Rizzolatti, G., & Destro, M. F. (2008). Mirror neurons. Scholarpedia, 3(1), 2055.
[Note that the main reading has been updated. Bonini et al (2022) was too hard to understand, Rizzolattti & Destro (2008) has been substituted for it]
What might be the functional role of the mirror neuron system? A series of hypotheses such as imitation, action understanding, intention understanding, and empathy have been put forward to explain the functional role of the mirror neurons. In addition to these, it has also been suggested that the mirror neuron system represents the basic neural mechanism from which language evolved.
Bonini, L., Rotunno, C., Arcuri, E., & Gallese, V. (2022). Mirror neurons 30 years later: implications and applications. Trends in Cognitive Sciences.
Optional:
Bandera, J. P., Marfil, R., Molina-Tanco, L., Rodriguez, J. A., Bandera, A., & Sandoval, F. (2007). Robot learning by active imitation. INTECH Open Access Publisher.
Cook, R., Bird, G., Catmur, C., Press, C., & Heyes, C. (2014). Mirror neurons: from origin to function. Behavioral and Brain Sciences, 37(02), 177-192.
Bonini, L., Rotunno, C., Arcuri, E., & Gallese, V. (2022). Mirror neurons 30 years later: implications and applications. Trends in Cognitive Sciences.
NOTE TO EVERYONE: Before posting, please read the other commentaries in the thread (and especially my replies) so you don't just repeat the same thing.
ReplyDelete**BLOGGER BUG**: ONCE THE NUMBER OF COMMENTS REACHES 200 OR MORE {see the count, at the beginning of the commentaries] YOU CAN STILL MAKE COMMENTS, BUT TO SEE YOUR COMMENT AFTER YOU HAVE PUBLISHED IT YOU NEED TO SCROLL DOWN TO ALMOST THE BOTTOM OF THE PAGE and click: “Load more…”
Delete________________
Load more…
________________
——
After 200 has been exceeded EVERYONE has to scroll down and click “Load more” each time they want to see all the posts (not just the first 200), and they also have to do that whenever they want to add another comment or reply after 200 has been exceeded.
If you post your comment really late, I won’t see it, and you have to email me the link so I can find it. Copy/Paste it from the top of your published comment, as it appears right after your name, just as you do when you email me your full set of copy-pasted commentaries before the mid-term and before the final.
——
WEEK 5: Week 5 is an important week and topic. There is only one topic thread, but please read at least two of the readings, and do at least two skies. I hope Week 5 will be the only week in which we have the 200+ overflow problem, because there are twice the usual number of commentaries: 88 skies + 88 skies + my 176 replies = 352!. In every other week it’s 2 separate topic threads, each with 88 skies plus my 88 replies (plus room for a few follow-ups when I ask questions.
When I first read the title of the section language evolution in the context of mirror neurons, I thought it was going to talk about language acquisition, and how we hear combinations of sounds from humans around us and imitate them, acquiring the grammar based on statistical learning or any other mechanism. But it was really surprising to read that the involvement of mirror neuron system was more ancient than our current way of communicating. Learning that our primary mode of communication was through gestures that were immediately understood when interacting with someone as we were projecting ourselves in their speech and their intentions was interesting. And I wonder, if because we’ve evolved to primarily use words to communicate, our ability to understand other people’s mental states has been affected. We know that facial expressions are rather easy to read and interpret in that they stimulate mirror neurons in our anterior insula, but if we were to only use words to communicate our mental states, would it be more difficult to empathize?
ReplyDeleteThe first thing that mirror neurons seem to enable us to do is to imitate the movements of others. How do they do it? That's still not known. And we already knew people could imitate movements.
DeleteThe guess is that the capacity to imitate a movement is somehow based on the similarity between the pattern of performing that movement and the pattern of seeing someone else perform that same movement.
But imitation is not yet communication.
And gestural communication is not yet language.
And motion is not the same as emotion. (Or is it?)
What's certain is that if we could not FEEL our internal states, it's unlikely that we could communicate them in words. And being able to communicate what internal states feel like using movements (including vocalization) that accompany (or even resemble) what internal states feel like probably preceded being able to describe them in words.
I get your point on saying that feeling our internal state brings us to the ability of communicating about it, and that humans probably used signs to express how they felt before being able to vocalize it with words. Nevertheless, I don't see how imitation differs from communication. If we take young children for example, that act of playing has a huge impact on their development, and one of the first games that is also seen in non-human primates, is actually the imitation. Studies have shown that reciprocity in a game brings children to experience more positive feelings during the interaction and it could be seen as a way for the child to express to his partner that he’s enjoying the interaction and that he’s interested. So in that case, how does that differ from communication?
DeleteYes, imitation can be communication, but it's not language (Weeks 8 and 9). What do you think might be the difference between language and other forms of communication?
DeleteImitation being a form of communication rather than a language is reminiscent of the portion of the article that discusses some imitations being seen through facial expressions, and facial expressions in itself are not a language. It also touches on the use of imitation in learning and social situations among species displaying how it is not simply just a form of communication. With imitation and communication as a whole being described as a broad range of things, I think language is a bit more niche and complex in its own individual way. As we have discussed symbols previously, not all communication is fixed with symbols (such as imitating facial expressions); however language has fixed symbols and rules for writing or speaking. I believe one of the main distinguishing factors is that imitation and other forms of communication are often limited to observable behaviors compared to language which can dive into complex concepts and thoughts and is less universal than other forms of communication.
DeleteThe special kind of communication that language is will become clearer as we go along.
DeleteTwo things to muse on are (a) the difference between showing and telling and (b) the amazing fact that you cannot communicate everything by showing (why not? examples?), but you can communicate every proposition by telling (do you believe that? why? or why not?).
(And remember that (1) "information" is the reduction of uncertainty about which (of a finite number of alternatives that matter) is the RIGHT one (think of the (vegan) sandwich machine) and (2) remember that to categorize is to do the RIGHT thing with the RIGHT KIND (i.e., category) of thing.)
I would say that one of the distinctions between imitation and other forms of communication relies on what each one is based on. To imitate, you need a behavior, a facial expression or a movement to replicate whereas language for example is based on the manipulation of different symbols. We could also separate those 2 concepts in terms of when it arises in a child’s development. Imitation is frequently seen in children but tends to disappear as we grow, whereas language is acquired through experience and continues its development even when the frequency of imitating behaviors decreases.
DeleteIf communication purely relied on showing, we wouldn’t be able to describe abstract concepts that require more than sensorimotor experience with the world. For example, the concept of “believing” cannot be shown, but we can surely explain how we feel about it, or maybe show some examples but we won’t be able to show what believing looks like. Furthermore, we try to make sense of what we see and this involves interpretation. So, despite what the person is trying to explain to us through showing it, we might have a biased interpretation of it, and thus increase the interpersonal gap.
DeleteOn the other hand, telling requires language or other forms of communication, this way we can address every concept that we know. Nevertheless, even if we are able to describe everything, including our feelings in a specific situation, we wouldn’t be able to explain why it felt that way to feel this, in other words, we face the other-mind problem and we need to assume that others are able to feel the same thing as we do.
This paper defines mirror neurons as they were first identified, and explores how the definition and understanding of their role in neural activity and cognition has evolved in the ensuing 30 years. I am curious to the degree to which MNs represent their own unique cell type and to what degree ‘mirror neuronness’ refers to a characteristic shared by most neurons to varying degrees. As the article states, there are MN populations covering all the sensory modalities and ones that demonstrate different extents of sensitivity to other- or self-initiated behaviour. The utility of having MNs seems obvious - in being able to learn from the consequences of others’ behaviour and to build social connection through reflection and understanding of others’ experience. It seems, based on the descriptions provided in this article, that there are many examples of neurons that exhibit MN-like qualities in addition to their own roles in mediating the behaviour of the agent themselves, which to me raises the question of how useful it would be to have a distinct class of MNs versus having MN-like activity be a facet of neurons already capable of facilitating the behaviour in the agent themselves.
ReplyDeleteAn additional question regards the authors’ mention that it is unclear whether MNs are innate or emerge following learning. I am curious how this might be tested, especially if we imagine that MNs could emerge in a later developmental stage. Since we do not know much about a unique molecular basis for MNs (and indeed, seem to have a fuzzy definition for what constitutes a MN in the first place), it seems like a big ask to identify them during development. But, it also seems unlikely that they are an emergent result of learning only (without some innate basis) considering how pervasive and preserved they are across individuals and species.
Mirror capacity is the capacity to imitate the sensory shape of perceiving a movement with the motor shape of producing that movement. We already knew we could do that kind of thing. Now we know that the neurons and areas that are active under those two conditions (producing and perceiving) are the same or near one another, but we still don't know how they do it.
DeleteMirror capacity is probably partly (or in some cases) innate, and in other learned (or partly learned). But we still don't know HOW MNs do it.
The empathy and emotion part of this reading was particularly interesting to me, as I wonder if there are potential implications for research and treatments for psychiatric illnesses wherein there is a lack of understanding for what others are feeling, like in ASD, or a lack of caring for the outcome of others and their feelings(more commonly observed in ASPD). I remember reading about a study where individuals had a harder time identifying facial expressions when they had a gel mask on their face(interrupting the sensations they could feel on their own faces, and their movement). I also remember a study wherein forcing a facial expression can evoke feelings associated with that expression(eg. A half-smile can evoke positive feelings). It would be interesting if any of this understanding of observation of others' movements/expressions and our own physical sensations could help train empathy in a clinical context or at least help us reverse engineer an understanding of what empathy really is.
ReplyDeleteI was also very interested in this effect when it was presented in the paper. I’ve heard of similar studies conducted on how physical expressions of emotion can lead to a genuine experience of that emotion. Perhaps this information of mirror neuron functioning can be used as potential treatment methods in the future. There has been some work on deep brain stimulation to treat some neurological conditions and while I don’t know entirely all of it’s applications, it seems like we might be able to use this knowledge in an effective way for potential treatment or at the very least symptom management.
DeleteWhile the existence of mirror neurons are fascinating and thinking of their potential implications is exciting, we still don't understand how they work. We don't yet have a comprehensive model that explains how they might contribute to our ability to understand and empathize with others. They also don't help us reverse engineer a T3 and don't help explain the "how" of cognitive science. Definitely a lot we have yet to uncover about their inner workings.
DeleteJosie, good questions; especially relevant to a vegan activist like me! The deliberate smile's effect on mood might have some effect on mood, but is not a strong enough antidote for major depression. But there may be ways to strengthen feelings of empathy; we'll discuss them in Week 11:
DeleteAnimal Sentience and the Mirror Neuron Initiative
Jenny, deep brain stimulation is a bit far afield of cogsci, at least for now.
Miriam, spot-on.
Josie, I also found these discoveries very interesting, and it made me think directly of the Psychic Sponge Effect, being when you feel something simply by being exposed to someone else feeling the thing (because of a great empathy).
DeleteMiriam, it's certain that they can't help understanding the "hard problem" but it seems like nothing really helps explaining it these days. And mirror neurons don't necessarily serve the purpose to explain it because why would every phenomenon happening in the brain be useful to identify the "Hard Problem" if the hard problem is itself still a very tangible concept ? Maybe the key to answering the hard problem will be a sum of some concepts, or even one key concept, that don't even englobe mirror neurons.
Forget the HP for now and focus on the EP: What would reverse-engineering mirror capacity help explain?
DeleteReverse-engineering mirror neuron capacity would help us explain the EP (how and why we can do what we can do). MNs would help explain the how of things we can do by researching the structural and functional pathways of MNs. If we could do this and implement these pathways into a T3 and it was indistinguishable from humans in regard to mirror capacity, cogsci would get closer to solving the how of the EP. MNs would help explain the why of things we can do by researching how MNs evolved into a genetic trait. For example, it could be that MNs became a sustained genetic trait because it was advantageous. If one homo sapien mirrors a more skilled homo sapien this is advantageous to survival or reproduction and over generations, even a small advantage of mirroring more skilled peers, could become a dominate trait.
