I once gave a (perfectly awful) cognitive science lecture at a major centre for brain imaging research. The main project there, as best I could tell, was to provide subjects with some or other experimental tasks to do and take pictures of their brains while they did them. The lecture was followed by the usual mildly boozy dinner, over which professional inhibitions relaxed a bit. I kept asking, as politely as I could manage, how the neuroscientists decided which experimental tasks it would be interesting to make brain maps for. I kept getting the impression that they didn’t much care. Their idea was apparently that experimental data are, ipso facto, a good thing; and that experimental data about when and where the brain lights up are, ipso facto, a better thing than most. I guess I must have been unsubtle in pressing my question because, at a pause in the conversation, one of my hosts rounded on me. ‘You think we’re wasting our time, don’t you?’ he asked. I admit, I didn’t know quite what to say. I’ve been wondering about it ever since.
See also:
Grill-Spector, K., & Weiner, K. S. (2014). The functional architecture of the ventral temporal cortex and its role in categorization. Nature Reviews Neuroscience, 15(8), 536-548.
ABSTRACT: Visual categorization is thought to occur in the human ventral temporal cortex (VTC), but how this categorization is achieved is still largely unknown. In this Review, we consider the computations and representations that are necessary for categorization and examine how the microanatomical and macroanatomical layout of the VTC might optimize them to achieve rapid and flexible visual categorization. We propose that efficient categorization is achieved by organizing representations in a nested spatial hierarchy in the VTC. This spatial hierarchy serves as a neural infrastructure for the representational hierarchy of visual information in the VTC and thereby enables flexible access to category information at several levels of abstraction.
Hmm.. I’m not sure how much I took from Fodor’s piece. The general goal behind these readings is to explore the relevance (or lack thereof) of neuroscience in cognition. Here, Fodor questions the obsessions with localizing different functions to parts of the brain. Reminiscent of Turing, Fodor questions why, if your engine is working and functioning, is it important to know where its carburrettor is? Similarly, if a machine seems to be cognizing and we have no reason to doubt that based on its behaviour, why must we question its structure and location of said structure? He also articulates that materialism hinges on the acceptance that mental states have neurological correlates. If localization studies are proposed necessary to elucidate mental states, then one moves away from materialism and into dualism. This made me think of Searle’s argument that Strong AI (computationalism) essentially is a hidden dualism because by suggesting implementation independence (i.e. the hardware doesn’t matter, only the program does), there is the notion that these two are separable. What do other people make of this discrepancy? Overall this week’s readings were interesting but I am not convinced one way or the other. At times I am a proponent that neuroscience is irrelevant to cognition (it is arbitrary as long as the function is the same), but in light of Spector and Weiner’s piece (2014), it is hard for me to conceptualize the visual system’s categorization ability without its specific hierarchy.
ReplyDeleteHi Jessica!
DeleteYou raised some of the exact points that I was thinking of. When Fodor makes his engine analogy, he's saying that knowing where the carburettor is in the engine gives us no additional information that can serve towards our understanding of the engine itself. In my opinion, however, if there's damage to the engine, understanding what the carburettor is responsible for, where it is and what it does can help us understand what we need to do to fix the engine. I think that this concept is important in cognition, particularly when talking about reverse engineering the brain. I think that beginning with a basic understanding of structure and location can help us understand how different parts of that structure interact, and function.
Jessica, the gist of Fodor's paper is that he doubts that neuroscience can reveal the causal mechanism for what we are able to do. Do you think Spector & Winer (2014) explains how we categorize? What is their explanation? And would it help pass T3 for vision?
DeleteKristina, Fodor does not doubt that studying the brain helps if you want to fix the brain, when something goes wrong. What he is doubting is whether it can show us how it does what it does when everything is going right.
It may not be the central question of this article, but I felt like it had to be mentioned: consciousness is not a concept that we understand half as much as other brain functions.
ReplyDelete“I admit, for the sake of the argument, that consciousness is correlated with certain neurons firing at 40 Hz cycles; and that some bits of the brain light up when we hear nouns but not when we hear verbs”
brain immaging has shown how specific tasks (actions, words said or heard,..) are linked to metabolic activity in a specific more or less consistant part of the brain. But consciousness is still for the most part a mystery. Some patients with complete hemispherectomy still have consciousness, which proves that no specific localized bundle of neurons allow consciousness. The study of split brain patients also teaches us about how consciousness can't be located in either hemisphere nor both. In any case, the concepts of "self", "mind", "consciousness", etc, cannot be put in the same boat as motor actions when it comes to their brain localisation. Fodor said "I’m a philosopher, not a neuroscientist, so perhaps I’ve badly missed the point.", he questions the necessity to understand localization of brain functions, but there seems to be a clear correlation between how much we understand a function and our knowledge of the localization of that function.
Fodor’s argument is interesting with respect to current class discussions of implementation independence in computationalism, although I found Fodor’s proposal (that pinpointing loci of mental processes in the brain is a superfluous task) weak in its execution. I agree with Flinois’ point that our understanding of a function is proportional to our knowledge of the localization of that function, especially when we uncover the neural circuitry and connections that exist between different brain regions. For instance, our near-complete understanding of language function is due in part to the fact that scientists have mapped out the different brain regions involved in language understanding and execution, as well as from different lesion studies. Conduction Aphasia, or the inability to verbally repeat linguistic structures, is explained by a disconnection between Broca’s area and Wernicke’s area, involved in verbal execution and understanding respectively. Localization of function, and the subsequent connections between these different loci, sets the foundation for understanding how the function is achieved, almost step-by-step.
DeleteIn regards to computationalism, where the goal is to emulate cognitive processes in the form of computer algorithms, implementation-independence may hold, so long as we are only concerned with the output of the algorithm, while physical details of the hardware are irrelevant (as is with the Turing Test). However, in neuroscience, where the goal is the pursuit of complete understanding, uncovering the neural circuitry that is responsible for a cognitive process is essential in enriching our understanding of that very function.
Julia, I don't think Fodor's points were about consciousness (the hard problem) but about cognitive capacity (the easy problem)
DeleteJaime, "our near-complete understanding of language function"? You mean you think we're able to explain how the brain passes T2? Fodor is talking about the causal explanation of our language capacity, not the clinical explanation of aphasias. Fodor is a computationalist. But do we even have any causal explanation, whether computational or dynamic or hybrid, of how the brain produces language capacity (T2)? And can studying the brain lead us to such an explanation, the way studying the heart has led us to an explanation of how the heart pumps blood?
I thought Fodor's reply to Benjamin Bly illustrated the gist of his article nicely. of his. Seemingly the next two statements of Fodor's contradict each other.
ReplyDelete"given that it matters... mental functions have characteristic places in the brain, why should it matter... where those places are?"
"...the way the structure of our minds depends on the structure of our brains. Nobody even knows which brain structures our cognitive capacities depend on."
Given the latter statement, is it assumed structure and neural imaging are relevant after all? The difference between the latter statement and a brain image showing activity from thoughts of tea cups is what is the hypotheses that they are testing. The difference being one is a formulated hypothesis that can be tested and the other is a scientist with a camera. Fodor's despair seems to come more from a lack of hypotheses and direction, especially bigger picture hypotheses that connect with theories of the mind, rather than acknowledging structure/ function is important. I think most neuroscientists would agree that the aim of neural imaging is not phrenologist in nature ie. capturing the parts of the brain active when thinking of teacups, baseball, lampshades etc. Rather, like patient H.M., the aim should be isolating areas with known function as this is especially helpful in the medical field. I argue that teacups are not important but isolating specific systems such a vision and then figuring out which other systems they connect to is. For instance, if a structure was associated with a specific behavior or emotion, it could be discovered to be smaller in patients with say psychopathic tendencies, and biologists could isolate the gene which results in this. Multidisciplinary research can result in micro to macro scale connections of the brain. Unless Fodor is a dualist, how can understanding the brain not lead to hypotheses about the mind?
I definitely agree that Fodor’s issue with brain imaging has more to do with hypothesis-formation and a lack of theory-driven research. It looks like he’s taking issue with researchers searching for answers for unformulated or ill-formulated questions. In addition to this and putting aside the clinical usefulness of brain imaging, however, I think Fodor is also driving at another point: that answering the question of where things happen cannot actually help solve the question of why and how we do what we do. If this is his argument over and above the way research is conducted, then I think he does make a defensible point in the context of current research. But I don’t see why he would condemn the future of brain localization research as well.
DeleteKathryn, Fodor is asking whether learning where and when things happen in the brain that are correlated with what we do and can do can tell us how what is happening gives us the capacity to do what we do and can do. (Austin seems to agree.)
Delete
DeleteFodor is arguing that brain imaging techniques used to investigate correlations of behaviour and neural activity in the brain is a fruitless endeavour. He seems to insinuate that scientists are actually making circular progress with research. By circular research I mean scientist are continuously returning to disproven or un-popularized theories (for example phrenology). Another example of this circular research is that Fodor had a sense that research was being conducted to try and strongly disprove the notion dualism. Dualism it the theory that there is an entity outside of the brain that is responsible for consciousness and thought processes. Fodor claims that perhaps scientists are futilely trying to find physical locations in the brain that may correlate with neural processing to disprove this theory that has already been disregarded by the majority.
