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.
I agree with many of Fodor’s points. Yes, scientific research is expensive. Yes, public taxpayer dollars go into funding scientific research. Yes, it’s a waste of time and money to fund studies that aim to see what parts of the brain light up when people think about forks. With that being said, neuroimaging techniques can be used for much more meaningful and practical experiments than making mental maps of what lights up where, when people think about forks. This is why I disagree with Fodor that it shouldn’t matter where mental functions are located in the brain (if they indeed are located there). With technologies like tDCS and TMS we have ways to non-invasively affect specific parts on the brain. This leads to interesting ways to think about neuroimaging research techniques. If we can find brain regions that correlate with mental states like anxiety, depression, fear, etc. we could attempt to see if we can treat them locally with tDCS and/or TMS. Also interesting, if we find brain regions that correlate with happiness and relaxation maybe we can try to induce them using the same techniques.
ReplyDeleteI think that Fodor makes an excellent point when he says that “the availability of the new technology is running the science rather than the other way round” sometimes it seems as though researchers seem to be asking questions specifically so that they can apply a new technique on it which will perhaps make a greater impact in the scientific community. Scientific exploration is often limited by the techniques that are present to test hypotheses. In the past, our tools were not precise enough, did not provide us with the opportunity to look at diverse questions but now we perhaps have another problem. Today we have high-tech tools like brain imaging and so brain research has limited itself exclusively to questions which can be directly tested by these methods, which can lead to an extreme amount of reductionist thinking without consideration of more broad questions.
ReplyDeleteI also think however, that Fodor is wrong to say that any effort to determine where different mental functions are located is a waste of time. In science, structure often goes hand in hand with function and so by understanding the structure of a system we get one step closer to understanding the function of that same system. True, it doesn’t really matter what that structure is in the end, just that a particular function is adequately performed, the structure itself remains an essential part of the way in which a function occurs. Just as the hardware regardless of its configuration is essential to running the program, so too is the brain to its functions. Understanding the structure of the human brain is a means to an end, not the end in itself and so long as it doesn’t completely detract us from the more important functional questions, it is an important step in neuroscientific research.
Fodor suggests that finding out where and when something happens in the brain when you do something does not explain how or why you can do it. It just gives you correlations, not a causal explanation.
DeleteAs to structure and function, compare the bladder and the brain. Both can do things. The bladder receives input from the kidney and its output is the urine it expels. The brain receives input from the world and its output is everything we can do (vegetative and cognitive): T3.
It is easy to see how structure helps explain function for the bladder, kidneys and heart (or a pendulum, clock or computer). But for our T3 capacity this is not at all obvious.
(And if computationalism were correct, there would be no way to determine from the brain's hardware what software it is running.)
(Fodor is a kind of fuzzy computationalist, packing a lot of illicit content into computational "representations" in the "language of thought," which, as we'll see next week 5, is ungrounded.)
I sometimes get the impression that neuroimaging research forgets that measures of brain activity are proxies for brain function (for kid sib: a proxy is an indicator for something we are interested in, but not the thing itself). And importantly, they are proxies without a sufficiently good explanation tied to them (how/why does this pattern of activity ‘code’ for thinking about something specific). You could only really claim that you are actually measuring brain activity (via brain imaging) if you have an explanatory theory of how/why brain imaging reveals brain activity (and with what accuracy it does so). Researchers know that statistical correlations used in brain imaging do not imply causation about brain function, but they’ll remark that “this area seems to be implicated in X, while this area in Y” – but why, and how? Consequently, I would say that corroborated predictions involving measured brain activity (e.g. “as predicted, this area lights up when we show participants X”) feign scientific progress.
ReplyDeleteThe trouble with cognitive neuroscience is that it serves two empirical masters:
DeleteThe brain is a organ of the body, so its anatomy, physiology, biochemistry and genetics are of interest in their own right, as part of biology and medicine. Any new empirical fact about the brain is a scientific fact.
But the brain also generates all our cognitive capacity: all the things our bodies can do.
At this very early stage on the road to T3, it's easy to generate new facts about the brain in the form of correlations between brain activity (where, when) and all the things we can do.
What's missing is the how/why causal explanation. And it's not clear whether it can be bootstrapped from the correlations.