DeleteAs someone who studies linguistics, I found the theory of gestural speech as the origin of language very interesting as I had never heard of it before. The authors dispute the widespread belief that “human speech evolved from animal calls” by indicating four key differences. The difference that I found the most compelling was that human speech is usually directed at intended receiver(s) while animal calls usually do not have a specific recipient they are directed towards. The use of gestural communication through motor neurons makes a clear connection between the sender and the receiver of a communication signal, further solidifying the gestural theory of speech origin.
ReplyDeleteYes, the perception/production mirroring is most pronounced in human language, and it's there in miming too. More on this topic in Weeks 7-9)
DeleteThis reading made me wonder how (and if) mirror neurons could possibly provide another insight into the Other Minds Problem (which, to reiterate, is the problem that we can never know if others have conscious thought or feeling since we can never get inside another individual’s head.)
ReplyDeleteIn “Emotions and Empathy,” the author presents a study about how presenting a participant to faces expressing the emotion of disgust activated the same mirror neuron system as when the individual was presented with a disgusting odorant. Hence, the author concludes here, that mirror neurons may be involved in “feeling the same emotions that others feel.” Similarly, in “Intention Understanding,” the author discusses evidence suggesting that IPL mirror neurons appear to not only recognize but also to “understand the intentions of [an observed] action’s agent.”
Though this does not completely eliminate the Other Minds Problem, it does cast light on whether it is correct for us to say so strongly that “We can never get inside another individual’s head,” and if the Other Mind Problem needs to be reworded in less extreme terms.
Good points. And this shows a way in which Darwinian evolution also provides a "Periscope" on the Other-Minds Problem (OMP) with all the mirror-capacities: perception mirrors production and production mirrors perception What is produced may be a movement or a felt state.The mother smiles, the baby smiles back, and it's clearly not just imitating the movement, but feeling what it feels like to smile.
DeletePerhaps the most powerful perception/production mirroring is with language: not just the speaking/hearing mirroring, which is a mirror-neuron function too (and songbirds as well as parrots can do it), but the understanding/meaning mirroring: You say "the cat is on the mat" and I don't just hear the sounds "the cat is on the mat" but I also understand what you are meaning to say. This is much more difficult to reverse-engineer than the perception/production mirroring of seeing and imitating mpvement, though for (toy) roboticists even that is still a big challenge (but surmountable).
Hi Ohrie,
DeleteI have to say I share your thoughts regarding mirror neurons, though connecting them to the other mind’s problem is a fascinating possibility I had not considered. Originally, when I was reading the Language Evolution section of the paper, I couldn’t help but think that mirror neurons, in addition to possibly being the foundation of language as the gestural theory of speech argues, could also be related to how we can interpret prepositions. Perhaps if we wish to solve the question of the homunculus and how it is able to “understand” prepositions, mirror neurons would be an element to investigate.
I think I am sort of touching on the idea you are alluding to Dr.Harnad when you speak about “the cat is on the mat”. Perhaps evolution evolved this periscope because a key aspect of efficient learning is being able to infer intention as being able to do so allows for effective imitation.
Do you have any hunches about how there might be a mirroring between the production and the understanding of language (propositions)? It's not as simple as the mirroring between the production and the perception of speech.
DeleteTowards the end of the paper a hypothesis for what led to human speech was mentioned; namely that the path to speech started with gestural communication, and sounds were associated with the gestures, and then later the sounds became more dominant. The mirror neuron system possibly supports this hypothesis because of mirror neurons’ ability to make motor copies of observed actions. With this translation from an observation to activations in our parieto-frontal neural systems, we may be able to understand the feeling of what it is like to do that similar gesture and therefore understand ‘why’ that person made that gesture.
ReplyDeleteIf the gesture from the person communicating with us is causing activations in our brains that may help us ‘mimic’ making the gesture ourselves, then we’re close to causing the gesture and therefore closer to being in a similar state to the person that is causing the gesture, and therefore (the last therefore I swear) we may be understanding them because we’re in a similar state.
Yes, that's right, and it's quite remarkable. We'll be discussing the nature and evolution of language in Weeks 6-9 pretty intensively. But the transition from gesture to speech came after gesture became language. Gestures had to stop being imitative or iconic; they had to become arbitrary symbols. (Does that sound familiar?) Then it no longer matters if you use gesture or vocalization, and vocalization has huge advantages. (Can you identify some of them?)
DeleteIn terms of the advantages of vocalizations, my immediate thought was that vocalization offers a wide breadth of symbols to be used. Speech consists of very subtle changes in tongue position, constriction of vocal tracts, control of vocal chords etc. which produces a multitude of different sounds. And somehow, our brains are able to detect these subtle phonetic differences, and interpret all these different sounds as symbols. But then I realized we could have evolved to pick up on very subtle differences in hand/arm movements or even facial expressions, and learned to interpret these as symbols instead, and used these as the primary building blocks for language. Thus, I realized this isn't really a distinct advantage of vocalization.
DeleteOne actual advantage of vocalization is that it can be used as communication even when someone isn't looking at you or you can't be seen. Evolutionarily, if you are in danger and need help, you could call out and be understood by another human being without them needing to see you.
Yes, evolution could have optimized gesture discrimination, but vocalization is inherently faster and works in the dark, and at greater distances, and out of sight, and when your hands are occupied... Any other candidates?
DeleteBased on the paper, “speech in humans is not, or is not necessarily, linked to emotional behaviors.” I think the term ‘emotional behaviors’ is a weasel term and it is not well-defined. I believe that every behavior comes with emotion and the boundary to distinguish what counts as non-emotional behaviors is unclear. And what is the relationship between emotion and intention for adult human? Can you ever talk without emotion? For example, for animals, and maybe also for infants, emotion, such as feeling sad, is the absolute intention for them to communicate, or ‘call’ from the article. But how about adults? And does emotion the same as feelings? And if so, it seems like we are trying to address with the hard problem here.
ReplyDeleteAnd since the mirror neuron system involves the empathy, which is the state that you feel what other is feeling, is it another evidence to prove that we need neural system, or at least mirror neurons, to solve the hard problem, which is ‘how’ and ‘why’ we feel. So, this means that we need to have T4, as it with neural network, in order to have feelings.
Hi,
DeleteI wanted to add on to some things and maybe help you understand the “emotional behavior” part you were confused about. I think what Rizzolatti and Destro meant in that part was that when animals make vocalizations, those sounds directly reflect their emotional states, but that human speech is more complex than animal calls where our emotions may not be conveyed when we communicate.
As for the relationship between emotion and intention, I would argue that our emotions often drive our intentions. However, I would be hesitant to draw a direct relationship between emotion and intention, since I think intentions are not solely based on our emotions. I guess that would depend on what you define emotions to be. For example, I would ask you whether you think “doing the right thing” is apart of your knowledge or your emotions. I am asking this question because human actions can contradict their feelings/emotions, but if we’d consider this to be another emotion altogether, then I want to know how we define the relationship between thinking and feeling.
Also, I do think that people can talk without conveying their emotions, maybe (maybe not?) we always feel something, but I don’t necessarily think it is always conveyed in communication.
Eugene, You made some good points, but you over-reached. This should clear it up: Yes, emotions, because they are felt, are feelings. But emotions are not the only feelings: You can feel hot or cold, tired or energetic, the feel of touching a smooth surface or a rough surface, light can feel brighter or paler, a sound louder or softer, and seeing different colors or forms is as varied in what it feels like as hearing different musical instruments: All those sensations are FELT states. That includes a completely unemotional reading aloud of the telephone book!
DeleteAnd it does not end there, because, as I keep demonstrating with Csenge and Daniel, it FEELS LIKE SOMETHING to understand English, whereas to hear Hungarian, if you don't understand it, feels like what it feels to hear the sounds, but not what it feels like to understand the meaning (as with Searle and Chinese -- and that's true of ChatGPT too, despite all the languages in which it can chat: it doesn't understand a thing because it does not feel a thing: it just manipulates arbitrary symbols, following rules for combining and recombining them, with the help of the "Big Gulp".
Every (waking) instant is a series of states that it feels like something to be in. So, yes, of course the Hard Problem is involved, and cogsci has to try to explain. Just as cogsci has to explain how and why we can DO all the (cognitive) things we can do, it also has to explain how and why we can feel. We'll see in Week 10 why that's not so easy to reverse-engineer...
Selin, good points, but see above: We are not zombies, but we feel a lot more than just emotions.
Thank you so much for the clarification! I understand now if correctly: Human does talk without emotions, and feelings is a larger term that could include emotion, involuntary (or vegetative) states, and feeling of feelings itself.
DeleteThough I am still confused about the part where to draw the line between intentions and feelings. Thank you as well for your comment, Selin, and I would actually agree on the idea that intention always comes with emotion.
If we defined intention as 'urges to complete instinctive needs, such as hunger or thirst', I think emotions come together at the appreance of intentions and may even remain after urges are finished.
Lastly, I would like to mention the quesion from Nagel's 'what is it like to be a bat'. Even though we cannot be sure if what we feel is the same as what other's feelings, but 'Empathy' seems to be the way to have a peek.
Like mentioned in the article, it seems that we need to have this mirror neuron system to be able to develop empathy, so that we feel like 'the ChatGPT does not feel like understanding...'
Once again, this could be the evidence to have T4.
Intentions (volition) will be discussed when we get to Libet and 'free will'.
DeleteI had to sit and think a while about how this paper pertains to (one of) the question cognitive science aims to answer: How are humans able to do what they do?
ReplyDeleteSo what aspects of our capabilities do mirror neurons help us understand? They are somehow tied to our abilities to recognize actions being performed by others, to copy each other physically, to understand others' intentions, and to recognize emotions in others and identify with them. Furthermore, they may help us understand where language came from. If human communication began as gestural (as is hypothesized in the paper) then these MNs would allow the receiver of a message to grasp the meaning of a communication directly from observing the action without having to process the message. Observing an action in a certain context creates similar neural activity in the receiver as if they themselves were performing that action in the same context.
So then do mirror neurons provide us an explanation for how spoken language came to have meaning that could be directly grasped? By linking immediately graspable gestures with noises, we came to be able to associate the noises alone with the meanings previously carried by the gestures?
To me, so far every skyreading has aimed broadly at the question of how it is that humans are capable of doing ALL the things we are capable of doing. Instead, this reading discussed a specific component of the physical implementation of the "computer program" (for lack of a better term) that humans are running. I'm not sure I fully understood what aspects of our cognition these mirror neurons are meant to explain but I've given my best guess above. If anyone has anything to add, I'd love to hear it!
I think the idea may be that if motor neurons are responsible for so many essential cognitive functions, they provide further evidence for the fact that cognition can’t be just computation. The theory that motor neurons are involved in gestural language means that the development of language hinges on motor capacities, which a purely computational system wouldn’t have. To replicate cognition, you’d also need some system that is capable of perceiving motor actions and then copying them. This system cannot encode the perceived gestures symbolically, like a computer does, or the copying function would lose its ability to overcome the parity problem of communication. As a side note, if someone could explain the parity issue of communication to me I’d be very grateful.
DeleteI entirely agree with both of your comments. From my understanding of this article, the logic of mirror neurons is close to the feature of cognition provided during the seminar: MN and cognition are processed based on shapes. In particular, MN seems to have a weak equivalence, as through it, we observe the action and then produce the same action by ourselves as the outputs.
DeleteI asked ChatGPT what the parity problem of communication is, and it says, "the parity problem in communication typically refers to the issue of ensuring that data transmitted over a communication channel has not been corrupted during transmission." I think this answer is not very comprehensive. The parity problem is caused by the distinction between what the sender wants to express and the information the receivers get. This may be caused by the fact that even for one symbol, different people give it different meanings. For example, once I ask whether what they eat tastes good, they may wonder whether I am hinting to them to share their food with the others, while I want to express the shallow semantic meaning of what I said. I hope my answer doesn't cause you more confusion on this question, and wait for someone to provide an honest answer if my explanation needs to be corrected.