Fodor mentions that he may have missed the point to this research and I believe he has. Although there is some research that tries to answer unformulated or ill formulated questions, I think there is significant amount of other good research that uses brain imaging techniques to assist the medical field. For example the discovering of areas in the brain where language is localised or where memories are formed (I am using this description of localization of brain areas loosely), have helped scientists perform the most efficient surgeries on patients without causes unnecessary loss of function. This localisation research is definitely not futile and has helped treat patients with severe epilepsy or other forms of brain damage.
Nadia, what Fodor says here about dualism is irrelevant nonsense!
DeleteBut remember that clinical usefulness is not what's in question here.
As I mentioned in class, I can't help but think that where and when things happen in the brain (even if it only is correlated with what we can do) can only attempt to help answer how we do what we do. It would be foolish to study the mechanism of how a plane takes off without every studying the physical plane itself. No anatomy and neuro imaging studies haven't answered the 'how' yet but haven't they given us more information about the incredible thing that allows us to see, move and think? Isn't knowing something better than knowing nothing? And yes it would take an incredibly long time to study the brain and its many connections and intricacies to figure out the how or at least a hypothesis. But what is more likely reverse engineering a human in the form of a T3 robot which is able to see, taste, smell, hear, move, think, work with acetyl choline (mentioned in class) and pass for a human for its entire life span or to simply study the brain itself with the help of a giant or two and better imaging methods and studies? I think it's always dangerous to compartmentalize studying something- a good understanding of the heart comes from anatomy, microbiology, physiology, and some physics. Just one discipline would not provide a complete understanding. By no means do reverse engineers have to create robots with the information specified by neuroimaging studies but I do believe studying the brain itself can only help the field of cognitive science with additions to psychology, medicine, neuroscience etc along the way.
DeleteIn Fodor’s dismissal of studying the brain being a useful endeavor, he fails to see that brain scans are not merely used to answer questions about the mind and cognition. They are used as medical tools that can help in treating neurological illnesses, can be useful in brain surgeries etc. Although localization leads to a detailed map of correlations, brain scientists are not claiming that these are the key or path to a causal explanation of cognition. Although a lot of science has been filled with ‘science for science’s sake’ as illustrated in the story of the girl who meets Pavlov, many brain scientists work towards learning about and modeling the brain to get a better understanding of a healthy and well-functioning brain areas vs. brain areas that can predicts the development of future problems, or brain areas that are indicative of health problems. Just as hundreds of years ago, people examined and dissected the body to see what was going on in it, what was wrong, and what could be fixed, people have only recently developed tools to observe the brain more deeply. Therefore, I feel Fodor is going too far in suggesting that the brain technology we are currently using is waste of time.
ReplyDeleteI think that Fodor is right when he says that it is useless to do research just for the sake of creating more studies, research that does not have strong hypothesis it is trying to test, and research that is only about mapping areas of the brain (i.e. modern phrenology). Perhaps certain fields like social psychology would not benefit from brain imagery as much as a field like neuropsychiatry, since the former involves observing people’s behaviors and reactions, while the latter may have more emphasis on chemical and neural imbalances in the brain. I know this is a vague generalization about both the fields but any thoughts about how brain scans may be more useful for some fields/sub-field/questions and may be a waste of time/money in others?
I think Fodor fails to recognize that brain imagery is not just about localization, even though that is an aspect of it. In cases, such as studying split-brain patients, brain imagery provides data that would be impossible to make sense of by merely observing behavior. Yet I also understand to some extent why he is so skeptical about brain imagery, since every brain is not the same, and so we can never produce an overarching model brain. Furthermore, the expense and hassle of using a brain scanner questions whether it is even worth trying to use such a technology to diagnose patients and study the mind.
Nimra, Fodor is not doubting the clinical usefulness of neuroimaging (or of neuroscience in general); he is doubting its capacity to explain our cognitive capacity (the "easy" problem).
DeleteNimra, I agree that Fodor fails to mention the usefulness of neuroimaging in clinical research in his diary. However, the point he was trying to make is highly relevant in a world which puts its complete faith in the totalizing meta-narrative provided by neuroscience and neuroimaging. Look at any newspaper's science section and you will read bogus like scientists have discovered the secret of happiness or something like scientists have uncovered the secret to a successful life. This is clearly not the case even with something very specific like the ventral temporal location's in categorization. Fodor is just questioning the complete acceptance of the non-dual narrative that pin-points every thought to a spatial location in the brain. There is an ethical dimension to this debate too, which is the problem of concluding evidence based on neural correlates. Say that neuroscientists come up with an imaging technique to detect lies and the results of this technique could become legal evidence. This would have huge implications to society and morality. What would be useful however would be to regulate neuroimaging to certain avenues of research as u suggested, so that it does not pose to be an answer to how the mind works.
DeleteWhile Fodor’s argument that finding out where things happen in the brain doesn’t inform the functional questions of why and how do we do what we do, I think it’s a bit premature to brush off the question of where things happen in the brain. It could turn out that brain localization does much more than simply “[knowing] it scientifically”. It’s not that brain imaging is metaphysically privileged, but that it’s one of the few ways we can look at the organ which we presume to be responsible for why and how we do what we do. Looking at where things happen in the brain can eventually help answer the questions of how “the way the structure of our minds depends on the structure of our brains” or explain “the basis of cognition when it depends on a subtle interplay among brain regions”.
ReplyDeleteI agree with you, Austin. By using brain imaging techniques, scientists and researchers are collecting data. This data may not answer any immediate questions, but it can always be used in the future. It may be that this data collected will help formulate the questions needed to be answered.
DeleteAustin, you seem to think we can't get from when/where to how/why, and yet we can, after all. How? Can you give some examples?
DeleteLaura, of course it's useful (and necessary) to gather data. But the trouble is that data do not explain themselves. In the case of the heart, the suggest an explanation; in the case of the brain, it's not so clear that that they do, or can.
In principle, I tend to agree with Fodor that we can’t get from when/where to how/why because doing more of the same thing (perhaps with better technology as time passes) cannot generate causally explanatory results. But my opinion is that this argument applies to current research. I’m hesitant to condemn, as Fodor seems to do, the future of research on when/where things happen in the brain. But I must admit that my hesitation rests primarily on a pragmatic assumption that, in time, this stance is likely fallible. However, I want to give a shot at trying to think of how this might be. The best example I can think of is if researchers improve the methodology used in Benjamin Libet’s experiment on volition (which is itself a can of worms) and apply this to other mental or behavioural states. This may connect when/where things happen in the brain to how/why.
DeleteAustin, Libet was trying to find out about the timing of voluntary movement: Which comes first, the willing of the movement or the brain activity that causes the willing of the movement? (We're actually doing some work on that in my lab right now! Participate in the experiment if you are interested.) But how will that kind of thing (the timing of volition) help come up with a causal mechanism that can pass T3 (or T4)?
DeleteI was thinking how causal explanations are made in psychological studies. A longitudinal design often helps. The involvement of time made me think if the where/when of brain states can be connected causally to how/why we do things. For instance, substituting out volition for something like attention to certain objects, timing can help in saying whether brain area(s) X, Y, Z are involved in causing attention to certain objects. This one causal connection can be combined with other similar causal connections. If areas X, Y, Z are involved also in causing another mental process, then we might say these areas are involved in causing something in general underlying these processes. But this still doesn’t answer how and why. We can then look at what events cause areas X, Y, Z to change. Now we have a link between a group of events causally related to a group of psychological outcomes. Over and above confirming what we already know, the value of where/when in this case would be that it can identify general connections between events and psychological outcomes based on brain areas. These general connections can inform underlying factors at play over and above what we already know about what events cause what outcomes. Admittedly, however, this stance is no longer purely where/when as it involves interacting with the environment, which is why I still tend towards agreeing with Fodor despite my hesitation.
DeleteAlthough Fodor does make good points in arguing that knowing where 'thinking about a tea pot' would be a relevant way to spend research money, there are many important implications for investigating locations of mental processes.
ReplyDeleteFor example, it has be recently discovered that the brain regions for speech overlap with the brain regions for music, but not exclusively. This evidence can be used to help individuals who have suffered legions in their speech production loci by teaching them to sing their speech, and eventually allows them to recover their ability to speak. Without the advent of neural imaging we may never have come up with that solution.
It seems that Fodor is frustrated that cognitive science is spending it's resources on brain scanners when he is interested in larger questions, like the origin of consciousness. I would argue, however, that it is productive to take advantage of the resources and technologies we have, and explore to the extent we are able with those technologies. His statement 'difference between a scientist who has a hypothesis and one who only has a camera' seems to suggest that a scientist with a camera serves no function, but I would argue that a scientist with a camera has a higher likelihood of developing a hypothesis than one with nothing.
I agree; also, I think that while Fodor seems to have tried to pick as useless of an example as he could think of, you could even find some use in his teapot brain area. For example:
DeleteWoman walks into doctor's office: "Doctor, it's the strangest thing: my husband seems mostly ok, but he suddenly has no idea what the tea pot is."