My initial impression of Fodor’s stance on neuroimaging is that it is not powerful enough to be useful to us. However, upon reading this article again, it seems that Fodor doesn’t doubt the clinical importance of neuroimaging and that is it very significant in medical practice (specifically to know where language areas are in our brain so that neurosurgeons can avoid them during surgery). Instead, Fodor doesn’t think that all of this money spent on functional localization can help us explain our cognitive capacity. Finding out where functions happen in the brain doesn’t answer why we do these functions and how we do these functions.
ReplyDeleteI think Fodor should’ve distinguished between the clinical importance and the cognitive importance of neuroimaging. For clinical use, it provides ample amount of information for treatments and interventions that we can provide in the future. For example, research about music therapy for interventions to reduce stuttering have advanced with the help of fMRI studies. But in terms of cognitive importance, neuroimaging doesn’t do much to help us solve the easy problem.
Celine, I completely agree with you that what Fodor was implying is that neuroimaging has no cognitive importance, but I really don't think he did a good job in articulating this point in his controversial article. I thought that his stance was too harsh and also reread the piece several times to see if I can derive new meaning from it. Maybe I am missing something, but he seems to stating that he sees no benefit whatsoever to functional mapping of the brain. This is quite unjust to the field of neuroscience - the field is young and it is growing. The brain is perhaps the most important organ in our bodies besides the heart, and certainly the most complex. We are nowhere close to fully understanding it but through neuroimaging, we can start to gather little pieces of knowledge to build up a bigger picture. No, neuroimaging is not the right way to go in understanding cognition. However, I do believe that if we ever successfully reverse-engineer a T4 or T5 passing machine, neuroimaging will be valuable in seeing whether its 'brain' resembles one of ours.
DeleteIf we successfully reverse-engineer T3 (Isaure), cogsci's work is already done.
DeleteI agree with Fodor that finding out where certain mental states are located in the brain is not a particularly interesting endeavor, but it is certainly a necessary one. Using his own analogy, sure identifying where the carburetor is in your car isn’t very informative if you already know how it works, but that’s not the case if you’ve never opened the hood before. If you took someone who knew nothing about cars and asked them to figure out how the carburetor works, they’re not going to get very far if they can’t even find it. “Where” may not be the answer we’re ultimately looking for, but it’s the first step in identifying “How”.
ReplyDeleteI definitely agree with that! I think Fodor is too quick to dismiss the implications that brain imaging holds for further research and clinical endeavors. He also overlooks the fact that neuroimaging, as a scientific field, is still fairly young. It would make sense that there are a plethora of neuroimaging studies being conducted at the present time, because neuroscientists and clinicians are still in the stage of "car novice" (to use your example) when it comes to functional brain localization.
DeleteThis paper was also written in 1999 – almost 20 years ago! I wonder whether Fodor would still stand by this view, following the advances and developments reached in neuroimaging in the last two decades…
DeleteI feel like the new technology wouldn't change his opinion too much. The advances in neuroimaging may allow us to study the brain more accurately, and in better resolution, but it's still just focused on where in the brain the activity is happening, and not so much how it's happening. I think that's a question that neuroimaging alone will never be able to solve.
DeleteThe goal of cogsci is to explain how the car can do what it can do. The car's function, like the kidney's, is vegetative rather than cognitive. The brain's function is to generate our capacity to do everything we can do: T3, the "easy" problem. Fodor is suggesting that knowing where the caburetor -- or mirror neurons -- are does not help explain how they generate our capacity to do what we can do. (He died a month ago, but I doubt the last 20 years of research changed his mind.)
DeleteHe was a great lover of opera, though, and cats (or, rather, a cat).
I find it contradictory that Fodor is doubting if functional localization is worthy of widespread scientific endeavors. Fodorian modularity in the 1980’s propelled the concept of modularity in evolutionary psychology and set the criteria for what can be considered a module. When he asked, “Who cares?” it is obvious that he, in fact, cared at some point in his career.