First, the existence of mirror neurons does not actually explain anything because we already knew we had these mirror capacities, so all we learned now was that they are correlated with activity in certain sensorimotor cells and areas. We do not know how or even whether)those cells and areas cause mirror capacity.
DeleteHowever, MNs do show that mirror-capacity (some sort of pattern-based perception/production congruity) seems to occur in several different sensorimotor modalities in the brain (seeing, moving, hearing, speech).
Sensorimotor transduction (receptors, effectors) are T3 and non-computational -- at the least at their endpoints, but probably also in the processing in between. But since we do not know how mirror capacity is produced, we are only guessing.
ChatGPT was right about parity. It's just a way of checking the fidelity of signal transmission. (Parity is also discussed by Pinker & Bloom in Week 8.) In the context of mirror function, Rizzolatti & Destro are overstating the case, since we don't know how MNs work. R & D seem to be using "parity" metaphorically.
I’m interested in the part that links visual perception to the motor system to illustrate the limitations exhibited in the case of visual perception alone, resulting in an incapacity to provide information about the observed action and what it means to perform this action. The logic behind this limitation is the need for multiple systems to cooperate, which reminds me of the concept of complexity. This is a weasel word, but an appropriate one to describe this situation. However, regardless of whether the activation of the mirror circuit is of low or high complexity in providing an understanding of the observed action, the problem of underdetermination remains. As we cannot determine how the mirror circuit accomplishes this or whether the process involves other circuits or not.
ReplyDeleteThe conclusions of emotion and empathy, when combined with the conclusions of the language evolution section, lead me to believe that realizing T4 could solve the hard problem. Since a T4 machine would have to be physiologically and physically indistinguishable from a human, it indicates that if we want to build a T4 machine, we have to build a machine whose structure and function are physiologically identical to that of a human. According to the idea of reverse engineering, if we are able to realize the construction of the hardware in the first place, it is more likely that the software will be realized as an incidental product of the hardware, e.g., humans have so many innate "software" functions. Thus, if we build mirror neurons and circuits into the prototype T4 machine, we can test whether the mirror mechanism achieves parity and direct comprehension in communication, as well as intact emotional feelings. This is because emotional feelings result from the activation of circuits that regulate the corresponding emotional responses.
I wanted to discuss the implications of the emotion section on the hard problem. I see that Jinyu has already commented on the subject. The phrase from Scholarpedia “these experiments suggest that feeling emotions is due to the activation of circuits that mediate the corresponding emotional responses” implies the hard problem, i.e. explaining why we feel. “Circuits mediating emotional responses” seems to be hinting towards computation, that there is an output created by a process (algorithm?) that occurs in mirror neuron circuits. However, we still have much to understand about mirror neurons and how the circuits operate, it may turn out that they use analog processes and not computation. This uncertainty means the hard problem is not solved. I believe we need to fully understand how mirror neurons work BEFORE we can incorporate them into a T4.
DeleteJinyu, kid-sib could not follow you.
DeleteNicole, yes, we have no idea how MNs produce mirror capacity. See other replies about feeling.
And what does each of you mean by "computation"?
This paper seems to show that mirror neurons play a very active role in actions that we consider to be highly cognitive, including prosocial decision-making and empathy, and that the way that we do these things has a lot to do with our own motor systems. As an example, it's been shown that our capacity to make judgements about people's emotional state based on their facial expressions is largely linked to our own capacity to make the same facial expressions. It's clear that a T2 system can't make facial expressions like we can, and in theory a T3 system should be able to. This form of empathy is only one of many of our cognitive functions, but it seems to reinforce the idea that T3 is the minimum for reverse-engineering human understanding and thinking.
ReplyDeleteIt's certainly true that the motor system has a lot to DO with how and why we can DO what we can DO. But so does the sensory system; and mirror capacity is sensorimotor: perception mirrors production and vice versa.
DeleteHuman’s primary speech production language areas in the brain (areas 44 and 45) lie adjacent to our premotor and primary motor cortex (areas 6 and 4), and we know that the frontal cortex was developed later in evolution. This makes sense to me that mirror neuronal activation in the motor cortex were associated with original forms of language, which Rizzolatti and Destro allege to be gestural language. Speech production possibly evolved later alongside the motor cortex for more precise and efficient communication.
ReplyDeleteRizzolatti and Destro separate human language from the evolution of animal calls and connect it with motor areas in the brain. This stuck out to me because humans are very unique in our development of reading and writing which require motor functions. If our original language was based on gestures and highly connected to the motor cortex and mirror neurons, writing seems like a natural development of motor language communication. As we attempt to solve the hard problem in cognition (how and why we feel) and mirror neurons require feeling for learning, reverse engineering mirror neuronal activity is likely an important piece of the puzzle.
Reading and writing came rather late in the day to have channeled the evolution of language. Think of other reasons gesture might have been a much better starting point (if not the best ending point).(Weeks 7 & 8)
DeleteThis first reading made me think about the Bobo Doll experiment – a study about observational learning in children. The children’s mirror neurons would fire when observing the adult – performing whether aggressive whether empathetic behavior on the doll. An interesting finding was that self-identification to the adult was crucial in observational learning. The more the child would identify to the adult the more likely they were to act the same. Also, the children would be more likely to mimic the adult’s behavior if they had seen it in person then on a video recording. This makes me wonder about the self-identification properties a mirror neuron could have, or is that an info that is just relayed to them, then how? The digital versus real-life component is also very interesting – would that be different if the experiment was done today, would mirror neurons (or the neuro component in charge of this process) have evolved to be more responsive to digital images?
ReplyDeleteGood questions, but they are about the fine-tuning of mirror capacity when the main effect is not yet reverse-engineered.
DeleteIn Rizzolatti and Destro’s review of mirror neurons I found the discussion of mirror neurons and action understanding the most interesting. This section examines the hypothesis that the function of the mirror neuron system was to aid in action understanding. Here they argue that “mere visual perception, without involvement of the motor system would only provide a description of the visible aspects of the movements,” but would not provide the meaning of the action, or why it was carried out. Thus, mirror neurons provide “a real experiential comprehension of the observed action” to the observer. This discussion of motor neurons relates to a larger question that I’ve had through this class: namely, how we are able to learn from and feel like we understand the sensory world and stimuli around us. Last week's discussion of Searle’s CRA has caused me to consider what processes enable us to feel like we ‘understand.’ This exploration of mirror neurons has provided a window into how we are able to feel like we understand, and what mechanisms may be underlying this feeling.
ReplyDeleteGood questions. MN capacity is not yet reverse-engineered (and probably involves more than MNs), but perception/production coupling seems to cover a lot of cognitive territory.
DeleteI found the reading on mirror neurons very interesting, especially the discussion on their functions. In fact, mirror neurons are thought to be one of the neural mechanisms from which language evolved, the basis of action understanding, as well as a key player in certain emotional content, such as empathy. While mirror neurons give us the ability to replicate or “mirror” others’ cognitive abilities, I do not think it explains how we are able to do so. In other words, knowing WHAT mirror neurons do is not enough to understand HOW they do it, or to reverse engineer it.
ReplyDeleteSpot on.
DeleteThis week’s reading on mirror neurons was interesting for many reasons. Firstly, I had only ever learnt of mirror neurons in the context of child development and learning (in which they play a big role). However, I never knew that they were the basis of so much of what we can do in terms of higher order function. The paper got me thinking back to the Turing paper’s and what it means if we want to build a successfully passing T3 machine. I now assume they would need something similar in terms of what mirror neurons do for us. Something that allow them to integrate sensorimotor experiences but also understand and, in a sense, predict what others will do by simply observing actions (as stated in the intention understanding section). I suppose if we could reverse engineer mirror neurons, it would also be possible to create empathy, which in hindsight I never would have guessed that it was related to mirror neurons. The implications of this paper, with the background of the last few weeks is interesting, especially in trying to see how it all fits together. Although I doubt mirror neurons fill all the gaps from the previous weeks, it gives us a sizeable step forward beyond the “cognition is only computation” arguments we debunked the first week.
ReplyDeleteGood points, but remember that they are all just fantasy until mirror-capacity has been reverse-engineered. But perception/production mirroring is certainly something that humans (and many other species) can do, somehow.
DeleteI think my original comment is erased at some point 🤔at least I cannot seem to find it🤔 so I shall post again
ReplyDeleteIt is very interesting that the mirror system is located in close proximity to the human language system in the brain. As the article states that the mirror neurons are associated with understanding actions within their contexts (sequenced actions trigger more firing in neurons), doesn’t this mean that similar to how an understanding of symbols is required for language, an understanding of gestural-symbols is required for activating mirror neurons? In which sense both systems are manifestations of cognition, that need an understanding of our interactions with everything around us; which means that prior symbol-groundings are necessary to trigger the mirror neurons. But, is it possible to know at which point are the gestural symbols grounded enough to activate mirror neurons? But then, this seems like questioning at what point can the cognition capacity of a child be sufficient as an adult.
(I originally posted on the 22nd, if this thread becomes a duplicate of that one, please ignore. My apologies!)
Close anatomical proximity is not a causal explanation; it's just a possible clue. The rest is speculation without a mechanism. So it does not explain anything. But, yes, mirror-capacity can be seen to underlie both the perception/production mirroring of (1) speech and hearing and of (2) language and understanding: What's the difference?
DeleteSpeech and hearing are physical and physiological processes, whereas language and understanding requires our interpretation for meaning ('a kutya fekszik az agyon'). Can I say that speech and hearing are the tools (as well as gesturing) for language and understanding? And MNs are give a possible explanation to HOW?
DeleteYou can say that, but it doesn't explain anything. With speech, you can say that the "shape" of the mouth-movements and the sound patterns they produce when you produce them are somehow isomorphic or analogous to the sound you hear when someone else produces them, just as with gesture and imitation. But what sort of shape similarity can there be between (M) the words you produce to express what you mean to say and (U) the words you hear and understand what they mean? There's something similar there,between M and U, but not what the words sound like when you hear them or feel like when you produce them; but what is it?
DeleteThis week‘s reading delves into the biological mechanism from where functions like intention understanding, empathy, and language evolution branch off. The mirror neurons fire when we feel emotions and when we see others expressing emotions (or we recognize an action). I wondered whether mirror neurons would also be found in animals besides monkeys and humans since studies show that many emotions in animals physiologically mirror those in humans. This suggests that similar mechanisms of mirror neurons for empathy and emotional understanding could exist in a wider range of species. Mirror neurons are crucial for our ability to understand what others are doing, thinking, and feeling (actions, intentions, and emotions). Computers and AI struggle with this, but if we can grasp the “what” and “why” of mirror neurons better, it could help us make progress toward solving the difficult problem of understanding consciousness.
ReplyDeleteYes, other nonhuman species have mirror-capacities and MNs. But it's HOW that is not yet understood. (The WHY is easier in this case: the really hard WHY question is about feeling.)
DeleteThe end of this paper emphasizes that mirror mechanisms do not explain the complexity of speech, but can help explain an aspect of the evolution of language. I felt that this was a really important point, since the original use of gestures, which activated some mirror neurons to help us understand the agent’s state, was closely related to the things in which they actually refer to (this communication form was highly iconic/imitative), whereas speech now almost completely consists of symbols that arbitrarily refer to things. If we hear a word now, and a mirror neuron is able to activate an area in our brain that could produce that same word, we still don’t know why or how that arbitrary string of sounds has meaning for us.
ReplyDeleteVery good point. But there are important reasons why gestural iconicity are important to get language started, but also strong reasons why arbitrariness was needed to migrate it to the much more efficient vocal/auditory modality: Do you have any hunches why?