Knowing this, the doctor could have an easy idea of what might be wrong.
Yes, this is a very silly example, which I doubt would ever happen, but the greater point is that the more we know about the brain, the better. The closer we can get to a COMPLETE understanding of the brain (though 'complete' is probably impossible), the better we will be able to help people with brain problems, as well as to further our quest for artificial intelligence.
I agree! I like Adrian's example on how useful neuroimaging is, in contributing to speech rehabilitation. Since neuroimaging is kinda a newly-developed method that enables us to see brain activity, I honestly think what we need is a bit of time to let the imaging technique develop more, and THEN draw the conclusion on whether it is really not helpful in knowing cognition. It could be almost useless in helping us to know "how we understand", but it could be helpful in bring us insights of what tasks are those tons of neurons involved in. This is still a way to reduce uncertainty though, right...?
DeleteAdrian, Dominique, Alison clnical versus cognitive (T2, T3, T4) questions again...
DeleteThe attractiveness of fMRI is that it is very likely to produce “statistically significant” results. If you pester a subject in a scanner with a specific task over and over again, it would be more surprising if a BOLD signal wasn’t measured in one particular area of the brain more than the rest. Whether or not these results are useful or not is probably unimportant to the graduate or undergraduate student trying to publish a paper. Because there is still so much to be learned about the brain, it seems scientists want to know everything about it, no matter how trivial.
ReplyDeleteThere are undoubtedly many beneficial uses of functional brain imaging, but the cost and time required for these experiments should make one consider the real implications of any possible findings. I think the use of these techniques in the clinical field are much more valuable, as they will only be implemented if they are found to be beneficial (through rigorous replication and analysis: physicians don’t like to be sued). This is in contrast to a study showing viewing a teapot “activates” this area of the brain, which only contributes to a rather large wealth of unusable literature.
I think you make a great point! Fodor focuses on the "teapot" and "nap" example to support his point that this type of research isn't leading us anywhere, but I think the way he offers his opinion discredits functional brain imaging and its use in the clinical setting. While a lot of money and resources go into fMRI and PET, it is not all for trivial reasons. It is also important to map out brain regions associated with language in order to minimize damage during surgery in a non-invasive way. Fodor does mention the usefulness of these technologies for surgeons by stating, "And, no doubt, if you’re a surgeon you may well wish to know which ones they are, since you will wish to avoid cutting them out," but this statement undermines the importance of knowing the specific region per specific case. While discovering brain regions that are important for recognizing trivial objects is unnecessary, I believe it is still extremely important to focus on brain regions that are responsible for functions like speech, learning, moving, and visual processing, for a neurosurgeon would not want to "cut these out" even by a millimetre.
DeleteColin, Annabel Searle concedes the point about clinical usefulness.
DeleteI think Fordor is right, to a certain extent. We know phrenology is wrong, but it still seems to linger in the way that we talk about the brain and do research. However, surely a detailed understanding of the parts is necessary if we want to look at the bigger picture? The question then seems to be about the relationship between parts and wholes. To what extent does knowing one help you understand the other? It seems like we’re going to need to combine lots of different modes of investigation and data. None is sufficient but all are necessary.
ReplyDeleteI agree with you Auguste, but I think that the main question we should be asking is about the relationship between the parts and wholes rather than focusing on having a detailed understanding of the parts themselves. For example, although we know the hippocampus is important for memory, we also theorize that memories are stored in cortical areas as well. In this case, having a detailed understanding of the hippocampus wouldn't tell us everything we need to know about memory, which is why we need to focus on the relationship between parts and wholes.
DeleteI agree with the idea of focusing on the relationship between the parts and wholes rather than a detailed understanding of the parts themselves. I am wondering if you guys think there is a way to achieve such knowledge about these relationships without knowing the details of the parts? Perhaps Fodor intended to emphasize that the detailed parts are not alone sufficient, we need to explore relationships and other phenomenon to achieve understanding. However, to do this, the details are still necessary with what we currently have available to us as a starting point. The details of functional localization are not the end, but I would argue that they are definitely a means to the end of elucidating the relationship between the parts and wholes and many other ideas too.
DeleteI don't think the lens that we currently tend to look at the brain with (functionalism and localization) is due to the cultural trope that phrenology caused, but rather that our current technology lends itself to it. I don't think we tend to look at the brain as a bunch of parts because we like that tradition, but rather because we are bent on looking to empirical evidence as the forefront of our knowledge, and the most empirical technique we have at the moment is fMRI. Grants tend to be allotted to scientists conducting empirical research and so it is natural that most of the work in systems neuroscience right now is using this method. He didn’t say so outright, but perhaps Foder thinks that more grants should be allotted to scientists building AI neural nets and implicated in deep learning research. However, looking to the future, when we have some ideas of what some aspects of our “software” might look like via computer science, we will need to have a good physical representation of the brain to integrate it with. Furthermore, many deep learning researchers guide their research through looking to the latest in neuroscience.
DeleteI'm not sure how parts vs. "wholes" get us from when/where to how/why. But Fodor is a computationalist, and no happier with neural nets than with neural imagery! (And there was no "deep learning" in 1999: But as we'll see later, Fodor does not really think any kind of learning is important: he thinks most of cognition is innate. But in all that he's on weaker ground than when he is just sceptical that where/when leads to how/why in explaining cognition.)
DeleteAuguste, I agree that it is important not only to look at the parts but the whole as well. Annabel, I think that it is important to have a detailed understanding of the parts. For example, a lot of neurological disorders can be caused by small damage to the brain. Knowing which parts are affected help us understand the brain as a whole.
DeleteWhy, as a computationalist, would Fodor not be happy with neural nets? Is he a formal computationalist that he doesn't think dynamic systems (matter and energy) are necessary for cognition? He just thinks that formal computation/ algorithims for our cognition were predestined at the big bang and that's it?
Delete‘But we already knew that,’ she says, in some perplexity. ‘Now we know it scientifically,’ Pavlov replies. It may be that some such thought also motivates the current interest in brain localisation. Granted that we always sort of knew that there’s a difference between nouns and verbs, or between thinking about teapots and taking a nap, we didn’t really know it till somebody found them at different places in the brain
ReplyDeleteI find it disappointing he rejects scientific methods based on the assumption that somehow “common sense” is enough. Simply looking back at human history, so many times was science used for fraud or for justifying horrible acts. Indeed, bad scientific methodology once reinforced the idea that skin colour changed our intellectual capacity. It was some 1950’s white people’s “common sense” that kept them defending segregation. Relying on our present “common sense” puts us at so much risk of making biased, wrong, subjective assumptions. This is why I find disappointing that the author somehow minimizes the importance of objective, measurable, calculable and replicable facts.
I absolutely agree, Josiane. In fact, in my developmental psychopathology class, our Professor specifically mentioned that "common sense" science is not enough. She used an example of how many parents used to believe that by delaying the tasting of common allergenic foods, children would be less likely to develop allergies. However, studies have found that this "common sense" is wrong. Therefore, I agree that Fordor is incorrect in suggesting that scientific methods are not always needed.
DeleteIn bringing up that story I don’t think it was Fodor’s intention to say that objective/measurable facts aren’t necessary, rather I think he was criticizing the scientist for going through an unnecessary procedure (drilling holes in the mouths of dogs), not finding out anything new, only confirming what he had already suspected, and then assuming that he had gained some new knowledge because what he had thought before now has an objective/measurable correlate. I think Fodor just wants us to be more critical in determining what questions can and cannot be answered by localization studies and whether or not the data we’ve found from those studies is really answering the questions that can be answered. He admits that these studies have clinical applications, but can they contribute anything else to figuring out how the brain works?
DeleteI think your reply was spot on Ramya. Foder is basically stating that he advent of new technology has allowed some neuroscientists to become a bit lazy. They now have the means to produce new results that add to the literature, but since they didn't have a good research question, these studies don't really tell us anything new about how the brain works. Localization studies can (vaguely) get at the WHERE, and maybe with better technology eventually get at the WHEN of mental phenomena, but this method, BY ITSELF, will never get us to the HOW. Scientists are fooling themselves if they think that localization studies will get us closer to understanding the mechanisms of cognition, without mixing methodologies, and formulating precise research questions.
DeleteWhere I think this article is lacking is that he does not acknowledge the difficulty of getting at the how, and doesn't offer any better ways of getting there. I think if you are going to be as cynical and critical as Foder was in this article, you need to offer some tangible guidance as to how to improve the work... Just my opinion.
Localization studies alone***
DeleteI think this is precisely where Fodor's frustration lies – with the advent of any fancy new kind of technologically-advanced technique, there will be a section of researchers who hide behind the glamour of such new technology, and use it to mask an otherwise poorly designed study or poorly defined research question. Methodology classes tend to be some of the most universally loathed classes at universities, and while people acknowledge them as important, any methodology class I have taken at McGill has been lackluster. To my mind, there are a few facets of methodology that I think Fodor wants to tease apart in his strongly worded article – the "technological" methodology, and the epistemological methodology. It's all well and good to learn how to use an fMRI machine, for example, and what kind of data we can glean from using such tests, but a poor understanding of how this data can be manipulated and intersected with other methodologies can lead to data that is of little value.