ReplyDeleteHis analogy to the carburetor’s relation to the engine is weak. The brain is a system, like any other organ. It is important and, arguably, worthy of the time/funding dedicated to localization to understand each part of the system-- it matters where because of its connectivity to other parts and how the connectivity affects the whole. He furthers his analogy by asking if the energy is functional, why we would need to know where the carburetor is. By the same logic, that would be asking why are we trying to understand cognition, if we know that we are cognizing. Needless to say, structure and function are not mutually exclusive, and in a lot of cases, understanding structure helps to explain function. Although, I do agree that sometimes there’s a leap of logic between the two.
In my opinion, I don’t think functional localization can show us the causal mechanisms for our cognitive capacity. I think it’s a building block for connectivity and even that can be a building block for something greater. Surely where and when neurons fire will not answer how and why, but it's a start.
Fodor's "modularity" was functional rather than anatomical. It was also very vague; its influence was ideological rather then empirical. But he always denied the relevance of brain structure for discovering or explaining cognitive function.
DeleteWe know we are cognizing. We want to know how. And Fodor is saying that the when and where of its correlates will not reveal how.
It's good to keep in mind the difference between vegetative (clock, car, kidney, heart) and cognitive (T3) function.
I really like Fodor’s distinction between knowing something and knowing it scientifically. I think that a lot of neuroscience nowadays is trying to learn things scientifically by actually finding the brain region responsible for its occurrence as opposed to what it was in the past which was trying to figure out exactly what the brain was capable of. Where I disagree with him is that I do not think it is a worthless pursuit. Once we find a region in the brain that, at least could be, responsible for a particular phenomenon, we can start to examine it more in-depth and we can learn more about the phenomenon by manipulating the brain. Region-localization and brain imaging research can also lead to better diagnoses for neurological diseases and eventually, perhaps lead to treatment. For example, brain imaging research has led to significantly more accurate diagnoses of language disorders and many others. Perhaps Fodor is more concerned that the correlations of some the studies seem to be spurious and that the fact that a study finds that region X lights up during task Y is not enough evidence to really localize a phenomenon to a region… In any event, I do not think that is a sufficient reason for the practice of brain-imaging to be labelled useless.
ReplyDeleteI completely agree with your point. Knowing something with proof and experimental data allows us to come to a solid conclusion about how to use this information. If we know something scientifically, for example a localized brain function that we can back up with facts and data, then we can start to help people with neurological disorders and various mental health problems.
DeleteFodor agrees about the clinical usefulness of brain imagery; what he denies is its usefulness for reverse-engineering T3 capacity. Cogsci is not looking for the locus of a correlate of cognitive capacity: It is looking for a causal (T3) explanation of how the brain (or anything) generates that capacity. There are lots of facts about the brain as a biological organ, but which ones are relevant to its causal capacity to generate T3?
DeleteFodor discusses the fundamental differences between rationalism and empiricism re: mental states and functions in the brain. He says that empiricists believe in a "equipotential" brain, where "all mental states reduce to patterns of associations," whereas rationalists are in favor of the brain being divided into distinct regions, possibly infinite, of function. My question for him is: why can't it be both?
ReplyDeleteYears of neuroimaging research has shown us that brain/mind functions are definitely localized in certain areas, sure, but there are also areas that overlap in function, or have neuronal pathways that are highly interconnected (a physical overlap, if you will). In addition, certain functions cause multiple brain regions to light up, showing that they are pretty de-centralized. I think it's limiting to choose one side in that particular argument, because there is tangible evidence in both camps; this situation is far from black-and-white, as I believe Fodor is trying to portray it.
The rationalism/empiricism in this context is bogus. Fodor happens to believe in a pseudo-Chomskian "big-bang" theory of the origin and nature of cognitive capacity: It's not "learned" (either in the organism's lifetime or in the evolutionary history of the species): it's just "there." (We'll discuss Chomsky's nativism and rationalism in weeks 8 and 9, but here it's just a Fodorian quirk.)
DeleteFodor is a computationalist of sorts, so that's his main reason for ignoring location and hardware. His "modularity" was a (vague) flash in the pan.
But the question to reflect on here is whether where/when can reveal how/why in the case of cognitive (rather than just vegetative) function (T2/T3).
Fodor expresses an inability to comprehend why the Times, and neuroscientists, are so interested in functional localization. He argues that even if we determine where/when things are happening in the brain, this does not give us an explanation as to HOW we do these things. And therefore, unless we are surgeons performing removals on certain brain areas, this knowledge is essentially useless. He explains this with an analogy of a car, "But why (unless you're thinking of having it taken out) 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?"