DeleteInitially, gestural iconicity, with its reliance on visual and motor systems, could have laid the foundation for language by providing a tangible and direct link between representation and reference, making communication more intuitive and understandable. This would activate mirror neurons, aiding in the understanding of actions and intentions. However, as human societies grew more complex, the limitations of a solely gesture-based language—such as the necessity for visibility and the difficulty in conveying abstract concepts—became apparent. The adoption of a vocal/auditory modality, despite its inherent arbitrariness, allowed for a more versatile, nuanced, and efficient means of communication. It enabled humans to convey a wider array of concepts, including abstract and theoretical ideas, beyond the immediate and concrete, and allowed for communication even in the absence of visual contact, thus fostering more intricate social structures and cultural development.
DeleteI feel that gestural iconicity is important to get language started because it guaranteed that “what counted for the sender of the message also counted for the receiver” in that direct correlation between form and meaning meant that additional information was not necessary for understanding.
DeleteAs for why arbitrariness was needed to migrate it to the much more efficient vocal/auditory modality, my guess is that there is a limit to the variety of gestures that correlate to their meanings, especially when we get to specifics as it may be difficult to adequately distinguish two similar concepts. Iconicity also puts a limit on the amount of things we can name, as there are only so many feasible gestures that one could realistically produce. Arbitrariness would take the load off by not only allowing us to name things as we go, but also whatever we please.
One of the hypotheses for the function of the mirror neuron system is that it allows us to understand the intentions of actions from others. Beyond knowing what the action is, it allows us to infer the "why", and this was shown in an fMRI study where context was important to the activation of these neurons when shown a hand gesture. The mirror-neuron impairment that is noted in people with autism is thought to be the cause of a deficit in understanding the behaviour of others, but the reading notes that this could be related to deficits in organizing their own intentional motor behaviour. I am curious how this relates to the T2/T3/T4 debate -- if passing the T2 test requires convincingly answering questions about the intention of actions when taking context into account, for example "why is the hand reaching for the apple", then T3 capabilities to create this action on its own might be necessary to make inferences in the behaviour of others. If the only way to make these inferences is through the mirror neuron system, it might mean T4 is the only way to reverse engineer cognition. I'm personally more swayed by the T3 argument.
ReplyDeleteWell we don't know how to produce mirror-capacity, so who knows whether it needs to be done the T4 way? But it certainly has to be done a T3 way. ChatGPT (T2) certainly does not have sensorimotor mirror-capacity (yet it can describe it, and give examples, in words). Why not chat with it about this and let us know what it says?
DeleteI tried some different prompts with context scenarios attached to actions and asked for it to describe the intention behind the action. It was able to do this pretty well, at least I think indistinguishably from a person. I asked "A woman has been working all day without a break. When she comes home, she reaches for an apple on the counter. Why does she do that?" and it replied, "One of the most common reasons could be that she's hungry after a long day of work without a break for lunch. The apple provides a quick and convenient snack to satisfy her immediate hunger. Apples contain natural sugars and carbohydrates, which can provide a quick energy boost. After a tiring day, she might be looking for a natural source of energy." This seems like a pretty good understanding of intention (without mirror neurons involved) to me.
DeleteAt first I was surprised that ChatGPT was able to give a correct account of the intention behind an action and believed that the context was what gave it the ability to answer the question correctly. However, after I thought about how ChatGPT works, I came to the conlcusion that it would be able to say that the person wants to eat the apple if we were to ask it why someone would pick up an apple, simply because statistically speaking, eating is very closely related to food-related words. That said, it's still interesting to see that a T2 is able to imitate intention-understanding capacity of people.
DeleteThis also relates to the machine mentioned in the Searle reading, the program that can understand implicit meaning in stories (reading between the lines). A T2 like Chat GPT could use banks of verbal knowledge from across the internet such as books, film transcripts, psychological papers and nutrition to determine what is statistically most likely to be the intention of the woman in this situation. The idea that mirror neurons activate in response to a sign reflects the subject's personal involvement in the act ("I could also, and have also, reach(ed) for things because of hunger) directly shows that the implementation a program that produces a similar response is different between T2 and humans.
DeleteAll good observations. But ChatGPT can do even subtler integrations too -- along with bloopers and oversights. But what's especially interesting is correcting its mistakes. I hate that's it been trained to make a ritual apology every time. But what's interesting is how (at least till the session's over) it updates its information and state, so that if you're on a topic where its Big Gulp would be useful to you, your corrections make it able to be even more useful to you. Brainstorming with GPT is an art worth learning...
DeleteWhen I give out the mid-term, GPT of course makes it easier to cheat at about B/B- level (not A-level) and if you have not been following, it can expose you if you parrot a blooper you didn't know was a blooper.
But I don't mind at all if, when you HAVE been following the course, you try to improve your answer by brainstorming with GPT. Just be careful, lest it gives you supplementary information that is wrong!
(That's why I can't have TAs helping me with either teaching, or marking or skywriting in this course unless they have taken the course a couple of times already (and even then).
But I'm paying for the uniqueness of the course with my waking hours and my sleep, with an N of 88 = 2 x 88 = 176 skies to respond to and and evaluate every week, now the ceiling's been almost doubled, while you each just have to do 2 per week...)
But so far I'm not complaining. I get something out of a kind of brainstorming with you all too. It's fun as long as it doesn't kill me.
I found this reading very interesting, especially the part on how new findings indicate that the mirror mechanisms plays a role in fostering empathy as well. I had read an article in the past on how perhaps mirror neurons are an evolutionary response that drive our deep need to belong. We imitate and mirror others in order to be accepted by them, this was extremely useful for survival.
ReplyDeleteWell, social (and parental) behavior are much more complex than that (and we don't know how MNs work), but something along those lines...
DeleteI'm considering the implications of MNs in light of our ongoing discussion. Mirror neurons seem to play a fundamental role in sensorimotor grounding for both humans and animals because they appears to be essential to get “a real experiential comprehension of the observed action”. This suggests that mere perception is sufficient for a description-level understanding, but nor for a grounded understanding of the perceived action. Thus, since MNs require a sensorimotor system, the ability to experience and to interact with something may be a precondition to action, intention and emotion understanding. The discovery of mirror neurons in humans and animals, in my view, supports the embodied cognition theory and the idea that at least a T3 robot is necessary to pass the T2 Turing Test.
ReplyDeleteBut we didn't need to discover MNs to know we had mirror-capacities...
DeleteOnly a T3 would be able to reproduce these mirror neurons because a T2 would not be able to make facial expressions for example. But is it possible that a robot can reproduce the same facial expression as us and therefore, understand how we feel? Would a T3 really be able to feel empathy of some sort? Is reverse engineering possible in this case because we know what mirror neurons do but know HOW they do it.
ReplyDeleteNo, a pliable face is not enough..
DeleteI am not quite sure I understand the experiment of the 2 monkeys in the section entitled: Intention understanding. I don’t think I understood the results of the study and how they prove that mirror neurons are involved in understanding intentions. Indeed, they mention that they found that many of the neurons in the inferior parietal lobe discharged selectively when the monkey executed a given motor act such as grasping. Then they said that most of these neurons fire only when the coded motor act is followed by a subsequent specific motor act such as placing. So, do they mean that they found that these neurons were selectively firing when the observer saw the agent (in this case the monkey) performing a motor action that was part of a ‘bigger’ action (when I say bigger, I mean an action with a specific purpose) such as grasping for food to eat it? Also, I’m not quite sure how this experiment worked, did they have a person watch the monkeys perform these actions while being branched to an EEG to measure the person’s brain activity?
ReplyDeleteYes, part of a "bigger" action which included the goal of the action.
DeleteProbably observation and recording of both the action and the brain activity.
But there was no need to hurt monkeys; it can now be done with humans noninvasively, without the blood, fear and pain.
First, gesturing alongside speaking is not gestural language. (Why?)
ReplyDeleteAnd not only imitation, but deliberate mimetic communication is not gestural language either. But gestural language does exist too: What's the difference?
Searle couldn't have said anything about the "mirror mechanism" because MNs were discovered a decade and a half after Searle's (1980) article (and the mechanism still isn't known, though we've known we could all do it since the advent of language!).
What we showed with the example of Ohrie's 3rd Grade schoolteacher was not that we understand language without understanding, but without understanding HOW. We're waiting for cogsci to reverse-engineer it (and all the rest of our cognitive capacity) and to explain it to us so we can understand how we do it.
This is not an explanation, just a criterion, but language-understanding is: T2 capacity + the FELING of understanding.
If the theory that verbal communication evolved from gestural communication is true, and the mirror neuron system is necessary for grounding the meaning of others’ intentions/actions in our own experience, what does this mean for our mission of reverse-engineering?
ReplyDeleteFirst, it’s more evidence that mere computationalism isn’t enough to create something that thinks. At least a T3 would be necessary to ground meaning of gestures (and eventually words) in experience of the physical world.
But we still don’t understand how MNs do what they do, and even if language is evolved from gestural communication, and if the MNS is necessary for gestural communication and the transition to speech as the primary communication, is the mirror neurons system still necessary for speech comprehension? A phrase like “the cat is on the mat” doesn’t carry so easily. Mirror neurons certainly may be necessary for grounding some motor actions in experiential meaning, but they’re certainly not sufficient for understanding.
Hi Elliot!
DeleteYou bring up some valid points. Re: your comment on whether mirror neurons are still necessary for speech comprehension, I think they are necessary to some extent. I remember seeing in a PSYC class that when someone says a word like 'dog', we have different regions of the brain that activate (what a dog looks like, smells like, feels like to touch, etc.) which is what (presumably) allows us to 'know' what a dog is. If we apply this idea to all words, then I would say that mirror neurons system is still necessary for speech comprehension, especially when we talk about actions.
One part of this reading that got me thinking was when it outlined the two distinct series of information which one can gain from observing another individual’s action: the ‘what’ of the behaviour, and ‘why’ they are doing it. This made me think that mirror neurons may be important to categorization, which we defined in class as ‘doing the right thing with the right thing.’ From a young age, by observing how others use specific objects and their intentions when using them, we are able to form mental categories for both objects and actions. For example, if a child sees their mother using a hammer as a tool to hang up something on the wall, this allows them to a) categorize the hammer as a tool for nailing things into the wall and b) understand the concept of housework as something usually carried out around the house using tools.
ReplyDeleteOne question I had pertaining to the reading was whether the mirror neurons responded solely to visual observation of carrying out an action. In other words, if I was doing a physical task with my hands but had a blindfold on, would my mirror neurons fire the way they do when I observe another person carrying out the same task, or is it solely visual observation that causes these neurons to discharge?
As per your question, the author (in a later neuroscience review) talks about the role of mirror neurons and their responsiveness to both visual, motor, and, auditory related sensory inputs. And mirror neurons can fire not only when you see someone else performing an action but also when you hear or feel the same action being performed. So what I find interesting is the cross-modality cases like you put forth. If you knew what the process of hammering a nail looked/felt like (you knew what it meant) it would likely trigger mirror neurons. Though it should be made clear that the literature says that visual stimuli usually evokes the strongest mirror neuron response.
DeleteAgreed! Your observation about mirror neurons and their role in categorization is intriguing. Mirror neurons are indeed fascinating because they seem to play a crucial role in our ability to understand the "what" and "why" of others' actions. While the concept is often associated with visual observation, research has shown that mirror neurons can also respond to integrated cues of other modalities. This versatility of mirror neurons underscores their importance in our cognitive processes beyond mere visual perception, and leads to the question of whether machines need full sensorimotor abilities to "understand" as we do.
DeleteZoe, mirror neurons may aid you in imitating a movement, but not in learning what to do with what: to categorize you have to learn the features that distinguish the members from the nonmembers: seeing someone eating a mushroom won't reveal those features; you need a lot of trial and error to learn those.
DeleteAs William and Kristie point out, there are mirror neurons in many modalities (seeing, hearing, touch, with some cross-modal generalization. But if you already know how to do something with your eyes closed, you don't need mirror neurons (for that task).
Kristie, Helen Keller could pass T3 even though she could not see or hear.