DeleteI am wondering what/where Fodor thinks the resources and attention of neuroscientists should be focused on if he believes localization of function is of no current use. While some of his critiques are valid, he has no suggestions as to what would be of better use to be studying and researching. While I agree with some aspects and appreciate his ambition to quell some of the popular myths about the value of localizing brain functions, he seems to be criticizing to the extreme without suggesting a better alternative. This lack of suggestion makes me question whether there is currently a better alternative.
ReplyDeleteWhen he makes the analogy between the brain and a car engine and asks, “What part of how your engine works have you failed to understand if you don’t know that?” He suggests that this would only matter if you were thinking of taking a part out, so having surgery or dealing with brain damage. However, I do not think these are the only cases in which this information would be useful. He seems to be arguing that gaining the knowledge of where certain processes are is useless especially if we still cannot solve them. I argue the contrary and cite cases such as H.M. from whom we learned about the intricate processes involved in memory encoding. Localizing this function has spared many epilepsy patients the severe damage that H.M. endured and ensured that surgeons employ neuropsychological tests prior to surgery to ensure that cognitive capacity is preserved. This set the precedent for brain surgery not just in memory regions but also in all other brain areas. The knowledge gained shed light on many neuropsychological syndromes and contributed to the development of drugs and techniques to aid with memory-related issues (ex. Karim Nader’s memory ‘alteration’ utilized the knowledge gained from H.M. and subsequent localization efforts about where and how new memories are formed). Is there not tremendous value in that?
Aliza, I think you raise an important issue of when scientists fail to make their theories of use value. If the goal of research is to adopt refined models, what good use is it to critique without providing alternative theory? It leaves the reader in an existential vacuum.
DeleteBut, I think this may be his intent as writing an opinion piece. Limitations aside, he wrote a pretty thorough article (some years before this one) presenting a theory on the "modularity of the mind". So, I do think Fodor has tricks up his sleeves here, it isn't that he doesn't have theories to provide of his own, but he may be purposely probing the public to think further on what seems like very superficial claims. Notice how he almost satirically (?) titles this "Diary".
I completely agree with you Aliza. I think the point of functional localization goes far beyond trying to establish explanations for cognition. It seems to me that it is extremely useful to investigate where mental functions are neurally localized, not only for the neurosurgeon performing a resection but also in animal studies which help us gain a deeper understanding of the mechanisms of the brain. For example, finding the localization of a speech articulatory area in the monkey might allow us to further understand the mechanism of how articulation is produced and how to better treat language disorders. It may also tell us something about the importance of language and it's evolutionary basis. It is also worth noting that localizing brain functions plays an important role in the dispelling stigma associated with mental illness. One of the major barriers to mental illness recovery is the stigma associated with not having a "diseased" part of the body to point to and therefore feeling delegitimized within the illness.
DeleteI understand your critique of Fodor, particularly in how he underestimates the impact of localizing neural functions (both in terms of furthering our understanding of the brain and in potential clinical applications). However I'm going to (try to) defend one aspect of his argument which I found particularly compelling and is perhaps best represented in the concluding paragraph.
DeleteNeuroimaging techniques each have their limitations (e.g. low special/temporal precision). However, overall they are undoubtedly an important methodological technique and have greatly furthered our understanding of the brain. However, they are still, at the end of the day, just methodology. Neuroimaging does not excuse researchers from asking the hard (or the easy questions). They are not the answer, they are a tool for finding answers.
I think it's interesting you use the example of HM to show the importance of localizing brain function, when in fact most of these findings about memory were discovered without the use of neuroimaging. Discoveries were made because of thoughtful, creative experiments. I listened to Brenda Milner speak once and she made a similar point about methodology. She said (paraphrasing here) the same philosophical approach is required to neuroscience research even if the methodology is new.
To summarize, I think the gist of what Fodor is saying is that neuroscientists are getting lazy because they have this flashy new neuroimaging technique that mystifies people with cool, pretty images of brains lighting up.
I think Fodor’s article raises some valid points about how neuroscience receives extra attention in terms of funding and media coverage, and the curiosity of why people are so interested in the locations of function. This reminds me of the recent brain imaging studies of Cathart-Harris et al. and the administration of psilocybin/lysergic acid to research participants. In turns our that while under influence of the drug, there is increased activity in the occipital cortex, and heightened inter-hemispheric activity. Considering user reports where people claim to hallucinate, and experience heightened associative inferences (“aha” moments) while on the drug, are these imaging studies really surprising or at all relevant to psychedelic experiences?
ReplyDeleteTo answer Fodor, I think there is heightened interest and attention put into functional imaging research for the sake of the for-profit industry. It isn’t necessarily that the public is more interested in this kind of research than other theories of the mind, but it is that this news is what makes it public because that is what kind of research is being funded for/made available. The only reason I could rationalize desire to study the brain while under influence of hallucinogens is to potentially put together a case for the FDA regarding efficacy, motivated by the subsequent steps for pharmaceutical trials. This mostly a case of my own propaganda from usual trends (see recent MAPS studies with MDMA). I’d be curious to hear what others think.
This is not to deny the importance of localization. Fodor mentions that to understand an engine it is sufficient to know it's function, rather than the localization of the parts ("unless you want them taken out"). I don't think (and hopefully Fodor either) brain imaging research should be limited to surgery interests of "taking out" parts. That begs the question though of what kind of motives make for practical research, and which authorities push/pull that discourse.
DeleteFodor's article is understandably critical towards the neuroenchantment that is present in today's science. However, I believe that there are certain aspects of functional localisation that he fails to address.
ReplyDeleteCritically, functional localisation of mental processes to brain structures do not occur for the end-goal of functional localisation, but rather to piece together an organisational structure of the mind and brain. Analogously, the specific spatial location of an atom within a molecule is irrelevant but the specific configuration in which it is spatially located with respect to other atoms define its chemical properties. This is illustrated in the piece by Grill-Spector & Weiner where the hierarchical structure and the connectivity of brain regions have given us considerable insight into visual categorisation.
Even if we take the computationalist stance that mental processes are implementation-independent, there are certain physical configurations that could not support this mental programme. Since the mental programme is observationally opaque to us, our best option is to investigate the physical hardware that runs this programme, since these physical features make it possible for such mental operations, in order to understand how complex mental processes may be realised. Doing so may not produce a direct and immutable understanding of mental processes but may narrow the possible mechanisms that mental operations are accomplished: constrained by physical structures.
Fodor writes, "Unsurprisingly, rationalists thought there might be something to phrenology, but empiricists made fun of it. (The empiricists won the battle, of course; but my guess is they will lose the war.)"
ReplyDeleteHere, Fodor is guessing at who will "win" or "lose" the war... doesn't this imply that he cares about answering the questions he previously mentioned to not care about? He had previously mentioned, "...I wonder why the 'Times' cares. I wonder why anybody cares," when discussing functional localization. But, by making a hypothesis himself... isn't he indicating that he actually understands why people care?
It was more the ‘war’ Fodor talked about that I didn’t quite understand: is it the ‘war’ to shed light on neuroscience perhaps? If so, experimental procedures and evidence is what it would take to convince the scientific community, and many normal people, so I don’t know why he implied rationalist thought is what would actually give us the full ‘truth’ on brain science. And about this hypothesis that he makes – I think he’s making it more to poke at empiricists/imply rationalism to be better, not because he actually cares.
DeleteMr. Fodor’s piece takes an incredibly pessimistic view of MRI and fMRI techniques, and given my last few years of research spent in brain imaging centers and the Montreal Neurological Institute, I’ll take a stab at defending my field of study:
ReplyDeleteNeuroscience, like almost all biological sciences, is built upon incomplete information. The structure & function of the brain are possibly two of the most complicated things in the universe, and it may be impossible to fully understand them both, no matter how good our methods & technology are. With that in mind, scientists in the field of neuroscience are grasping at whatever bits of information they can get, and with this try and piece together an imperfect model of the brain, in order to make real-world decisions regarding the brain.
As Dr. Harnad said in class, information is about the reduction of uncertainty. If a patient has a stroke localized in the upper left portion of the backside of their brain, without the scraps of information neuroscience has gathered over the centuries, we probably wouldn’t know that the patient’s Wernicke’s area is damaged and what specific treatment to provide them with. While this is a basic and extreme example, it shows that understanding fragments of information about the brain can make a difference in people’s lives.
In my own lab, we study the cognitive neuroscience of music, and the pleasure associated with music via MRI and fMRI. While our research might not save lives or revolutionize medicine, the bits and pieces of information that we gather about auditory processing and reward mechanisms get added to the massive collection of knowledge regarding the brain, and pave the way for future neuroscientists to discover more. Without the technology of MRI and fMRI, research such as ours and thousands of other labs across the world would be without a quantitative and nearly direct measure of the activity of the brain, that is unattainable otherwise.