ReplyDeleteI do agree with Fodor that finding correlations between behaviour and neuroimaging may not give us an explicit explanation as to HOW the brain works, however I feel Fodor is overlooking the potential benefits neuroimaging technology hold. Just because it has not provided an explanation yet, does not mean neuroimaging never will. And therefore, I disagree with Fodor who implies that these experiments may be a complete waste of time.
The issue is not waste of tax-payer money but whether localization of correlates of cognitive function can lead to a causal explanation.
DeleteI for one am extremely grateful that brain imaging -- even if it is useful only clinically, and not for explaining cognitive capacity -- is sparing countless innocent animal victims from tortures that have proven completely useless for explaining cognition (and not much less useless clinically).
Here's a wonderful example of how noninvasive brain imagery can be used to answer questions about empathy in both humans and animals (though they still don't help for reverse-engineering T3): Decoding the dog's mind with awake neuroimaging.
Fodor raises some interesting points here about the intentions of scientists. From his point of view scientists are only interested in gathering information and data about whatever they can dream up. The newest technology drives which experiments they carry out, and are mainly about brain localization because that is what the newest technology allows them to do. In this way, they may be overlooking other methods and important questions to be answered. I agree with part of this statement, perhaps scientific questions regarding which area of the brain lights up when we think of a fork is one that does not require spending thousands of taxpayer dollars on, but not all scientists are like this and by localizing one thing, however trivial it may be, perhaps we are getting closer to a comprehensive brain map that allows us to treat and diagnose neurological disorders in ways we haven’t been able to in the past.
ReplyDeleteI also found his commentary of the way technology development drives research. It is definitely an easy trap to fall into, and we should discourage new researchers and clinicians from taking the easy route of allowing technology to make research seem important just because it has not been done before. Nonetheless, I think part of the excitement of new technologies is that we aren't always sure exactly what they will uncover.
DeleteWhile the funding issue is extremely sensitive, and perhaps funding allocation should be more selective, without testing the boundaries of new technologies, we may be missing something fundamental. Since this paper was published, brain imaging and localization studies have contributed immensely to both clinical and cognitive research.
Fodor does not deny the clinical value of brain imaging, just its value for cognitive reverse-engineering (T3).
DeleteThe virtue of noninvasiv brain injury is that it does not hurt anyone. And lots of innocent animals have been (and are being) hurt for what is not only also a waste of taxpayer money, but is mostly driven by career, curiosity, funding or fashion rather than a genuine prospect of saving lives (which would be the only justification for hurting animals).
I understand why Fodor may feel confused at the rising trend of neuroimaging to find out more about the brain. Of course he’s confused, this isn’t his area of expertise. Sure, a lot of money is being spent, but he could easily have written about the United States’ absurd military budget. Why are they spending billions of tax dollars on war weapons? Neuro imaging is an incredible concept that we could barely have imagined to be possible 50 years ago.
ReplyDeleteYes, finding out where the concept of a tea pot lights up in the brain sounds dull, but maybe the patient was in a vegetative state and they were trying to figure out whether these patients are aware or conscious. Maybe knowing where thoughts are localized can help understand human intelligence better? Maybe Fodor’s an old bat who should understand that the only way change happens is through innovation, however expensive that may be.
Isaure quite literally took the words out of my mouth, as I don't think I could have expressed how I felt reading Fodor's article any better than she did. To add on to her point, I find Fodor's comparison of finding the functional localizations in the brain to that of certain parts of the engine -- in this case the carburettor -- in a car to be naïve and too simplistic. He asks "But why (unless you’re thinking of having it taken out) 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?". Of course Fodor doesn't care where the carburettor is, because he is not a mechanic. However, if he gets into his car one day to find that it isn't running how he expects it to, he should hope that his mechanic knows where the carburettor is. If one day Fodor experienced a severe trauma and suddenly was unable to produce speech like he had in the past, I believe he would be grateful that someone spent the time and money to find out where speech occurs in the brain in order to possibly help him regain his abilities, or to ensure that they don’t damage his brain further…
DeleteTo suggest that neuroimaging techniques are only useful during surgery is too simplistic. We cannot pick and choose when something like this might be useful or not, it is either useful (though perhaps in some areas more than others), or not useful. Either we do the research, or we don’t. We cannot "half" do the research.