Professor, doesn’t figuring out what to do with what - i.e. learning to categorize - involve not just feature detection, but an understanding of the purpose of that category? For instance, to learn the category “tools,” and be able to categorize objects I’d never seen before as part or not part of the category, wouldn’t I to some degree need to understand what tools are used for and how certain features of objects which are considered tools aid in this pursuit? In that case, wouldn’t mirror neurons be at least partially involved in learning/understanding what objects are used for? Additionally, based on this reading as well as past discussions in other classes, I was under the impression that mirror neurons are “activated” both when I see someone else complete a task and when I myself perform the task. Thus, while they are integral to learning, they are not solely active during the learning phase. Have I misunderstood?
DeleteA few other students have already touched on the fact that our mere knowledge of the existence and roles of mirror neurons does not tell us how they do what they do, with which I completely agree. It is rather homuncular to hypothesize about function without knowledge of how to reverse-engineer MNs, but it regardless tells us a thing or two about Turing indistinguishibility; namely ruling out T2 on the basis of our innate need for sensorimotor integration of experiences (embodied cognition?).
ReplyDeleteI also found the optional reading, “Mirror neurons 30 years later: implications and applications”, quite interesting as it revisited core concepts to reveal that we still do not know how to answer the “how” question of their mechanism, which has given rise to controversies regarding the ontogeny of MNs. The main point of controversy lies in the fact that MNs might simply be the byproduct of more simple mechanisms of associative learning. Further, recent findings have uncovered that the mechanisms of social learning and behavioural responses may be an emergent property of a larger network, not restricted to individual mirror neurons but rather the sum total of MNs, MN interneurons, and other-selective neurons. This only serves to complicate the hypotheses of their mechanisms even more, and reveals that we are still far from understanding HOW mirror neurons do what they do.
Good points -- but stay away from "emergence": it's a... (you guessed it!).
DeleteIt's a way of seeming to say something, while saying nothing.
The paper mentions the hypothesis that language evolved first through the elaboration of gestures, and postulates that the existence and function of 'mirror neurons' offer support for this hypothesis. According to the author, mirror neurons facilitate our imitation of others' actions, understand the intention behind them and as such empathize with the person producing the action. The theory that language, the verbal kind that we are used to, evolved from gestures and subsequently became vocal is useful when we think about how elementary communication evolved into more complex and conceptual as humans transitioned from the stone age into the bronze age. If we run with this theory, it is interesting to think about the loop: did the evolution of language (from gestural to verbal) arise from changes in the way societies evolved? The other way around?
ReplyDeleteIn the context of this class, and thinking about "cognition is as cognition does", would it follow that a T3, a functionally identical 'robot'/computer, would need mirror neurons to empathize and therefore communicate in the same way humans do? Or would T4, a hypothetical robot with hypothetical mirror neurons alone be able to achieve this kind of identical behavior?
The paper seems to be attributing a very significant roles of mirror neurons in learning. It even suggests mirror mechanisms are the basis of language.
ReplyDeleteMy question is... what does that even mean, to say that "mirror mechanisms" are the basis of language?
What is a mirror mechanism one might ask? Something performed by a mirror neuron, one might respond. What is a mirror neuron? A neuron that performs a mirror mechanism...
It seems these neurons are mostly defined by their function, since they are not super specifically localized (even if they were it wouldn't mean much). But their function is still quite unspecified as it was derived from observating the activation of those neurons...
Am I missing something or is this really confusing?
Hi Emma! I agree that mirror neurons are a bit confusing. About your question on their connection to language: from what I understood, the article points to mirror neurons as possible tools for understanding each other; rather than needing to “cognize” in order to understand what someone is communicating, we just seem to automatically know. I think it’s easier to understand in reference to the gestural language the author is talking about: if someone points to you, you’re able to immediately recognize the motion and what it means, without much thought. My understanding from the paper is that in the language evolution hypothesis, the “mirror mechanism” is thought to give everyone a common foundation with which to build language, by giving us the ability to recognize and understand a common action, since mirror neurons fire when seeing someone execute the action, and when executing the action yourself.
Delete2609, I think you are conflating two kinds of "evolution."
DeleteOne is biological or Darwinian evolution (Week 7), in which genes code for biological traits and genetic variants are selected on the basis of whether they make the organism more likely to survive and reproduce.
The other is social "evolution", in which learned practises and skills are passed on from generation to generation on the basis of imitation or language.
Language evolves in both ways, but since we are the only species with language, the capacity probably began with adaptive genetic change. The gestural communication is not language, so that could have been a learned practice (though communication via pantomime is probably unique to our species too, hence likely to have a genetic component).
The transition from gestural communication to gestural language -- from iconic to arbitrary gestures, and from gesture sequences to subject/predicate propositions (the "propositional attitude"), and finally the transition from gestural language to spoken language) -- probably involved both genetic and social components. It's still mostly a matter of speculation.
Passing T3 would require mirror-capacity, but we still don't know the mechanism of mirror-capacity (as Emma and Lillian correctly note.
But mirror-capacity does not just apply to language (and in fact its involvement in language production and understanding is the most speculative of the authors' hypotheses).
Imitation is straightforward; vocal mimicry is more complicated, but probably still a straightforward mirror-capacity (perception/production).
The notion of "understanding" the "meaning" of movement is a lot more fanciful, however, and extending that to the understanding of a language sounds downright metaphorical to me -- though there are ways to make sense of it via category learning through feature detection, direct sensorimotor grounding of category names, predication and propositions allowing indirect grounding of more complex and abstract categories through (grounded) words and their inherited sensorimotor feature-detectors, mirrored in language (Weeks 6, 8 & 9).
This week presents the idea that humans started communication via gestural imitation which eventually evolved to use arbitrary symbols with semantic meaning. Why did we need arbitrary-ness? Infinite representational options…
ReplyDeleteBiologically, mirror neurons provide evidence that one of our most important aspects of cognition, i.e. language, evolved from sensorimotor transduction and non-computational means. Thus, symbol manipulation by a CPU is a reductionist view of language and its semantic associations to arbitrary stimuli. We are back to Searle: cognition is not solely computation. Think about American sign language (ASL): here, we have arbitrary gesturing that is NOT gestural communication, but real language, because of it is gesturing with meaning. Would understanding how we form these sensorimotor associations (between arbitrary symbols and their representations) actually help us understand cogsi’s ‘How’ question?
What need to be understood are (1) category learning ("cat") and (2) predication ("the cat is on the mat"). Reference is grounded by (1) and meaning is communicated by (2). Behind (or under) both is direct (sensorimotor) and indirect (verbal) feature detection.
DeleteOne possible "how" for feature learning (though probably not yet the right one) is deep learning (unsupervised and supervised).
In social psychology, we were taught that when Subject 1 mimics Subject 2’s facial expressions, it helps Subject 1 interpret and recognize what Subject 2 may be feeling. In the text, it is argued that although patients with schizophrenia cannot mimic or recognize facial expressions, they can be trained to by mobilizing facial muscles. I wonder how this kind of treatment, or Action Observation-Treatment, could be applied in the case of mental health disorders where patients suffer from emotional numbness. Could treatments including mimicry of “happy faces” help reduce symptoms of apathy overtime?
ReplyDeleteCould be -- but before cogsci know-how can be applied clinically it needs to be known...
DeleteMirror neurons are described as having the fundamental function of mapping “the pictorial description of actions carried out in the higher order visual areas onto their motor counterpart.” With regards to the development of language, the paper goes on to describe how the mirror neuron system allows direct comprehension of gestures and motor behaviours between individuals, because the mapping of visual input onto the motor cortex of the observer means “no arbitrary symbols are required.” This, on the surface, seems like a potential mechanism by which symbol-grounding could occur in the human brain! If a gesture is directly matched with a corresponding motor output, it seems as though the “symbol” of a gestural visual input can be grounded in sensorimotor information. This is fascinating, and seems quite plausible with gestural information, and perhaps even some vocalizations, because they are observed motor patterns that can be mapped to equivalent motor patterns to ground their meaning. However, I’m not sure how this mirror-neuron method of symbol grounding would work regarding verbal language… Linguistic symbols are still understood when we don’t see the motions of the speaker’s mouth, and even if we could, the speaker could be using words referring to abstract concepts with no clear motor referents, like “hermeneutics”. Can we bridge the gap between mirror neurons grounding gestural symbols or action-words, and highly abstract words? Or do more abstract words somehow ground themselves in meaning though some other system not related to mirror-neurons?
ReplyDeleteI was initially pleasantly struck by their reference to not needing symbols, and I found it intriguing how MNs appear to address the symbol grounding issue: it goes from perceiving (an action) to feeling (that action through MN activation), resulting in understanding “without any cognitive mediation”. But upon further reflection, I realized that in gestural communication, there's already a connection between the gestures and what they represent in the real world. This means there might not be a symbol grounding problem for gestures, as they are inherently tied to their meanings. However, this challenge still exists for language, where words use arbitrary symbols and must maintain a link between the symbol and its meaning. Perhaps delving into the transition from gestures to spoken language, those critical turning points, could shed light on how this symbol grounding problem in language can be better understood.
DeleteDaniel, first tell me what the symbol grounding problem is, and then tell me how your MNs could ground "apple" in the capacity to do the right thing with apples. And when you've done that, tell me how MNs do it for the castaway on the mushroom island. Mirroring is a concrete, eventually reverse-engineerable capacity, but it's not magic!
DeleteNatasha you are right that inherent in gestural imitation (communicating by pantomime)there is a direct connection between the imitative gesture and the thing being imitated (because of iconicity). So if you already know (somehow) what the thing being imitated is, and what's the right thing to do with the thing being imitated, then the gesture is already "grounded." That is one of the strong reasons to guess that gestural communication (pantomime) came before language (propositions).
DeleteBut pantomime is not language. And pantomime's scope is far more limited than that of propositions. (Can you give some important examples of things you can tell, but not show?) It might help to give some examples of things you can show but not tell -- unless your interlocutor has already been shown them.
This should illustrate why
(1) sensorimotor category learning,
(2) sensorimotor feature-detection and
(3) direct sensorimotor grounding
--have to come before
(8) indirect grounding through propositions.
--And it's why, in between,
(4) the gesture's iconicity has to
(5) make the mirror connection with whatever shape it's imitating,
--but then eventually
(6) the gesture's has to shrink and vanish into arbitrar shape
-- (but preserving the vanished iconic connection by shared convention), so that
(7) the "propositional attitude" can take over
--(is it genetic or learned?),
(8) providing the nuclear power of language to describe, define and explain anything.
--Then all that's needed is
(9) to migrate that nuclear power from the now-arbitrary gestural medium to the much faster and more efficient vocal medium (in which the whole process could not have started in the first place, because the iconicity of vocal imitation is so much more impoverished (except for calling cows "moo"). (Think of all the things you can mime gesturally but not vocally.)
Although this is not very related to our area of focus, I found the connection between mirror neurons and autism is intriguing. I've come to recognize that if mirror neurons are indeed crucial for comprehending the intentions of others, their malfunction may provide insight into the social challenges faced by individuals with autism. The common belief is that the main issue for autistic individuals, affecting their capacity to understand intentions, arises from a damaged mirror system responsible for interpreting other people’s actions. However, it's worth exploring the idea that the fundamental issue in autism could instead involve more basic difficulties in the organization of the motor chains. Conducting further research in this field could hold significant implications for improving interventions and support for individuals with autism. Studying individuals with autism could also offer additional perspectives and understanding of the nature and functioning of mirror neurons.
ReplyDeleteIt's all speculation, since we don't know the causal mechanism of production/perception mirror-capacitied, but if part of the deficit in autism is in producing one's own motor plans then it would be harder to perceive the motor plans of others
DeleteMirror neurons, from Dr. Rizzolatti and Dr. Destro's article, offer a unique perspective on how we perceive and internalize the actions and emotions of others. While the concept of mirror neurons is evident in assisting us in learning actions and possibly evolving complex linguistic abilities, I wonder then, in regards to Strong AI, if could we ever replicate intricate neural mechanisms in artificial systems beyond mere computation? Although the previous skywritings and lectures discussed this issue, I wonder if people's thoughts have changed after reading about mirror neurons.