I very much agree with your points on the importance of MRI/fMRI and localization. Say, for example, one was performing brain surgery to take out a tumor. If you don’t understand representation in the brain, you might end up damaging sensation to your hand or speech areas, or even change your personality. I think Fodor would take up the largest argument with your last paragraph, as you do not believe that you are saving lives or revolutionizing medicine - you are merely hoping it’ll pave the way for the future. Fodor would probably say that by studying music you are wasting money which could go to something more pressing. I however disagree, by focusing on pleasure and reward mechanisms it may help us further understand various psychiatric disorders and the best way to treat them (such as with drugs or behavior training or both).
DeleteI think you might be interpreting Fodor as more of a pessimist that he is.
DeleteHe notes the value of MRI and fMRI in the article ("if you are a surgeon you may well wish to know which ones [the areas in charge of language] are since you will wish to avoid cutting them out" p4). His argument seems to be directed at research that has implications outside of medicine (and I would argue that the reward mechanisms you mentioned are included in there). Rather I take his arguments to be more directed at scientists who are more interested in "localizing" rather than looking at pathways and relationships between cortical areas.
Fodor brings up the argument of whether we should adopt the view that specific mental functions are localized in specific areas of the brain, or whether we should view the brain as a homogeneous organism that can give rise to the same mental functions anywhere.
ReplyDeleteI don’t think that we can make such a clear separation between these two views, because they necessarily overlap. Just as it seems illogical to think that the brain has no specificity, it also seems illogical that one particular brain region could be completely responsible for a single task. I feel like there’s a common misconception in society that neuroscientists are focusing on a domain specific view of the brain, as was the case in phrenology (as Fodor mentioned, the popularity of this view is definitely being reinforced by the countless media articles dedicated to neural imaging findings). In reality, every mental operation that we can execute requires the involvement of the entire brain to be complete.
I have an example that I think will better illustrate my point. Neuroscientists have identified a “language region” in the brain, aka a number of brain areas that seem to be highly specified for language processing and production. However, our language capacities do not rely on the functioning of these areas alone. If we take the case of a patient with transcortical aphasia - a teenage girl who suffered from carbon monoxide poisoning, damaging every region of her brain, while leaving her “language region” in tact. Even though the area of the brain that seems to be specialized for language function was left undamaged, this patient still had severe language deficits. She was only able to repeat sentences that were said to her, and was not able to communicate any original thought of her own. While it is possible that certain brain regions are more implicated in a particular task than others, every task still requires the involvement of the entire brain in order to be executed to it’s full potential.
“I want, to begin with, to distinguish between the question whether mental functions are neurally localised in the brain, and the question where they are neurally localised in the brain. Though I find it hard to care about the second, the first clearly connects with deep issues about how the mind works…”
ReplyDeleteI disagree a bit with Fodor on this as I believe that both questions are important. Caring about where mental functions are neurally localized has attributed a lot to our understanding of why and how cognition works in the way that it does. This has also offered a great deal of information for individuals with brain injury who experience delusions or deficits in certain functions. Without knowing where certain functions are localized we wouldn’t be able to have answers to why individuals who have brain damage between their visual cortex and their amygdala perceive those who they love as imposters, a disorder known as Capgras syndrome. Knowing where these localizations are advances our knowledge and understanding greatly, and in situations of brain injury it provides us with many answers. However, that being said, the fact that certain functions may be localized doesn’t provide us with all of the information because multiple areas work together and are involved in mental functions. For example, there is a lot of research showing that there is an area in the temporal lobe specialized for detecting faces known as the Fusiform Face Area. However, individuals who have prosopagnosia have damage to this region and therefore cannot discriminate between faces. Importantly, though, they are still able to perceive a face as a face. Therefore, while this function may be localized in a certain brain area, it isn’t the sole region involved in face perception. Our focus should therefore not be entirely on where functions are localized but also how the brain works as whole in order for cognition to occur.
On the one hand, Fodor has a point that on its own, being able to point at such and such part of the brain as the teapot processor is not particularly useful. On the other, I think it can still have tremendous value, in particular in clinical research. Understanding the functions a particular brain region is involved in can help guide treatments and surgeries for patients with all kinds of neural dysfunction. But even beyond this, functional imaging is an important piece of the puzzle when it comes to decoding the brain. In conjunction with other methods, including patient studies, EEG, and animal research, we can build a more complete model of its structure and function, and overall organization.
ReplyDeleteI really enjoyed this article, Fodor is quite cheeky about some pretty interesting debates in cognitive science that are worth discussing. The relationship between the physical mapping of neurons and mental states is not quite established. I think it's important to keep in mind that it doesn't appear that Fodor is against arguments about the structure of the brain and the cognitive mind - quite the contrary really given his work on modular minds - rather he is arguing against neuro-imaging of concepts where the location of the concepts is the goal.
ReplyDeleteThere are lots of clinical applications, of course, to neuro-imaging, but the question is how can the location of semantic content of objects, scenes, faces, etc, help with understanding cognition? Though the paper was written in 1999, even a decade and a half later I really notice a lot of popularized neuroimaging studies in pop-science media, and often have similar thoughts to Fodor. For example this past year Gallant Lab took fMRI scans of subjects listening to stories. Using a model they were able to generate a huge array of data and essential map concepts used in speech all over the cortex. I highly encourage anyone to check it out, it's very relevant to this discussion and I would be interested to hear what Fodor thinks of this kind of project. On their website they have a couple points about why their research is relevant and what value it provides, it seems they are avoiding the argument that Fodor is making.
http://gallantlab.org/index.php/publications/natural-speech-reveals-the-semantic-maps-that-tile-human-cerebral-cortex/
I also really enjoyed this article. However, I disagree that Fodor’s argument is based on the localization of purely semantic notions. He does not just bring in semantic notions of language, but also the ability to ride a bicycle or construct a bridge. In this way, I think his conversation is of a much larger scope than what was done by the Gallant Lab. The Gallant lab’s study is neat, however, after (briefly) looking over it it seems like, just as you suggested, there are clear parts that Fodor would object to. For starters, how can the lab be sure that the same words don’t invoke different emotional feelings for different people? Isn’t it reasonable to assume that a story about an ocean might make one person distraught and another person nostalgic? I’m confused how they believed that they could separate all notions of subjectivity.
DeleteI believe that this idea relates back to Fodor’s article. How do we separate different mind-like functions from brain-like functions? It is clear that if we sever certain parts of the brainstem people stop breathing, but would Fodor be okay with this type of brain localization study because it doesn’t necessarily involve the mind (you don’t have to think about breathing)? But then how do we delineate between mental and physical functions of the brain? To return to Fodor’s example of riding a bike, is it possible that after riding a bike so many times it no longer requires any mind-like process and becomes purely physical? Or conversely could someone lose the physical ability to ride a bike due to a brain lesion but still retain the ability to think/dream about riding a bike? I think it’s hard to argue that the localization of physical abilities in the brain is irrelevant, since we know that areas of the brain are needed by humans to exist. So then where does Fodor draw the line between useful brain imaging and irrelevant brain imaging?
@Lucy, yes I agree Fodor's argument goes beyond the scope of semantics. With regards to the Gallant lab, if a lab was to also generate a syntactic map of the brain I think the point still stands, he's arguing that in order to look at why and how I can think about or preform riding a bicycle isn't illuminated by where. Though we know that our brain hardware affords us the ability to ride a bike, does knowing which parts of the brain are involved really help? When we talk about brain imaging it's easier to think of V1 of the occipital cortex as a visual stimulation button, but does activation in this region say how we can recognize an ocean, or activation of the amygdala say how we feel about the ocean. Even if we mapped every semantic and syntactic notion to regions of the brain would that helpful to understanding how that particular mapping does what the brain does. I think that useful in a medical/practical sense is hard to disagree with, but with regards to the the argument all brain imaging is irreverent.
DeleteI had two questions about Fodor’s paper:
ReplyDelete1.When he says, “if the brain does different tasks at different places, that rather suggests it does them in different ways” how is this true? This logic seems rather flawed. I guess it depends on what a ‘task’ is defined as. If a task includes the way it is accomplished then, yes, I agree. However, it seems reasonable to imagine 2 tasks being completed in different ways. This could also be independent of the exact location where the tasks are completed.
I guess this would just further feed into his argument. If we know the task and the location in the brain where it is done, it does not entail that we know how it is done.
2. I found the question he brought up very interesting, namely: “what if, as it turns out, nobody ever does find a brain region that’s specific to thinking about teapots or to taking a nap? Would that seriously be a reason to doubt that there are such mental states?” This seems similar to the other minds problem. There is no way to know whether someone is thinking about a ‘teapot’ or not. If every single thought were identified except for ‘teapot’ we would not be able to say the thought doesn’t exist because there is always a possibility that something exists which we are not yet able to perceive. However, it would make it statistically unlikely for the ‘teapot’ thought to exist.
2. There are complex connections between different brain regions. Fodor seems to be under the impression that:
Deletethink about teapot -> "think about teapot" brain region lights up -> the existence of a "thinking about teapot" mental state. Or something along those lines.
The fact that there's not a specific "think about teapot" brain region lighting up (or that there is, in fact, no brain region specifically devoted to thinking about teapots) does not mean that there is no mental state that is "thinking about teapots."