Fodor (who died last month) is not writing about wasting taxpayer money (and he does not criticise clinical uses of brain imagery). You have to see through the stylistic devices. He's saying that brain imagery won't help up explain intelligence (cognition).
DeleteCars are forward-engineered. They were built by us; that's why mechanics know what to do. Brains need to be reverse-engineered. (The mechanic was not someone who had never seen a car and been given one and told: "Figure this out" -- though perhaps he could eventually do so, because automotive function is a lot more like vegetative function, whether of the heart or the brain [balance, temperature, breathing...]).
I agree with most of the points brought up by Fodor on the futility of brain imaging and localization, however, only in the sense that isolates brain imaging from other domains. In fact, if we only look at brain imaging, the current brain imaging scientific literature is unbelievably messy that there is contradictory findings on every subject without any consistent data. However, this phenomena only shows us the limitation, the extend of what brain imaging can tell us, that we can only get a vague picture of different brain functions and their corresponding localizations, but never truly understand the brain. Meanwhile, correlation does not mean causation. If the brain lights up under certain task, does it corresponds to the task, or just a random thoughts or processes that were executed by the brain at that moment.
ReplyDeleteHowever, this is what we would get from looking only at the brain imaging studies, and ignoring other more detailed approaches on explaining our brain, and this is where I disagree with Fodor because isolation of one particular approach, trying to boil down its characteristics into simple terms without paying attention to the links and branches that this technology has with other methods is very arbitrary and ignorant.
For instance, although brain imaging does not indicate causation. Combined with case studies, we would get a sense of which part of the brain is responsible for doing what kind of task. For example, patient H.M., a epileptic patient who revealed the secret of hippocampus, inspired countless following researches that try to explain memory and learning. It shows a fairly consistent causal relationship that without hippocampus, patient HM cannot encode any episodic memory! Furthermore, with the help of fMRI, it makes it easier for scientists to pinpoint the general location of the neural circuitries that are responsible. With the help of fMRI, scientist can zoom into the biological process of the brain, instead of wonder around purposelessly, and get lost in the huge mesh of billions of neurons in the brain. And I think this is what brain imaging contributes to scientific literature.
Meanwhile, Fodor's understanding on brain is quite shallow. Brain is neither just about equipotentiality nor just localization. The brain is a highly plastic and interactive machine that changes in response to different stimuli. Neurons react to different signals based the nature of them, send the signals to other areas and get information feedback from connecting areas to make modifications. Knowing where is the process is not the only and ultimate goal of brain mechanism, but the first and necessary step to decipher the brain.
You need to ask yourself again: How can "where" reveal "how"?
DeleteIf planes, instead of being built by engineers, grew on trees, would we figure out how they fly by imaging where the correlated activity occurs when the plane is flying? (Perhaps, but if so, only because planes, like hearts, only perform vegetative functions, where there is a clear relation between the structure and the function. The brain's cognitive function is everything T3 (Isaure) can do.)
As others have pointed out, Fodor questions whether knowing where and when certain cognitive processes occur in the brain helps us understand how they occur. Clearly he thinks the answer is no. In other words, he doesn’t think that looking at the pattern of firing of neurons will tell us anything about how those neurons produce thought. First, I have some objections to how he argues that point of view.
ReplyDeleteHe describes neuroimaging scientists as being either obsessed with identifying brain regions which respond to specific objects and narrow classes of objects or observers of people performing entirely arbitrary tasks and taking pictures of the resulting activity.
I think this grossly oversimplifies the work a lot of these people do. Scientists aren’t excited by something like the discovery of a teapot-specific module, especially not for its own sake. For one thing, neuroimaging studies do not, as he implies, pick stimuli and tasks at random. They are often chosen with a specific theory or hypothesis in mind. “We think this part of the brain does x and not y.” The stimuli and tasks are chosen to identify where x ends and y begins. What is the shape of x? What conditions does x occur or not occur in?