ReplyDeleteAnother possible discussion question that I ponder is relevant to the reverse engineer problem is: as computation becomes more adept at 'interpreting' and mimicking communication, could they uncover or interpret unspoken human intentions or emotions more accurately?
Kid-sib does not understand what you mean by "replicate intricate neural mechanisms in artificial systems beyond mere computation". The last paragraph is also very vague.
DeleteIn the reading, it is said that mirror neurons might allow the observer not only to recognize the observed motor act, but also to understand the intentions of the action’s agent. This leaves me with some questions. Why would understanding one’s intention be important? Based on the reading, are we always supposed to understand people’s intention? I think not. What happens if our predictions of one’s intentions are not accurate? I think the word “understanding" might not be very appropriate in this case. I believe we can have the feeling that we understand someone’s intention but it doesn’t mean that it’s true. Mirror neurons might just create the illusion that we know what’s going on in other people’s minds but whether it is true or not is still up to debate in my opinion.
ReplyDeleteUnderstanding someone's intention in this context refers to comprehending the direct goal of their movement, not necessarily delving into their deeper motivations or thoughts. I think it’s important to understand the goal when observing an action, as it helps us relate to others and place their actions in a meaningful context. This aspect of understanding becomes particularly interesting when considering our earlier discussions on the true meaning of understanding, and what criteria is necessary for accepting that a system understands, as raised in previous conversations (such as one of the skywritings from week 1). I think that MNs bring us closest to what understanding something feels like, as they give us a sense of what doing that thing ourselves feels like, as pertinently explained by Marie in the comment below this thread.
DeleteLili, We have no idea how we do it, but, MNs or no MNs, we can (sometimes) tell the difference when a movement is made (1) just to stretch or rather (2) to reach for an object. And sometimes we can tell (from past experience) whether someone is praising you because you really deserve it, or they're trying to (vegan)-butter you up. In the case of movement-intention "understanding" and language understanding Rizzolatti may have overextended his inferences from the limited correlational evidence (and no causal mechanism).
DeleteNatasha, you are right to distinguish evidence from over-generalization.
This reading was interesting to me as I am particularly interested in social cognition (kid-sib how we understand and learn from what others do) and theory of mind (how we -at least try to- perceive what others feel). The existence of motor neurons seems to suggest a connection between what we perceive in others and what we feel. It doesn't explain how this connection happens, but it seems intuitive that the connection should exist. We have the intuition that understanding feels like something as we saw in the CRA and this seems connected to the idea that to understand what someone is doing (or trying to communicate) we need to have some idea of what it feels like to do the thing ourselves.
ReplyDeleteGood synthesis, and generous to Rizzolatti.
DeleteIt was interesting to read about how the motor system effectively imitates sensory input and operates as a core mechanism from which different functions such as those pertaining to empathy and intention and action understanding branch off of it. Ultimately, we don't know how the brain can create or lose mirror-neuron capacities. Considering this critical point, it seems almost irrelevant to consider why there exists the same connection between activity in the visually perceptive areas of the brain and activity in the motor areas without expressly viewing them in the context of a kind of reverse-engineering. I think above anything mirror-neuron symptoms illustrate how difficult this process is and adds another layer of complexity to the problem.
ReplyDeleteYes, lack of a model for mirror-capacity, or even evidence of whether it is innate or learned, should suggest some circumspection in interpreting and extrapolating from the correlation between the exercise of mirror-capacities and the activation of MNs or MN areas.
DeleteMirror neurons links an action and an agent together, through "mapping" the observed action to the motor system of an observer. It's really interesting to think about the clinical benefits that MN research can have, yet I am struggling to connect MNs to the Turing Test and reverse-engineering a possible T3: Observing an action of another serves two purposes. “One is 'what' the actor is doing; the other is 'why' the actor is doing it”, thus one of its functions is thought to be intention understanding. I was wondering if this is the same as mentalization? or is that just another weasel-word? Regardless, what confuses me about MNs is that yes, we can observe the neural correlates of mirror neurons in fMRI studies, and find/create situations where they fire. However, I don't get how this connects to reverse-engineering and even the easy problem (we don't know how MNs are doing their function properly). How can mirror neurons help us to understand how people FEEL, or what it feels like to experience their subjective experience. Even if we can see the neural correlates of "intention" or "feeling" via mirror neurons, intention understanding seems like its inherently linked with understanding how it feels like to be the other, which is not observable, so how can this help us to reverse engineer sensorimotor capacities or T3?
ReplyDeleteThe discovery of MNs provides a brain correlate of a capacity we already knew we had, in several different domains. It reminds us that these capacities have to be part of passing T3 (and T4) and are not possible with just T2 capacity produced by computation alone (e.g. ChatGPT). It is interesting (if correct), that whatever the causal mechanism is, it may be involved in all of these different production/perception domains -- correlating not just produced movement by oneself and perceived movement by others, but perhaps even felt states in oneself and felt states in others.
DeleteAccording to the reading, mirror neurons are important in our perception of others’ emotions and expressions, the understanding of the reason behind another person’s actions through mimicry, and thus the display of empathy. What I found most interesting is the finding about how it might be related to the inability of autistic children to relate to people and life situations normally. If that were the case, I wonder if understanding how mirror neurons exactly work and being able to replicate it, is the only thing we are missing to building a model that is almost indistinguishable to humans.
ReplyDeleteThe mechanism underlying mirror-capacity is not enough to pass the T-Test and reverse-engineer cognitive capacity.
DeleteAs a linguistics student, I really enjoyed the reading about the theory that the mirror neuron system could be the progenitor of language. I thought of how this theory fits in well with modern biolinguistics, especially the belief that the sensorimotor modality through which language is externalized is completely arbitrary. Psycholinguistic studies have shown that sign language has a complex syntax and phonology that are grounded in the same principles as natural language syntax and phonology that have been observed in speech, which contests the old hypthesis that sign languages were artificially invented, like writing. The evidence in this reading that point towards the even stronger claim that sign language precedes spoken language are very convincing to me. This claim also made me think of Chomsky's conjecture that language might not have evolved as a means of communication at all, and that the fact that we can use language to communicate is simply a convenient side effect. He suggests that the evolutionary purpose of language is to "link the interfaces" (I don't entirely understand what he means by that), and that it doesn't make sense that language evolved for communication because what we perceive in the externalized language (e.g., in speech) does not capture the syntax by which language is actually constructed in our heads. But mirror neurons could offer an alternative explanation to that: because "thanks to the mirror neurons, what counts for the sender of the message also counts for the receiver", perhaps the mirror neuron system sort of "fills in the blanks" of the imperceivable syntax behind sentences the receiver hears. Maybe because the sender of the message follows this abstract syntax to say what is spoken, mirror neurons force the receiver to reconstruct that syntax instinctively from what is spoken alone, despite the lack of the syntax in the physical message.
ReplyDeletePlease read the other comments and replies. How can the unknown causal mechanism of (mirror) capacities we all knew that we -- and therefore our brains -- have, explain anything about language to us? We know we could produce language, and we knew we could understand language. What has the discovery of MNs told us that we did not already know? Let's leave Chomsky till Chapters 8 and 9, because it would just (irrelevantly) muddy the waters here. We need a new word, a cognate for "weasel" word (which is the empty shell of a word with no content), but a word that refers to the empty cell of a non-existent explanation that nevertheless gives us the impression that it explains something.
Delete[There will be -- in Chapter 9, on Chomsky -- a discussion of whether Chomsky's distinction between the syntactic constraints on language (Universal Grammar) and the (shall we call them) "cognitive" constraints on thought, might tell us something about why ChatGPT (because of the constraints on the structure of its unimaginable Big Gulp) has managed to internalize enough from all that stuff to help prevent GPT from producing much more nonsense even though it is incapable of thinking, understanding or meaning a thing. (But this has nothing to do with MNs or mirror capacities.]
Although we may not completely understand how mirror neurons work, could we play with ways of activating them with potential applications to domains such as disease treatment. For example there are many diseases such as delusions that are highly related to issues in processing of social context situations and group interactions. If mirror neurons are more responsive the more familiar the subject looks, then would mirror neurons be activated with extra strength when seeing a VR representation of oneself, and possibly help with brain plasticity in training the motor areas of the brain that be be degenerating? Or if a patient was presented with an avatar that looks like themselves, and the avatar exhibited a certain emotion like joy or trust towards another, would this help the patient experience such emotions themselves, or could this help the patient perform better in an area of their lives (the way that mental rehearsal helps athletes prepare for a performance)?
ReplyDeleteStimulate people's MNs when we don't even know what MNs do? (And I thought Rizzolatti was getting a little too carried away with MNs!)
DeleteWhy VR representations of oneself? Wouldn't a mirror do? Why not train directly with a smile aerobics video, or a full cheering calisthenics video? (This was lightly mentioned in some of the other skywritings.)
By reading weekly so far, I realized some of the points and required techniques that cognitive science tries to solve. From this week's mirror neuron (MR) theory, closely integrated are social psychology, linguistics and neuroscience. The ultimate goal is reverse engineering in cognitive science, With the discovery of MR, it requires to study its bio-physical properties through research methods in the field of neuroscience, EEG, and observing the firing mode of neurons. Social psychology and linguistics play the role here of studying why this kind of neuron appears in the process of biological evolution, especially in monkeys and humans.Furthermore, realizing reverse engineering in cognitive science requires biology. Material preparation at the bio-physical level, psychology, linguistics decoding and trying to reproduce the human brain, artificial intelligence part to simulate the computation part, perhaps this is a solution to realize the Turing test at the T3-T4 level.
ReplyDeleteHowever, is it really possible for one to manage knowledge with diverse disciplines? Perhaps not all computation, then whether the next trending direction of cognitive science is to calculate the portion of each field and connect them together such that existing knowledge could be effectively linked.
Kid-Sib had trouble following this. Please read the other comments and replies about what MNs do and don't tech us.
Delete
DeleteMirror neurons are just a primary step. I think this may be the connection between consciousness and Darwinian evolution. Evolutionary psychology is not enough to solve all problems, especially the differences between individuals and the problem of configuration. Just knowing evolution and neural distribution provides limited help for us in reverse engineering. Therefore, in addition to considering biology, society's impact on people is also important when implementing reverse engineering. The impact also needs to be included
I have a quick question; in the introduction it says that Mirror neurons are found in the premotor cortex and the inferior parietal lobule. Yet, the Emotions and empathy part of this paper talks about the insula and never explicitly talks about the Mirror neurons. It solely states that the insula contains a “neural population” active both when one experiences an emotion and when observing the facial expression of that same emotions. Does that mean that the ‘neural population’ consist of Motor neurons, or that the neurons in it just have similar properties to MN?
ReplyDeleteWe need causal models for mirror capacities. We already have more correlations than we know what to do with.
DeleteThe part that appealed to me the most in the reading was the one explaining the relationship between mirror neurons and empathy. The data in the fMRI experiment has been able to demonstrate that the insula may contain a neural population active both when an individual directly experiences disgust and when this emotion is triggered by the observation of the facial expression of others. Similar data allowed us to make the link between the experience of felt pain and the observation of a painful situation in which was involved another person loved by an observer. It brought me back to our psychology introductory class. We studied a simulation system to understand facial expressions where the idea was that reacting to someone’s emotion generally implies a micro motor-somatosensory feedback that enables us to better understand its thoughts or emotions. We mimic others and these micro feedback allow us to activate similar emotions as their own. Since sensorimotor simulation is the core of empathy according to this model, I was wondering if a T3 would ever be able to process and reproduce emotions through the same process. Maybe T3 is enough for showing empathy, maybe we don't need a T4 with mirror neurons for that.
ReplyDeleteYes, empathy sounds like one of the many production/perception mirror capacities. But please read the other comments and especially my replies.
