electrical activity with behavior. For example, a lot of what we know about the various stages of sleep come from research using EEG.
Researchers had previously trained locked-in patients to communicate by changing their own EEG signal (e.g. more positive voltage to answer "yes" to a question, more negative voltage to answer "no"). But this specific patient wasn't having any of it. So instead they tried getting the patient to communicate "yes" by imagining the taste of a lemon, and communicate "no" by imagining the taste of milk. Apparently, imagining these things changes the amount of saliva in your mouth, which alters the saliva's pH, which can be measured. Amazingly, the patient was able to learn to do this just fine, answering questions with an accuracy of 89%. So they were able to communicate with the patient via the patient changing the pH in their mouth! Pretty cool, I think. Because the patient was able to answer these questions better than chance, I'd say that she was still conscious, even though she couldn't move any other muscle in her body. After all, our "ground truth" when it comes to whether someone is conscious is whether they can verbally report on their mental state. And that is pretty much what this patient was able to do, "reporting" by imagining something that happened to change the quality of her saliva.
In hindsight, it's clear that this patient actually hadn't lost all muscle control. She had lost all "direct" muscle control, but, via her imagination, she could indirectly control the muscles that control how much she salivates. In a way this is very similar to Pavlov's dog, who learns to salivate when it expects to get some food. I wonder if there are other things that she could imagine that would change other biological indicators, or if changing saliva is perhaps the "easiest". Maybe there is something special about cognitive control of food-related behaviors, maybe in some way they are more "basic" and more likely to survive brain trauma...
Check out the original research here:
- Wilhelm, Jordan, Birbaumer 2006 - Communication in locked-in syndrome - Effects of imagery on salivary pH
Also, here is the list of questions that they asked the patient: List of questions (.doc)
If you're interested in more articles about neuroscience and consciousness, check out my Why Study the Brain. Now that I look back on that article, I've realized that I no longer identify that much with significant parts of it. But it does include some interesting science tid-bits.
Read comments (13) - Comment
« Visualizing a Worm's Neural Network
Tugging Bubbles in a Box »
- May 19, 2009, 6:26a
Wow, ingenious is right -- thanks for sharing this.
- May 19, 2009, 9:41p
dear god. this is ingenious, but i am just thinking about this person who is locked in. i think they are likely totally insane. or almost. how could you not be, with no ability to communicate with the world?
when i read the diving bell book, at one point i was reading on the bart. it made me feel so claustrophobic that i almost threw up and had to get off the bart and breathe.
just reading your blog post is making me nauseous.
hope you are well!
- Aug 14, 2009, 3:02p
Is a worm conscious?
- Aug 17, 2009, 7:22a
short answer: i don't know whether a worm (e.g. C. elegans) is conscious.
long answer: hmm, what a damn tricky question.
there are a few issues. first, it seems that it is in principle impossible to know if something other than yourself is truly conscious. if you define consciousness to be the phenomenon of having a subjective experience of the world (which is how i use the word most of the time), it seems plausible that another creature could act as if it were conscious even though it had no subjective experience. it would answer your questions, behave "normally", and even say "yeah, i'm conscious", but there may in fact be nothing that it is like to be that creature; in other words, the creature may act as if it has subjective experience without ever having any - it would fake it. and in principle, it seems impossible to know when something might be faking it.
this applies equally to other humans as to other organisms, such as the worm C. elegans. in general, though, when it comes to humans, I seem to make a sort of similarity argument: i'm conscious, and other humans seem to be a lot like me, so they're probably also conscious. so in practice we make assumptions that seem to be consistent with common sense, but in principle confirmation of consciousness in another creature seems unknowable.
second, let's just forget the first problem and say that yes, there exists a behavior that can only exist if a creature is conscious. so if the creature can act in a certain way, then i conclude that it is conscious. which behavior would I use as the test for consciousness? in practice, being able to respond sensibly to questions seems like a nice test, but as in the article above, verbal response seems sufficient but not necessary. also, it doesn't extend very well beyond humans, as we seem to be the only species that has a highly-expressive language (at the very least, we don't have good ways of communicating with other creatures in their own language, if they have one). intuitively, i think my dog is conscious, yet I communicate with her in extremely simple ways. and i can imagine that even if i couldn't communicate at all with my dog, she might still be conscious. my conclusion that my dog is conscious is based largely on empathy and similarity of response - she responds to things (e.g. hunger, anger, treats) in a way that i also respond to those things.