Further, these studies aren’t conducted in a vacuum. No individual study of functional specialization claims to singlehandedly unlock the secrets of how a particular brain region cognizes. The point is to make incremental contributions to solving a more complex problem. Maybe knowing which cluster of neurons fires in responses to vegetables isn’t particularly interesting on its own. However, if we’re trying to figure out how the brain categorizes, or how it extracts meaning from language, then that cluster becomes relevant.
But none of the above gets to the heart of his take on the question. Is the study of functional specialization at all useful in our quest to understand how the brain produces thought? I strongly disagree with Fodor’s point of view. If what we really want is to figure out how we get from neurons to thought, then isn’t it crucial to know what things are happening where, and when?
Isn’t it important to our understanding of memory to figure out that the hippocampus doesn’t store memory, but rather aids in the consolidation thereof? Isn’t it relevant to know that there’s functional specialization within the hippocampus supporting different types of memory? If you want to know how cells produce thought, it’s necessary to know which cells are relevant to whatever kind of thought you’re talking about. You can’t know how a car works if you don’t know what the parts are or where they go.
Neuroimaging studies have contributed to the identification of place cells in the hippocampus, which contributed to O’Keefe’s theory of the hippocampus as a cognitive map. So not only does it contribute to our physiological understanding of thought, but even of the cognitive framework and processes in question.
In that sense, I’m not sure what neuroimaging even leaves out. If it tells us (maybe indirectly) about the physical building blocks of the stuff that produces thought AND what the output of that stuff looks like, what hasn’t it done? How are we supposed to figure out the way something works, other than looking at which parts do what at which times?
You may be right that what we know about the brain will help us reverse-engineer T3, but do you know of any capacity that we had no idea how to generate until we found something in the brain that revealed how? Motor neurons, for example: do they reveal how to generate mirror capacities?
Delete"How are we supposed to figure out the way something works, other than looking at which parts do what at which times?"
Good question, but when it comes to cognitive rather than vegetative function it's not clear that the strong C/T (Weak AI) is not the more promising option.
Fodor discusses and breaks down the rationalization of the majority of neural research focusing on brain localization. He brings up the arguments of both empiricists and rationalists as a way. To try. And understand why knowing where things occur in the brain is important. For empiricists, they view the brain as 'equipotential,' that is, that the brain can perform different functions in different places, whereas rationalists preferred the brain to be organized in different geographical areas. The question he brings up though is if we know there are places in the brain for certain functions, why does it matter where exactly these places are in the brain?
ReplyDelete"She runs into Pavlov, who explains to her why he is, rather horribly, drilling holes in the mouths of dogs: it's to show that expecting food makes them salivate. 'But we already knew that,' she says, in some perplexity. 'Now we know it scientifically,' Pavlov replies" (Fodor, 1999).
Overall, while Fodor makes some valid points concerning the potential lack of necessity to understand where all functions occur in the brain, I do think he tends to undervalue the importance of brain-imaging. Brain-imaging is used in order to understand where certain functions and neural connections are carried out in the brain, yes, but these findings have incredible clinical implications that are important in several cases, such as in individuals with brain damage. Without knowing where these things occur in the brain, how will we explain when a certain function is gone? How will we anticipate what effects a metastasizing tumour will have if it grows into neighbouring areas? I think Fodor is too quick to dismiss the benefits of brain-imaging in his critique.
I agree with you Tina. While perhaps it may seem unimportant to "find a brain region specific to thinking about teapots or to taking a nap", having a functional account of the brain will not only aid our clinical decisions and treatments, but during the process of localization - we may discover something of great clinical importance that was not within the original scope of our research.
DeleteIn addition to this, I think that it matters to learn where certain functions are located because it can give us insight into the genetic and associative aspects of the brain (as discussed in the Cook et al paper). During the process of investigation, we may find ourselves more equipped to answer questions about how the brain is accomplishing cognizing about tea pots or naps.
I also agree. I think brain imaging is necessary for the advancement on knowledge in the field. Because of this technology, so much research has been achieved and more treatments and measures have been sought out due to the advancements on functional localization. Although expensive, the benefits regarding research itself, definitely outweigh the costs.
DeleteFodor agrees on clinical uses. Does brain imagery tell you how mirror neurons cause the mirror capacity they are correlated with?