DeleteT3 is a test. A T3-passing robot like Anaïs passes the test. You can't be sure any thing other than yourself feels, but other people are close enough. And it you can't tell them from other people by what they can do, then that's enough too. We don't usually look into one another's brains to be sure they feel. But "Stevan Says" that feeling could involve some biochemistry too; so it may turn out that without at least that much of T4 , it's not possible to pass T3.
Yes, T3 requires mirror capacity, but it's not clear whether (or why) it would require MNs -- rather than just the means to do whatever it is that MNs do (if they do anything)!
This sentence in the article impresses me: "Humans mostly communicate by sounds. Sound-based languages, however, do not represent the only natural way for communicating. Languages based on gestures (signed languages) represent another form of complex, fully-structured communication system. "(Rizzolatti & Destro 4)
ReplyDeleteI have my own opinion on this statement. This reminds me of a question I occasionally read in a book in high school, "If you had to choose again, would you rather be a blind person or a deaf person?" I remember I asked many friends and even my parents, they all chose to be deaf, which is the same as my choice, because it is a very scary thing when you can't see anything.
In life, if there is no vision and no pictures, the horror is far greater than not being able to hear. In fact, many times, I think people communicate more through their vision. There are many strange things in life, and many explorers go there to see and experience them instead of listening to them or from other people. In addition, the frequency range of human hearing is only 20Hz-20kHz. Compared to other animals, we can actually hear very, very little.
Seeing is a richer sense. And you can imitate more things in the world by gesture than you can by sound. That's why it sounds like a more likely way for language to first start. (This is not about a normal child today, with a speech-prepared brain: It's about the transition from before language evolved, to after. Please see the other commentaries and replies.
Delete(But I'm not sure whether people are more distressed, in the short or long run, from losing seeing or losing hearing, or from being born without hearing or without seeing.)
DeleteI've heard of mirror neurons before, we've discussed them here on skywriting. this source was overall an informative survey through the functional role of this interesting phenomenon. but, I have a bit of an issue with the concluding statement that because of mirror neurons we have solved the essential question of " how does what is valid for the sender of a message become valid for the receiver" potentially I am being overly picky on wording, or maybe I've just missed something, but I believe this to be an overstatement.
ReplyDeleteThis is a valid objection, but it has come up in other replies. Please read the replies.
DeleteThe reading talks about how mirror areas enable one to understand why an agent performs a specific motor act. At 7 months of age, infants are able to sit independently and reach for objects. That is also around the same time they build an understanding of other people’s actions. 6-month-olds were shown repetitively a person reaching for a ball and looked longer when the person then reached for a doll, meaning that it went against their expectations. This happened even when the position of the objects was reversed, so they knew the actions were object-directed. Therefore, this provides an argument that AI needs to be physically able to interact with its environment to see why an agent does certain actions.
ReplyDeleteBy "AI" do you mean a T3 robot? Or just a computer (computationalism; T2). It makes a lot of difference in this course.
DeleteI'm well familliar with mirror neurons, which many science communicators (which played a large part in my decision to study Cognitive Science) I enjoy have discussed them.
ReplyDeleteSomething that came to mind is a possible link between imagination, other's minds, and mirror neurons. The fact that they fire while one either executes or sees an other perform an action reminded me of a paper I stumbled upon (Isaac, A. R. (1992). Mental practice: Does it work in the field? The Sport Psychologist, 6, 192–198.) where I learnt that imagining an action and performing it has very similar activity in the brain, and my thought was: is this monkey-see-monkey-neuron-do first and foremost imagining oneself in the other's position?
This could have interesting consequences in the evolution of empathy, which some monkey species seem to lack, although mirror neurons could still make sense as a sort of "narcissistic projection", the monkey imagines itself as the other, which could in some places throughout the evolutionary line turn into true empathy.
Mind-reading capacity, empathy and altruism are not narcissistic in mammals and birds, who must feed and protect their young if they are to survive and reproduce the parents' genes; so although Dawkins has called genes themselves fundamentally "selfish" rather than altruistic (Week 7), parental interest in the wellbeing of their young (a genetic trait) is not narcissistic (and it's not even self-interested, except in the same sense that feeding and protecting oneself is self-interested).
DeleteI’ve always been interested in the physiological aspects, especially the neurocircuitry, that underlie certain abilities observed in living organisms;“Mirror Neurons” by Dr. Rizzolatti and Dr. Destro covers a fascinating example of this, by relating the mirror neuron system in the brain to crucial brain functions like the purpose behind an action or even empathizing with others. More specifically, many experiments mentioned in this article made correlational links between the mirror neuron system and a plethora of these different functions— for example, the authors used functional magnetic resonance imaging to track mirror neuron activity; increased activity in these specific cells were observed when context (which could provide insights on intention) was added to a hand movement. Although it has been established that there exists some sort of relationship between this function and these neurons, it is still not clear to me how this works. I’m curious as to whether current research has been delving into these results on a deeper-level— what exactly are the details underlying this increased activity? The usage of the word “activity” in itself in this article also seemed too vague and broad to me— what exactly is meant by “activity”? What exact processes are cells executing in these circumstances?
ReplyDeleteAs well, although anatomy is usually thought of in objective terms, I’m wondering if mirror neuron systems could vary depending on certain factors/influences as well. For example, if organisms have different capabilities of understanding and varying levels of empathy, would they have slightly different mirror neuron systems? In what ways would they differ, both in terms of physiology and the processes they undergo? What are some factors (ex: genetics, species type, living conditions) that may influence mirror neurons?
Please read the other comments and replies: No one knows yet how MNs produce mittor capacities (if they do).
DeleteAs I was reading this article, this part of the article caught my attention: “It is generally assumed that the primary deficit in intention understanding found in autistic children, is due to damage of the mirror system as the system responsible for understanding the actions of others. However, one may wonder whether the primary deficit in autism lies indeed in the incapacity to understand others or rather in a more basic deficits in the organization of the motor chains. In other words, the fundamental deficit in autistic children resides in the incapacity to organize their own intentional motor behavior”. After reading this, it made me question about psychopaths. Does this infer that we conclude that psychopaths' empathy deficiency may also result from damage to the mirror neuron system?? So I did the research myself and found a research paper on the relation between mirror neurons and empathy-related regions in psychopathy. In the paper, the scientist discovers that people exhibiting high traits of psychopathy have alterations in these mirror neuron structures, which results in a reduced ability to empathize with others’ distress. This made me question once more, can the mirror neuron system be damaged during a child’s growth because of environmental issues? For example, growing up with parents that have hypoactivity of mirror neuron system, thus lacking empathy. Could these environmental issues affect the child’s mirror neuron system?
ReplyDeleteI recently read a paper that analyzes whether childhood maltreatment, such as emotional abuse and neglect, is associated or correlated with differential activation of the mirror neuron system (i.e., IFG, insula, cingulate gyrus, etc.). The extensive findings of the study demonstrate that sexual abuse and emotional maltreatement during childhood development may lead to the delayed development and recruitment of the mirror neuron system. More specifically, the level of emotional and sexual abuse, as well as emotional neglect, were associated with an increased activation of the left IFG during emotional mirroring and understanding. In summary, individuals who experience emotional maltreatment or sexual abuse in their childhood show an altered and/or delayed mirroring of other people’s emotions. So, to your question, it does seem that environmental factors can affect a child’s mirror neuron system.
DeletePlease read the other comments and replies on this. It is not yet known how MNs produce M-capacities. But it is already known that early abuse can produce psychopathy, wether or not the effect is detectable in MN activity.
DeleteIn this week's reading, I found 'imitation' as a core mechanism of the mirror neurons to be particularly interesting. To my understanding, when one observes an action being performed by another individual, this triggers the activation of the mirror neurons to replicate the observed action (by means of mapping it onto its motor system). The replication aspect is as if it was a secondhand experience that was taking place internally (the observer's brain), in the absence of the actual external experience. The system also recognizes the underlying intentions of actions, aiding in the comprehension of why actions are being carried out. The intricate connection between imitation, intention, and empathy is fascinating as we have a mental replica, via mirror neurons, of emotions displayed by another person, fostering a sense of empathy that could not be felt by purely mimicking the movements of another person (eg. Smiling).
ReplyDeleteAlways read the other commentaries and replies, so you don't just say things others have said.
DeleteMirror neurons are a very interesting phenomenon I have encountered many times across the years. They have been implicated in many important cognitive functions such as empathy, imitation, and motor learning. Mirror neurons act as a liaison between visual input and motor output, and potentially underlie a variety of functions. New to me is their role in language development, especially in signed languages, which is logical with their strong basis in motor operations (gestures), but I would also believe to be relevant in aspects such as articulation or lip-reading in spoken language that lead to the development of human communication.
ReplyDeleteWith respect to cognition and its relation to the Turing Test, we have often spoken of what mechanisms we need to emulate cognition, it would be interesting to see how we would go about establishing a version of this and what its impact would be on the function of the machine.
Mirror capacity is part of T3. Always read the other commentaries and replies, so you don't just say things others have said.
Delete
ReplyDeleteWhile reading this text, I particularly found the discussion of mirror neuron systems in individuals with autism very interesting. When the discussion around emotion and empathy was brought up, I immediately thought about psychopaths. If psychopaths are unable to feel certain emotions and therefore are unable to relate to how another person may experience these emotions, would that be similar to those with autism and could you say that they would also have deficits in their mirror neuron system?
Mirror capacity is part of T3. Always read the other commentaries and replies, so you don't just say things others have said. Comment on what others have said; add something.
DeleteAfter reading through the article on mirror neurons, I was fascinated by the fact that this finding could be one of the keys that may be able to explain humans evolutionary superiority over not only our close ape relatives, but over many other animals that lack the capacity for emotion, understanding, imitation, and language development. I quickly researched if humans are the only animals that are thought to have mirror neurons and to my surprise, mirror neurons have only been found in humans, macaques, and songbirds. This however is followed by the belief that the only reason we have not found mirror neurons in other species is simply due to a lack of looking. This in itself fuelled my question on how many more animals possess these types of neurons that we are not even aware of, and of those species, how many may be able to perceive action and intention understanding that we do not know. Surely the incorporation of these mirror neurons can increase an animal’s ability to possess emotional intelligence and with that, a higher state of consciousness. Can consciousness then be found in many more species that we even thought possible? And if so, how else could we be able to conduct experiments to either prove or disprove this.
ReplyDeleteGood points. But what do you mean "higher states of consciousness"?
DeleteI also did some googling after the readings Stefan, and found similar information on the topic of mirror neurons. It is likely, I assume, that different animals other than the ones we have discovered do also have mirror neurons. What I am unsure of though, regarding your reply, is how the discovery of mirror neurons in new animals relates to if 'consciousness then [can] be found in many more species.' Neuroscience defines mirror neurons as, 'a mechanism whereby we experience empathy and recognize the intentions of others by observing their behavior and automatically matching their brain activity,' however it does not grant the animal a 'higher state of consciousness'. I'd argue consciousness is consciousness regardless of how you interpret the actions or emotions of others.
DeleteNicolas, I understood mirror neurons to merely be an observed category of neurons that have the specific behaviour of firing both when an action or perception is established by oneself or by another person/animal. Do you think it requires "feeling" empathy? One might imagine a feelingless robot that identifies these commonalities between its own sensorimotor perceptions and others', and fires a signal in either case!
DeleteAsk yourself how could the discovery that there is a correlation between activity in MN areas of the brain and the exercise of capacities we all knew we had tell us anything about how the brain does anything if no explanation of the causal mechanism that produces the capacity is given? This is the reason the Fodor reading was paired with this one. What have you learned from this?
ReplyDeleteIt seems to me to be a recurring theme to confuse the "functional" and "feeling" meanings of different words, like "understanding" or "empathy". It seems clear that mirror neurons would exist for the purpose of computational empathy, ie algorithmically predicting what other people feel and what those feelings imply about their future actions. That is, in fact, one purpose of mirror neurons discussed at length within the scholarpedia article. It is also, clearly, relevant to inclusive genetic fitness, and is present, we can infer, in most animals with non-trivial neuronal architectures, since it is a central use of sensorimotor capabilities.