I totally agree with you! I think he is very anti the 'connectionist' model. Many abstract thoughts can be formed by the activation and inhibition of numerous brain regions. However, the question still stands. Let's say that all of the patterns of activation and inhibition are identified for every possible thought, except for 'teapot'. Now could that signal that the person does not have that concept?
DeleteValentina, when Fodor says “if the brain does different tasks at different places, that rather suggests it does them in different ways” I don’t think he is trying to make a strong logical distinction. Absolutely I think he would agree that it is quite possible that two locations could be structurally identical beyond location.
DeleteHowever, it wouldn’t be very functionally or evolutionary unwise to have two areas that do the same function. Plus if two different areas functioned identically we could never predict whether task A or task B would occur in area X or area Y. Therefore I think what he is saying is that if the brain does something in two different places there is a reasonable hypothesis that it does them in different ways. I think this is important because it shows the limits of his stated rejection of localization questions. As even his detractors point out at the bottom of the article, he is an founding proponent of the modularity of the mind, therefore it is strange for him to care so little about location. That said, Fodor is pretty playful. I think he is trolling cognitive neuroscientists (but not clinical neuroscientists). I think he makes a strong distinction between the brain and mind. For him the mind is computational, therefore it is implementation-independent, and therefore while the mind is modular in his view, I think he sees the brain as the computer running the mind software. Since computers can be equivalent, perhaps he feels the physical implementation is less interesting than the underlying cognition, of which the how-ness remains intractable to probing.
RE: But we already knew that,’ she says, in some perplexity. ‘Now we know it scientifically,’ Pavlov replies.
ReplyDeleteAs most of my classes have been related to neuroscience I have never questioned why we need to localize functions of the brain and what they do. It's just normal at this point. In terms of the quote above I think that knowing something without proof is not enough. People need hard facts and evidence to come to a conclusion even if they could come to the same one using common sense. Although Fodor makes some interesting arguments about why we don't need to localize concepts I think that this kind information makes a point more valid. It is easier for people to accept certain information if there is evidence to back it up and I think that localizing certain brain functions is information that needs to be studied in order to have a clear understanding and belief of what the brain does and how it does it.
This comment has been removed by the author.
ReplyDeleteFodor makes a compelling argument – it is difficult to see the purpose of activity-specific localization of brain activity. I agree that neuroimaging of “when and only when whoever owns the brain is thinking about teapots” seems like an odd place to be spending scientific grants. Two things stood out to me about his article:
ReplyDeleteFirstly, this argument sounds familiar: It has always been easy to say in hindsight that research was unproductive. However, scientific discovery is often unpredictable, and we are (as of now) unable to say which areas of scientific research will be fruitful. Brain localization may not be productive today, but this research may someday aid us in recreating the brain – or it may not. If the brain is not only a computer (as we have maintained in class), we are not dualists, and the physicality of the brain is necessary for feeling, how else would we go about trying to understand it?
I suspect that brain-scanning seems unproductive because of the enormity of the task, and how little we actually know. He states that the location of the carburetor is irrelevant to understanding what it does, but if someone from the stone age found a car, how would they know where to look? Who is to say where the next breakthrough will come from? He argues that there are more important questions relating to the brain, but historically answers come from unexpected places. Penicillin was discovered because of unwashed dishes, microwaves a result of melted pocket-chocolate, and an apocryphal story suggests that gravity was discovered while fast asleep. While teapot scanning seems almost laughably obscure, perhaps that lab will uncover the next great age of cognitive science.
Secondly, I understand (and share) his frustration surrounding “X laboratory has discovered the location of Y in the brain” type arguments. It seems to me though that this is much more the fault of media wanting to make science exciting than the brain scanning itself. Further, he mentions that this research is expensive – it is hard to convince someone to part with their money for an unproductive study, and more than a few scientists have been guilty of dramatizing their findings for this reason. However, neither the media or over-excited scientists reduce the usefulness of understanding the brain, as expensive and slow as progress may be. Even if no progress is made, as I have argued above it is impossible to know ahead of time. Maybe the next time they figure out the neurological basis of love they will be right?
That all being said, we have been barking up this tree for at least 18 years – and in that time, less has been accomplished by brain-scanning than optimistic predictions suggested. Through this sort of reasoning Fodor raises some concerns, but like many critiques he fails to provide a replacement – If not by fMRI, how should we move forward with the physiological end of brain science?
"What if, as it turns out, nobody ever does find a brain region that's specific to thinking about teapots or to taking a nap? Would that seriously be a reason to doubt that there are such mental states?"
ReplyDeleteI think Fodor fundamentally misunderstands the issue of localization. No, the fact that there is not found to be a specific brain region devoted to encoding "thinking about teapots" does not mean that nobody thinks about teapots. No neuroscientist would claim that. When one investigates whether something is localized in a specific area of the brain, the question is not to investigate whether or not this thing is actually thought about / whether or not this "mental state" (to use Fodor's term) exists.
"Or that they are mental states of different kinds?" I'm not sure what he means here.
"Or that the brain must be somehow essentially involved in both?"
Unless one takes a dualist standpoint, the brain is essentially involved in thinking about anything. Fodor later implies that he's not a dualist (at least by my reading), so this seems like a bit of a silly question.
"As far as I can see, it's reasonable to hold that brain studies are methodologically privileged with respect to other ways of finding out about the mind only if you are likewise prepared to hold that facts of the brain are metaphysically privileged with respect to facts about the mind; and you can hold that only if you think the brain and the mind are essentially different kinds of thing. But I had supposed that dualistic metaphysics was now out of fashion, in the brain science community most of all."
First, something being methodologically superior does not mean it is “metaphysically privileged”.
Second, what does he mean by "the mind"? Fodor starts out by referring to “the mind” and strongly implying that the brain and the mind are different things - ie. There are ways of finding out about the brain and ways of finding out about the mind. He then says there are "facts about the brain" and "facts about the mind.” In saying this, he begins with the implied premise that the mind and the brain are different. If his logic leads him to that same conclusion, it is because he himself made this distinction. It is Fodor, not neuroscientists, that appears to subscribe to the unfashionable ‘dualistic metaphysics.’
Fodor’s entire argument seems to revolve around his objection to studies of the brain having a “significant scientific import over and above their implications for medical practice” or, in simpler words, having a point. What is the point to or “import of” any scientific research at all? Fodor mentions his interest in astrophysics – What is the “point” or “scientific import” of learning about quasars and blackholes? Why is there a point to learning about “how far away everything is” (which doesn’t even have any medical implications) and no point to learning about functional localization? This is never explained.
This article made me think of how pop psychology developed off of psychology, and how the same thing seems to be happening with Neuroscience. As psychology became "applied" to daily life, there was a boom of articles, tv shows, etc. taking advantage of the public's fascination with these topics which lead to things like self-help.
ReplyDeleteGoing off of Fodor, 'pop neuro' might be leading in the direction of localization of function. In general, people seem to love the idea of it and it could be easily formed into a profitable business that uses pseudo-neuroscience to appeal to the public and make money. The media creates these trends that Fodor is wary of, which is very problematic for the actual science that is trying to be done. The media and the public favor 'sexy' things (like right/left brain) which may become counterproductive to the actual questions trying to be answered.
Already there are loads of books on how to strengthen a brain hemisphere (like drawing books)....
Jerry Fodor brings up a lot of great points. What interests me the most is his distinction between why and how. I tend to agree with him in that brain science is too focused on anatomy and functionality at this point. What does it matter, neutrally, how we think about teapots or taking a nap? Surely by thinking about these two things the thought itself exists? Does it matter if it sparks a rise in neural activity in the visual V1 cortex or the medial prefrontal cortex of planning and decision making? We still experience both visual and intent associations with teapots and naps. I think that once you get to higher abstraction levels of questioning minds/brains/thought, then brain mapping becomes less specifically empirically relevant and more abstractly relevant. If, for the sake of argument, we think of the mind as a program, the lines of code that would correspond to neural firing patters still can only explain function and not necessarily performance.
ReplyDeleteI can see your point. I think the for what we are concerned with in this class that the bigger picture of how the brain works is more important or relevant for understanding cognition. However, the only problem that I do see with paying less attention to the finer detail of the brain mapping is that different areas of the brain still work together in order to carry out functions so in order for a machine to be able to replicate human behavior, wouldn't it need to be somewhat concerned with the internal wiring and connections that exist?
DeleteI think the importance of anatomy shouldn't be completely disregarded from a medical standpoint. If people have certain trauma's, it's very useful to know localization of function. Even if it turns out this is just modern day phrenology, it would then be useful medically to know that this was not the case.
DeleteHowever, from the perspective of the class it's less interesting because we are looking holistically at the processes of the brain.
The author grossly misinterprets much of the literature and as a consequence has pretty weak logics in his argument.
ReplyDeleteFor instance:
"Similarly, if you think there’s such a thing as ‘general’ intelligence – what IQ tests are supposed to measure – then you should also think that designing bridges and designing foreign policies manifest much the same kind of cleverness, albeit applied to different tasks."