so now on to worms. i have nearly no empathy with worms, and there lives are so different than mine that it's really hard to tell based on intuition alone whether they're conscious. but you can do things in worms that are easier to do than in any other organism: you can identify genes, molecules, cells, and groups of cells that are required for certain behaviors. so if consciousness is due to some physical process (a big assumption), studying worms might be a good way to find pieces involved in that physical process - assuming, of course, that they are conscious (to some degree) to begin with.
i guess this is a very long-winded way of saying that i don't know whether a worm is conscious. there is a specific type of learning that is correlated with awareness in humans (a variant of Pavlovian classical conditioning), and i'm in the process of testing to see whether the worms can do this type of learning. if they can, it would be one small piece of speculative evidence that worms (specifically C. elegans) might be conscious. my hope is that over time we'll discover more behavioral tests that are correlated with consciousness in humans, and that worms could then also be tested. if i accumulate enough pieces of speculative data, the whole argument might become a lot more convincing. that's one strategy, at least.
- Oct 15, 2009, 12:53p
thanks for the very accurate answer.
let's assume that C. elegans is conscious (which i believe is true). is a bacterium conscious? is the fact that a system possesses a neural net that makes it conscious? or isn't just their reaction to a behavioural test? if the reaction makes it conscious then also a bacterium is conscious, since it replies to external stimuli as any other creature. aren't neural nets only there because of the size of the systems and therefore the necessity of a faster communication between parts than that achievable chemically?
- Oct 18, 2009, 3:26p
i don't think that just because an organism has a neural net it's also conscious. like i said, i believe you need to find a behavior or some other observable that correlates well with awareness or consciousness. the presence of a neural net could be an example of such an observable, but given that there are many examples of nonconscious states in humans who have neural nets (e.g. vegetative states, dream-less sleep), i don't believe a neural net is sufficient. a neural net may be necessary, as there don't exist any examples of conscious creatures without neural nets, but that is less helpful in identifying those creatures which are conscious.
(as an aside: i prefer using the word "organism" or "creature" over the word "system", as i think "system" masks the complexity found in biology that is not found in standard engineered "systems")
i want to be clear - i don't believe that just any behavioral response is sufficient to show that a creature is conscious. people talk in their sleep and sleepwalk all the time, ostensibly without any consciousness at that moment (or at least very low levels). it's actually pretty interesting to record yourself sleeping - i got a nightvision camera that can do this, and i do all sorts of strange things in my sleep. one night i woke up holding a light bulb in my hand, with no memory of how it got there. the light bulb had been lying on my nightstand, so i just put it back, but it was a bit bizarre...
what i'm working with so far is a specific behavior called trace conditioning. i think i've blogged about it before; it's a specific type of pavlovian conditioning that involves specific time delays between the paired stimuli. i may blog about it in more detail at a later point. right now i'm working under the assumption that if an organism can trace condition, than this is sufficient for consciousness. this assumption is based on the fact that trace conditioning is well-correlated with awareness in humans, while other forms of conditioning are not. neither a bacterium nor a paramecium (a larger single-celled creature) are capable of trace conditioning (though no one has ever tried, as far as i know). no one has shown that c. elegans can trace condition. getting trace conditioning working in c. elegans is actually my main research project, and if i can show that they trace condition, then that's a tiny shred of evidence that it may be conscious.
for your last question, size does not necessitate the presence of a neural net / nervous system. paramecium are much larger than bacteria (~300 microns at its longest length, comparable to baby C. elegans larvae), and they are single cells without any neurons. they do have ion channels and they conduct electric currents, which only goes to show that you don't need a neural net to conduct currents. even plant cells have been shown to conduct currents, so many cells likely have this property. neurons are generally defined by shape (a cell body with at least one process reaching away from it) rather than by their ability to conduct.
hope this is helpful (and not too confusing).