ReplyDeleteWhat relevance to the conversation, then, does phenomenological empathy hold? It is, under our current scientific understanding, a subjective phenomenon. The inference of others' intentions, imitation, and the conjecture that language comes from these interactions, is fascinating; but even for philosophical zombies, ie animals that have our brain's computational structure without its feelings (if those are possible), it is clear that empathy would still be a meaningful concept, insofar as we treat it computationally!
My apologies for the late writing professor. Reading this paper made me feel inspired about all the avenues that this research has and will surely continue to open up. Even if continuing to pursue research from this empirical and/or neuroscientific approach may not get us any closer the "how" (qualitative) functionality of MN's, the horizon of applications are exciting to me. For example, recently there has been a lot of anecdotal evidence of athletes using virtual reality technology as a successful addition to their training protocol. From my understanding, they use it to aid in visualization training, which in conjunction with highly realistic VR simulations, brings up interesting questions around the necessary stimulus requirements of MN's. Ultimately, even if we cannot jump the gap between "what" to "how" with these approaches to MN research, perhaps as applications and further research unfolds we will gain novel ways to ask questions that bring us closer to the "how".
ReplyDeleteIn the section about intention understanding wouldn't the use of mirror neuron only apply to a very simple or primitive set of actions? Can mirror neuron activation explain our understanding of larger-scale actions? For example, instead of just grabbing an apple, let's say a monkey has to go through a little obstacle race to eat an apple. At what point would a monkey that sees this happening realize that the monkey inside the race is aiming for the apple? And would that understanding be derive from the activity of mirror neurons?
ReplyDelete**BLOGGER BUG**: ONCE THE NUMBER OF COMMENTS REACHES 200 OR MORE {see the count, at the beginning of the commentaries] YOU CAN STILL MAKE COMMENTS, BUT TO SEE YOUR COMMENT AFTER YOU HAVE PUBLISHED IT YOU NEED TO SCROLL DOWN TO ALMOST THE BOTTOM OF THE PAGE and click: “Load more…”
ReplyDelete________________
Load more…
________________
——
After 200 has been exceeded EVERYONE has to scroll down and click “Load more” each time they want to see all the posts (not just the first 200), and they also have to do that whenever they want to add another comment or reply after 200 has been exceeded.
If you post your comment really late, I won’t see it, and you have to email me the link so I can find it. Copy/Paste it from the top of your published comment, as it appears right after your name, just as you do when you email me your full set of copy-pasted commentaries before the mid-term and before the final.
——
WEEK 5: Week 5 is an important week and topic. There is only one topic thread, but please read at least two of the readings, and do at least two skies. I hope Week 5 will be the only week in which we have the 200+ overflow problem, because there are twice the usual number of commentaries: 88 skies + 88 skies + my 176 replies = 352!. In every other week it’s 2 separate topic threads, each with 88 skies plus my 88 replies (plus room for a few follow-ups when I ask questions.
I really enjoyed this reading because i remember taking this in one of my courses from last semester! It’s so fascinating how mirror neurons have been posited to have functional roles, and influence the way that we behave on a social level. I never would have imagined that basic things such as action understanding, imitation, intention understanding, empathy and even language acquisition are based on intricate neural mechanisms that literally mirror the neural mechanisms associated with those functions or actions!! I actually would not have thought that for action understanding and imitation, the visual system is alone is insufficient. It makes sense, I mean if I do not engage the motor system, where do I map out the visual input. I think the concept that intrigued me within intention understanding is the research on ASD children that suggests a dysfunctional mirror neuron system leads to their deficits in understanding why and how people around them act. It was interesting that Rizzolati and Destro proposed the possibility that the deficit may also rely on an insufficent engagement of the motor cortex. Its so fascinating how even though our brain compartmentalizes individual components to respond to certain input, they interact with each other to result in a larger, intentional action.
ReplyDeleteThis week’s reading on mirror neurons and especially the section on action understanding made me think of last week’s reading critique of the CRA by professor Harnad. As discussed in the reading, the mere visual perception of an action does not equal an adequate understanding of the intrinsic components of the observed action. Therefore, the activation of the mirror circuit is essential to provide the necessary experiential comprehension that humans are capable of in understanding the action of others. In thinking about this with the CRA by Searle, I came to realize that the mirror neuron system in its role as providing action understanding as well as intention understanding is something that Searle would have considered as necessary for a computer machine to be “intelligent” (as in having intentional mental states and actually having semantics (meanings).
ReplyDelete*Repost, first comment deleted*
ReplyDeleteBy reading some of the previous responses, I understand that we will examine language more in depth in later weeks, and that at the moment it's crucial to understand that imitation is not language, although imitation does play a role. I nonetheless found the following quote fascinating, particularly its implications on how we might understand language and what might make it that we can 'feel' that we understand language, as mentioned in Searle's text and in class- "thanks to the mirror mechanism, actions done by one individual become messages that are understood by an observer without any cognitive mediation". Crucially, it's stated that understanding happens without understanding. I do think that for both gestures and language, there is most often a conscious element. We can point to what we think a gesture means. Then again, the point here is that it's the actual imitation itself that does not necessarily require understanding. For example, in conversations we often gesture without being aware of it.
*Edit*
Before my comment was deleted, Professor Harnad pointed out that gestures in speech are not gestural language. I understood this to be because they do not aim to convey any meaning on their own, instead just being accessories to the conversation such as to add more emotion.
Mirror neurons are so interesting. One of the first things I did after reading the web page was check if dogs have mirror neurons, and they do. I wanted to check because there are differences between the way different animals interact with us. So then, I checked whether goldfish have mirror neurons. I was expecting the answer to be no but, despite getting answers only from questionable sources, the answer apparently is yes. I was assuming based off of my other assumption: because most fish that can fit in a small fish bowl don’t really connect with their owners, feel their emotions, and comfort them (which is the opposite for dogs and other bigger mammals such as dogs). Mirror neurons work in different capacities in different animals, is the conclusion I came to. Open to further discussion :)
ReplyDeleteI have heard and read about mirror neurons in previous courses, but I had never really considered the widespread functions that mirror neurons could mediate. This paper really opened my eyes to how important mirror neurons are to several of our cognitive capacities, including empathy, intention understanding, and especially the evolution of language. The most significant importance of mirror neurons, however, continues to be the fact that it provides an observer with an experiential understanding of observed actions by demonstrating to the observer the intrinsic components of the observed action. From this, mirror neurons enable us to learn a new motor action by replicating an observed action now present in the motor ability of the observer. Thus, mirror neurons are important for skill learning.
ReplyDeleteBased on our lecture and readings on categorization from the week of October 16, mirror neurons (i.e., simply observing and imitating the behaviors of others) are not enough for category learning. However, I believe that they are intimately connected. Category learning is a matter of leveraging the motor skills we have learned from our mirror neurons, applying them to interact with the environment around us, and obtaining feedback (physical, social, etc.) from those interactions in order to learn to do the right thing with the right kind of thing. Mirror neurons are closely related to categorization because they allow us to learn the motor skills needed for the “systematic differential interaction between an autonomous, adaptive sensorimotor system and its world” necessary for category learning.
Mirror neurons are neurons that activate upon viewing another person execute a motor movement towards an object alongside when oneself preforms a motor movement. In the article we see that one of the possible reasons that we have these neurons is for empathy, language, and imitation—among other things. I’ve recently been on a Girard deep dive and find the implications of mirror neurons within Girard’s memetic desire curious. Memetic desire, simply, is that we want what we want because others also want it. Could it be that these neurons provide some biological backing for Girard’s theory?
ReplyDeleteMirror neurons help us understand other people, their intentions and feel what they’re feeling. Mirror neurons could be part of the biological machinery that helps us tune into and recognize others desire.
Though this does not explain why mirror neurons do what they do, or why we may desire what others desire.
I am skeptic about what is it in language that is actually activating mirror neurons. The reading suggests that it is the meaning of the language that is valid from the sender becomes valid for the receiver. However, I was reminded of a performance art in which the artist was talking to the audiences using the language created by herself. The audiences were nevertheless able to tell what the artist is doing by her tone: whether she is giving a speech or just casually chatting. There are more and more experimental art that is using completely made-up words but still able to express the artists’ emotions. This makes me wonder what is activating mirror neurons during communication might actually be the act of communicating instead of the meaning of the language. The receiver’s mirror neuron is activated when they understand that the sender of the message is trying to communicate. What the mirror neuron shows is only the understanding of the action, but not the content. The comprehension of the content might be happening elsewhere. Other skywritings have pointed out that we can use people who are having trouble in communication (such as those who are suffering from autism or aphasia) as subjects in the research of MNs and language. I think a more ethical solution could be to present a healthy subject with language made of made-up words and observe the response from MNs.
ReplyDeleteI learned about Rizzolatti’s study before, but I never thought about mirror neurons’ function on language evolution and emotion. Rather than mimicking behaviors, the function of mirror neurons is more about how to interact with the external stimuli. Based on my understanding, it plays the role of a programmer from input to output. To be specific, it maps the pictorial description of actions carried out in the higher order visual areas [the inputs] onto their motor counterpart [the outputs].
ReplyDeleteThe article Robot learning by active imitation emphasizes the importance of developing social robots with learning capabilities to adapt to unknown environments and perform new tasks. It highlights the limitations of task-specific learning algorithms and proposes a solution for real-time responses through active imitation. Active imitation involves the robot observing a human performing a task and then imitating the task in real-time. This approach allows the robot to learn new tasks quickly and efficiently, without the need for pre-collected training samples or a predefined representation of the task. Real-world would-be recognizing diving signals and human poses.
ReplyDeleteIn the article Mirror neurons 30 years later: implications and applications, found the part about the practical application of mirror neuron research in neurorehabilitation very interesting. Mirror therapy involves observing and imitating movements and can be used to improve motor function in patients with neurological diseases or orthopedic injuries. Action observation can increase the cortical excitability of motor brain areas, leading to plastic processes that facilitate the subsequent execution of relevant daily actions. It is suggested that mirror neuron research may ultimately enable the refinement and personalization of neurorehabilitative interventions in various clinical populations.
ReplyDeleteThe reading Mirror neurons 30 years later: implications and applications describes the history of mirror neurons from when they were first discovered as just neurons that fire at the sight of motor movement and at production of that same motor movement, to now. Mirror neurons’ function has been extended to emotion, spatial location, decisions, direction of attention, beliefs etc, therefore playing a significant role in social interactions and cohesion within a community. In the context of this class, knowing that MN activate in response to motion and emotion can be used to understand the development of language. If seeing an individual producing a sound activates MN in your brain, then it can be hypothesized that MN are involved in comprehension. This can work in the case of speech sounds, but considering that vocal language likely evolved from gestural language, this mirror theory is even more likely.
ReplyDeleteI am struck by the profound implications of mirror neurons for our understanding of empathy and the evolution of communication. These neurons do more than just mimic observed actions, they seem to be integral to how we decipher intentions and emotions. This understanding could potentially help us understand the pathways through which our ancestors transitioned from simple mimicry to complex language. It's a compelling narrative, but one that requires us to decipher much cognitive complexity to truly grasp how these neural reflections contribute to our uniquely human social interaction and linguistic ability. It's become clear that understanding mirror neurons is not just about seeing a reflection.
ReplyDeleteI think the past and future MN debate in the reading “Mirror Neurons 30 years later: implications and Applications” to be quite interesting. It delves into nature vs nurture, depicting the interplay between genetic predispositions and environmental influences. Neonatal imitation’s ambiguity reflects the challenges of clarifying innate mechanisms from early ‘experiences.’ Comparing audiovocal MNs in birds and mammals hints at a shared genetic foundation and raises questions about universal cognitive capacities across species. The link between monkey genetic variant and altered MN emphasizes the nuanced relationship between genetics, neural plasticity and functional arcitecture–highling how subtle genetic variations shape social and cognitive behaviors. The debate not only explores action observation but shows the complexity of genes and cognition.
ReplyDelete