Many neuroscientists agree that even the best IQ tests are fairly flawed, and perhaps not the most indicative measure. However, even IQ tests have multiple subtests (i.e. Verbal Comprehension Index (VCI), Visual Spatial Index (VSI), Fluid Reasoning Index (FRI), Working Memory Index (WMI), and Processing Speed Index (PSI) for the WISC // Verbal Comprehension Index (VCI), Perceptual Reasoning Index (PRI), Working Memory Index (WMI), Processing Speed Index (PSI) for WAIS)
It's hard to imagine any of the contemporary authors getting published in the Times advocating that bridge-building and foreign policy writing are interchangeable tasks under the umbrella of general intelligence.
Fodor's impatience with the literature for, granted, individually meaningless discoveries of localization is frustratingly close-minded.
In 1508, Leonardo DaVinci and others knew that the elbow was a point of flexion and tension, should he not have examined and illustrated the inner workings of cadavers then because of the redundancy of seeing "where" these incredible anatomical processes happened?
I strongly agree with your criticism of how Fodor makes a huge and unwarranted generalization with his general IQ claim. I scratched my head when I read that paragraph.
DeleteIn regards to the DaVinci example, I don't think it's comparable, because DaVinci dissecting the elbow is a purely mechanical exploration of the elbow bending function. While neural firing can also be seen as mechanical, the corresponding mental experience of whatever function you're looking at is not physical. To me, these two situations are apples and oranges.
For me, the goal of research in neuroscience is ultimately about human health. It’s undeniable that the brain is the most important organ in the human body (maybe alongside the heart). Even minor damage to the brain can result in severe impairments or potentially even death. Therefore, it is critical that we attempt to learn about the brain and try to understand how it functions so that when injuries happen, we can appropriately fix and limit the damage. However, the brain is one of the most complex things in the universe and we’re nowhere close to understanding it. It’s impossible for us to understand the brain with a single research experiment. Therefore, we have to put together all the little pieces of information to produce a clear map of the overall function. That means that any new information moves us one step closer to understanding the brain and who wouldn’t get excited about that. Secondly the attention from the New York Times will help inspire people to see possibilities in the field of neuroscience and potentially help in our quest to understand the human brain. With how important the implications can be, one extra helping hand today, could save millions of lives in the future.
ReplyDeleteI agree! Fodor mentioned that resources like time, money and people trained to do science-graduate students-are limited and therefore he feels bitter when money is spent on brain research in neuroimaging. I also think that promoting brain science indeed can help raising more awareness in brain research and help to gain more resources for the field! I guess what Fodor wants to convince others is that, he believes neuroimaging is not the correct way to understand cognition. However, I do believe that neuroimaging will have the use to help us verify whether a successful reverse-engineered thinking machine we create in the future, resembles to our brain or not.
DeleteI am easily drawn in by a persuasively written essay. I finished and found myself agreeing with Fodor wholeheartedly and unquestioningly. Thankfully this has worn off slightly. I do find certain parts of Fodor's argument important but I believe he is throwing the baby out with the bathwater* when he argues that localization studies are not helpful in answering question about the how the structure of the brain maps onto the mind.
ReplyDeleteI do appreciate Fodor's willingness to discuss science as a field dictated by trends - sometimes at the cost of other (perhaps more deserving disciplines). Stemming from that his distate towards of "trendy" science being performed without a hypothesis, as a tool of observation is understandable. As integral as tinkering might be to the field of science, it is important to consider the viability and value of one's work in the context of grants.
However, I find there to be confusion between localization studies and studies at a systems level. Fodor focuses his frustration on imaging and localization studies - not systems based, interactional models. Despite this, all the responses he receives defend the systems based point of view - something that Fodor himself seemingly agrees with (see: his response to Benjamin Bly). It seems as if Fodor wants the "photographing" to stop and for someone to start building systems based hypotheses based on this information instead. In this situation, I hope Fodor gives due respect to the tinkering and localization that allowed hypotheses of "subtle interplay between brain regions" to be developed in the first place.
This comment has been removed by the author.
DeleteI agree completely with this point! I think Fodor fails to acknowledge how young the field of neuroscience is relative to other disciplines. We are in an experimentation/"tinkering" phase because there is still so much about the brain that is not known. He also neglects the growing recognition within the field to move towards a systems-level analysis and view the brain in terms of neural circuits (rather than isolating functions to a localized area).
DeleteFodor suggests that in terms of theoretical neursocience, empiricist "typically hold that all mental processes reduce to patterns of associations, would like the brain to be ‘equipotential.’"
ReplyDeleteWhile, rationalists "think that there might be as many different kinds of thinking as there are different kinds of thing to think about, generally would prefer the brain to be organised on geographical principles."
He then argues that "unsurprisingly, rationalists thought there might be something to phrenology, but empiricists made fun of it.(The empiricists won that battle, of course; but my guess is they will lose the war.)"
First of all, it seems strange for Fodor to put this prediction of empiricists "losing the war" without any other evidence or argument to back up that claim, or any real hint as to what he means by that. Furthermore, it seems that this idea of rationalists' phrenology failing to catch on to be counter to his argument that localization of brain areas are overpopularized. What are localizations of brain areas (especially examples such as the mundane teapot-area) if not scientifically-backed phrenology? Moreover, how can he argue in favor of phrenology and rationalists and say that he expects them to win overall if he's against neuroscientific localization?
One potential solution to this problem is that he is saying that empiricists won at first and now rationalist phrenology is getting more popular again, leading to increase in research on localized brain areas, and that these theories are what will "win the war" in the end, much to his disappointment.
This comment has been removed by the author.
ReplyDeleteRE: "But whereas, historically, studies of the localisation of brain functions have often been clinically motivated, I take it to be currently the consensus that they have significant scientific import over and above their implications for medical practice……..And if, as I suspect, none does, why are we spending so much time and money trying to find them?"
ReplyDeleteI think that Fodor's critique is not entirely justified and his approach is highly close minded. His point of view's come from a very philosophical point of view, thus brushes off some applications of functional mapping. Although mapping the brain on functionality doesn't exactly tell us how we do what we do, localisations of the brain functions often have clinical applications. His point of view is highly unfair to the field of neuroscience because the studies contribute to understanding the structure and function of the brain that are highly complex and difficult to understand. Every additional piece of information from neuroscience is crucial towards understanding what makes us human. Brain imaging techniques has been applied to further our understanding of neurobehavioral disorders, such as Alzheimer’s disease, epilepsy, brain tumors, stroke, traumatic brain injury, and multiple sclerosis which affects a large population of the world. Therefore it has many important implications that should not be overseen so easily.
This comment has been removed by the author.
ReplyDeleteFodor's apparent bafflement with the media's fascination with the brain surprises me, particularly considering his background as a philosopher (and the origins of psychology in philosophy). The "mind body problem" dates back to Greek philosophy and has been debated for centuries since. Questions of consciousness and "other minds" are of fundamental philosophical importance. Now, advents in psychology and neuroscience have the potential to explain a range of cognitive functions including perception, language processing, memory, and even emotions. Therefore it's not at all surprising to me that, in comparison to other disciplines in science, it receives greater attention. The psychology of attention is just more immediately applicable to people’s everyday lives than the geology of sedimentary rocks. Fodor fails to realize that the fascination with the brain extends far beyond neuroimaging studies to the disciplines of psychology and neuroscience overall. I will concede that just because research is interesting does not necessarily mean it is useful. A great deal of neuroscience research likely does not/will never have direct clinical application, but the same could be said for research in astrophysics. Academic research is sometimes done just for the sake of satisfying our curiosity and expanding our understanding of the world, why single out neuroscience?
ReplyDeleteWith respect to the rationalist vs. empiricist debate, my opinion is that neither side can truly win. Yes there are some functions that can/have been localized to a particular area in the brain. However most complex cognitive functions (which in my opinion are the more interesting targets of research rather than a mundane "teapot") have been shown to involve multiple brain regions (e.g. visual perception requires integrating and relaying information from/to at least 30 regions). Information is processed and integrated in complex ways in the brain. Thus, localizing specific functions is not sufficient in explaining brain function, but it is necessary.
Isn't Fodor the one who also thinks that nothing genuinely new can really be learned? Similar to the Cook article, it seems to me that Fodor prescribes himself to the concept that thoughts/cognitive activity are computations that act on a specific function that is relatively independent of anything else going on. His criticism of this focus on localization of function is not entirely off-base as sometimes it lacks explanatory power, but associations play a powerful role in clinical brain studies and micro-circuitry can often provide powerful explanations for functionality. The idea that the brain functions are innate to its discrete elements without regard for the overlap/associations between those elements again dismisses the usefulness of neuro-circuitry in elaborating cognition and opting for a purely computational approach.
ReplyDeleteI agree. And based off of what evidence does Fodor purport that our all of our categorization capacities are completely innate? While there is rich evidence suggesting that UG must be an innate subset of the categorization capacity (POS, etc.), but this is not a basis for extending the claim across the entire domain.
Delete“Brain scientists are supposed to be materialists, and materialists are supposed not to doubt that distinct mental states have ipso facto got different neural counterparts. That being so, why does it matter where in the brain their different counterparts are?”