- Mar 15, 2010, 9:42p
Along with saliva, irises should change a significant amount (contract/widen) in response to what someone imagines or sees, right?
How accurate do you think an experiment on "iris contractions" in a person's reaction to certain images would be? (Or do you think recording saliva pH would be more accurate than videotaping someone's irises...?)
- Mar 23, 2010, 11:31a
i'm not familiar with any experiments that discuss pupil dilations due to imagined seeing. is there a specific paper you're thinking of, or are you just assuming this to be true? it's an idea that should be really easy to test.
measuring saliva pH was used in the study above, but i don't remember if they did the positive control: can a non-comatose person control the pH of their saliva to answer questions? i'm not sure (though i assume yes). if a non-comatose person can also control pupil dilations in response to questions, it seems reasonable to test this on comatose patients too (as you suggest).
- Mar 23, 2010, 5:36p
I have read of non-comatose people seeing specific pictures and having their saliva pH change accordingly.
How can a comatose person's saliva pH/pupil dilations change if they are unresponsive? If the neurons controlling the muscles are "dead," isn't the person only able to breathe?
(If a comatose person's nociceptors/non-nociceptors don't work, does that mean that they feel absolutely no pain either?)
- Mar 31, 2010, 9:31p
A comatose person may not be able to control some muscles (such as their arms and legs), but may be able to control other body functions (such as their eyes or saliva pH). It probably comes down to exactly why they are "comatose", and whether they have any residual function. I think you can probably be comatose for many reasons, not just because your neurons are dead. For example, maybe a connection to the muscles is missing, and the neuron needs to regrow the connection. But I think this is still very much an open research question, with little known (as far as I know).
Same goes for pain. As an example, there are stories of people who say that they experience pain while under anesthesia. They claim that they were actually aware during the surgery, and that they could feel, but they couldn't speak or respond in any way.
It's complicated stuff, very little of which is actually understood.
- Nov 18, 2010, 7:19a
Hey Nikhil! Is not consciousness a defining feature of all living organisms in their normal living state? Well, atleast thats what we learn in our high school biology! In that sense even C. elegans or for that matter any other worm should also be conscious! It is just that we dont know how these different organisms exhibit their conscious behavior... isn't it? or am i utterly wrong??
- Nov 19, 2010, 10:12a
Clearly some living organisms go through different states of consciousness (e.g. awake vs. asleep). So even if you're alive, if you're asleep all the time I don't think that that minimal level of consciousness (if any) would be enough for me to say that the person has conscious experience. So no, I don't think consciousness is a defining feature of all living things.
Also, the definition for life is hard at its boundaries: is a virus alive? It only has DNA, and depends on other organisms to reproduce. It's so simple that some believe that it's not even fair to call it alive. So if we're not sure whether it's alive, how do we know if it's consciouss? More interestingly, does a virus (or bacterium) have anything that it feels like to be that virus (or bacterium)? Maybe it's so difficult for us to imagine that we might say no. Or maybe all living things have consciousness, so then a virus must too? Hard to say.
- Nov 20, 2010, 8:29a
Well, a virus i think cannot be considered as living.. but yeah some do consider it as living.. its just a mass of protein with a little genetic matter.. so we cannot search for consciousness in a virus... but i mean all organisms which have nerve cells starting from phylum coelenterata or platyhelminthes i guess should be conscious! but as you said having neural networks may not necessarily indicate consciousness.. but then why is it so?
And yeah.. this is a very interesting thing that when we sleep we are in the subconscious state! And i feel sleep especially the REM stage should be one of the hottest area of research! Is this area really given such high importance?? I did read about your own sleep experiments with the night vision camera! That was really interesting! I'll try it myself!
But investigating genes associated with consciousness is a far fetched idea but equally interesting also! All the best bro!! I'll follow your work and hopefully one day you will come up with the genetic basis of consciousness!!