ReplyDeleteI don’t think that Fodor is right in calling neuroscientists (or rather the Tuesday Times) “closet dualists.” Perhaps it is true that a materialist “ought” not to care about where functions are localized in the brain, but this isn’t the same as saying that a materialist wouldn’t find this to be true or meaningful information, or even provide some knowledge about how functions are related in the mind. He can make the argument that learning this information isn’t a useful way to spend resources, but I don’t see that it has that much to do with whether a person is a dualist or a materialist.
Fodor raises some important concerns (where funding for research goes – the most sensational research vs. the most practical) and questions that a scientist must continually ask themself as they go about their career ("what do I want to actually get out of what I am doing? What’s the point?"). However, I don’t believe that his argument should be taken to the conclusion that finding out about the brain’s structure won’t tell us anything important about its function. Take, for example, neural networks. As far as I know, studying how neural networks are organized (their “architecture”) is directly relevant to figuring out how they can do what they do. And, as Fernanda point out in class, figuring out how the brain does what it does can help pave the way for the important medical applications that Fodor mentions in this article.
ReplyDeleteRE: “But why does it matter where in the engine the carburettor is? What part of how your engine works have you failed to understand if you don’t know that?”
ReplyDeleteKnowing where, means that we have greater specification of how it is that the brain is working. If the mind is an epiphenomenon of the brain, then understanding the mechanics of the brain should be a sufficient explanation for how we do what we do.
For example, knowing that neural networks work with parallel distributed processing serves as a model for better approximation when building AIs. More data serves to reduce the magnitude of underdetermination. Therefore, knowing “where” in the engine the carburettor is, as Fodor puts it, helps us to better understand “how” the brain is working, which in turn helps us reverse engineer the brain with greater precision.
‘You think we’re wasting our time, don’t you?’ he asked. I admit, I didn’t know quite what to say. I’ve been wondering about it ever since.
ReplyDeleteEssentially “what is the point’ --You need a question with an objective to get an answer that has substance. Fodor is critiquing the short sighted approach to some of these studies. However, I believe the statement ‘the where and when doesn’t answer the how and why” is too far sighted to go about any meaningful research.
Understanding the where and when of gravity never got us to the how and why- and that can be said for most scientific discoveries. But that doesn't mean we should stop searching for the where and when. The where and when is a good thing to focus on- in fact it is all we have the potential to focus on. The where and when of neuroscience helped us to understand things like unilateral neglect, different ataxia's, and aphasias. We shouldn’t stop searching for the where and when, but at the same time we shouldn’t go digging just because we have a shovel.
I disagree with Foder in that I don't think we should discount the worth of our work mapping brain regions. Neuroscience is a young field, and we need to do some tinkering at the beginning. These fMRI studies will eventually be instrumental in creating good simulations of the dynamics of the brain. I agree that it is silly to be overzealous, and aggrandize the worth of the findings of any individual correlation study, but taken together, the map of the brain these studies will provide will eventually help us solve the easy problem.
ReplyDeleteThis was a provocative opinion piece to read but even more juicy were the responses below (did you catch the exchange between Benjamin Martin Bly and Fodor where they threw one another's quotes at each other?). This strain of neuroscientific research does have a tendency to lead to the simplified and inane popular science articles which attempt to pinpoint certain feelings or processes to distinct locations in the brain (a modern day phrenology.) But I agree with the others in believing that Fodor's critique of functional neuroimaging goes too far in dismissing the technique outright. The examples of the teapot and napping, and how propinquity does not necessitate qualitative similarity is a straw man argument; no one is arguing against it. I find his assertion that "we know certain functions occur north of the neck and it doesn't matter how far north" to be excessively facetious. It does matter because knowing which areas are active in certain processes and not others can at the very least provide evidence to support either the empiricist or rationalist perspectives he discusses earlier in the passage. The alternative, not knowing anything about the localization of function, just that they occur and how it feels when they do, serves no explanatory function at all.
ReplyDeleteRE: “mental functions have characteristic places in the brain, why should it matter... where those places are?"
ReplyDeleteI understand that Fodor was trying to imply that simply conducting fMRI scans will not yield any answers to the hard problem of cognitive science. However, he did not articulate that point well enough. He merely states that he sees absolutely no benefit in conducting mapping experiments during cognitive tasks. When a patient sees a goes to see a doctor about a suspected neurological disorder, the doctor could make an educated guess as to where this problem may be occurring based on the patient’s specific impairments. Where would the information come from that allows the doctor to make this educated guess? Most likely brain imaging studies. I agree there’s not much use to mapping the brain when asking questions about our consciousness, but to suggest there is no benefit is ridiculous.
In this article, Jerry Fodor brings up the idea of neuroimaging research being capital-intensive, using scarce resources such as time, money, and computer power. We spend so much on a research method that does not even guarantee effectiveness and still yields some uncertainty. It is expensive, largely publicly funded, and it seems as though there is never enough money to do important research. Fodor thinks that there could be other possible methods of learning about the brain, but we are too stubborn and too persistent on using current neuroimaging methods. Perhaps there really could be other ways of understanding the brain that could tell us information not captured by an MRI machine. Functional localization tells us where neurons are active in the brain given a certain activity. Instead of trying to find where things happen in the brain, maybe we can focus on what things happen in the brain, how they happen, and even why they happen. These are all valid questions that could require non-neuroimaging techniques for further understanding. I believe this is integral in the study of cognitive science where multiple disciplines come together to try to study cognition. Philosophy, Neuroscience, Psychology, Linguistics, and Computer Science all have different things to contribute and I believe this is the point Fodor was trying to emphasize. In short, we should expand our options as to not limit ourselves to "where things happen in the brain," and we must also try to find other more efficient and less costly ways of studying the brain.
ReplyDeleteThe Fodor’s article seems to be a little controversial and appears to say that we probably will keep not finding solutions to the hard problem until we’ve literally dug as deep as we can into the physical brain. We correlate areas of the brain with specific mechanisms, though how sure are we that those areas actually are 100% the cause of the mechanisms they’re claimed to be responsible for? For this we need to look further into the brain and continue to do so until we stumble on something interesting, as with all science.
ReplyDeleteI feel that some discovery will be made at some point, but it continues to sound like we’re mostly just stabbing away into the dark until we finally run into something. Perhaps we're not looking in the right places, or are failing to link all five streams. How should Cognitive Science move forward in a better way? Is neural imaging really going to solve the problem? Are being too restrictive and ignoring other areas of knowledge?
This reading gave me "the whole is greater than the sum of its parts" vibes, but I feel as though knowing how the parts function is crucial in understanding how the whole produces an output.
ReplyDeleteFor example, I thought of this in terms of a camera. A camera produced an output (a picture), but it can produce different ouputs (pictures) based on how its parts are functioning. The aperture of the lens, the ISO sensitivity of the sensor, the speed of the shutter... all these parts have an effect on the overall output--how light or dark it is, whether there is a depth of field, how noisy it is, etc. And if you don't understand how each individual part plays a role, you don't necessarily understand how the outcome is produced. You just know that pushing the shutter button takes a picture. Knowing how they work is important in improving on them, modifying them, or what have you, and I think the same can be said for knowing how different parts of the brain work.
I think that people are interested in where functions are localized in the brain for medical reasons (e.g. diagnosing and surgery) and also because they want to know if a particular structure controls the “mind”. I think it might go back to the ideas of how consciousness is created. Is it a result of processes from the physical activities of the brain? Or something that you cannot recreate, like something similar to the concept of a soul?
ReplyDeleteOstensibly, we would need to focus more how the parts interact to create the whole, rather than the parts themselves. We’d need to mentally spread the information in such a way that we can see the properties of the parts, and the properties of the relations between parts, and the properties of the relations between relations, and the properties of the whole (the brain) itself, all at once. We can then see in a parallel fashion how when/where happens, and can start correlating it with how/why, almost like cooking. We need to understand the different ingredients themselves and how they interact to understand how the cake comes out the way it does. To focus on only one part or relation is to push the others to the wayside without considering how it works with other parts or relations.
ReplyDeleteI understand Jerry Fodor’s concern for the amount of money spent on brain imaging studies in comparison to other methods of studying the mind. It does seem to me these days that psychological literature is dominated by neuroimaging studies, while providing less grant money to other methodologies (that could be effective in different ways). However, despite my own skepticism of neuroimaging, I am not compelled by his argument for why it isn’t important AT ALL. The main utility of neuroimaging that came to mind, that Fodor did not even mention, is it’s clinical value. Being able to localize brain function is incredibly valuable in finding effective treatment methods and cures for a variety of brain disorders. Although Fodor seems to think that scientists are interested in brain imaging because the computerized colors are mesmerizing, and “just because.”
ReplyDeleteI think Fodor’s argument is valid that there should be a distinction between asking whether mental states produced in the process of cognizing depend on certain brain structures and the question about where these sites would be in the brain. However, I don’t think that the second question is unimportant or irrelevant but rather I see it as a key to solving the first question. We can use the data we gather from the ‘where’ question as pieces of evidence for answering the ‘whether’ question. This might not be the ideal way to go about answering the question but science and controlled experiments are the best tools we have at our disposal and the best they could do is to give us some correlational insights to the where question. With these insights, we might get closer to posing the right questions about cognition and its relationship to the brain.
ReplyDelete