DeleteI find Fodor’s perspectives on scientific experiments quite interesting. I think it’s normal that Fodor has such impressions on experimental techniques since first, he is not a believer in neural localization, and second, neuroimaging will not crack the code of the mind. As a philosopher, I think his interest in brain lies in the discussion of the mind-brain relation, which is essentially the hard problem in Cognitive Science. However, most neuroscientists, or at least those neuroscientists at the Cognitive Science lecture which he went to, focus more on discovering brain mechanisms and correlated cognitive functions. In my opinion, Fodor is not aware of the fact that brain is the most sophisticated part of the body, and Neuroscience is still in the developing stage. Therefore, the current primary emphasis in this area of research is to study the brain structures and related functions. Furthermore, most of the knowledge we have about the brain is from Neuroscience experiments with the assistance of neuroimaging, and understanding the physical brain, is a necessity to understand the mind. Hence, the contribution of neuroimaging cannot be simply put as making brain maps.
ReplyDeleteWhat about Fodor's questiion about how to get from where to how?
DeleteIn this reading, Fodor questions the sudden surge and popularity in scientific research of localizing functions in the brain. According to him, if a human (or machine) is cognizing and we can recognize it to be cognizing, why is it important to know the physical structure of its mechanisms? He claims that just because two concepts are similar in function does not mean they will be represented in close proximity in the brain, or have similar brain locations. Similarly, just because humans and machines do not have the same mechanisms underlying their functions, does not mean they cannot be executing the same functions. This is a form of the implementation independence argument, meaning the hardware doesn't matter, only the software does. However, in light of the mirror neuron article, it's hard for me to conceptualize a machine that integrates sensorimotor experience and can execute tasks based on that input, something that was crucial to the acquiring and functioning of mirror neurons according to the authors.
ReplyDeleteYou are right that analog capacities go beyond computation. They (but not their sensorimotor interfaces) could be simulated internally by computation, however.
DeleteSo empiricists, since they typically hold that all mental processes reduce to patterns of associations, would like the brain to be ‘equipotential’, whereas rationalists, since they 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.
ReplyDeleteIn his work Fodor portrays these two ways of thinking as completely separate, however it is not that simple (just as most things are not) when it comes to looking at the reality. Despite this, I think Fodor brings up some interesting points that we can use for modeling the brain. For example, a large part of his paper is spent questioning the reason as to why we study cognitive science and if the mappings that many scientists have dedicated their lives towards finding truly matter. I do understand how someone like Fodor (who describes himself as a philosopher) might not recognize the value of investing in brain imaging technologies. On the other side, he questions the science that seeks to explore the domain he is passionate about. After all philosophy is a sophisticated type of thinking and reasoning, which if we find most clearly in human beings. Therefore I find it ironic that Fodor is not more intrigued by the results of this research.
Fodor is a computationlist who holds that the hardware is irrelevant, and does not reveal what software it is executing. (This is true of a computer too: "Cyberimagery" would not tell you what program is being executed.)
DeleteFodor seemed to imply, when describing localization of brain functions, that the exact same brain regions will light up in all individuals when they do the same task, for example, thinking about teapots. Of course there are certainly many brain regions common to all individuals that are crucial for a specific task (think of all of the specialized regions for speech comprehension and production) but it is known that brains vary structurally among individuals (this is why extensive brain mapping is done prior to and during brain surgery). Considering that this article was written in 1999 (nearly 20 years ago, wow) I wonder if Fodor would change his account of brain function localization. Perhaps his account of over-generalizing localization is due to a misunderstanding on his part, or a misunderstanding in the greater neuroscience community 20 years ago.
ReplyDeleteYour example is clinical, not cognitive.
DeleteA month and a half ago I said “Perhaps his account of over-generalizing localization is due to a misunderstanding on his part, or a misunderstanding in the greater neuroscience community 20 years ago.” Now that I have a better understanding of localization as an approach to neuroscience, I retract my statement because localization was widely accepted; it was not unique to Fodor. Localization was widely thought to be the key to understanding the brain. It was thought that by mapping the brain and its functions, we could ‘solve’ it. Localization was shown to be wrong because it would still wouldn’t be possible to duplicate the brain no matter how extensively it was mapped.
DeleteI also admit that my example given in the previous skywriting is clinical, not cognitive. However, Fodor’s research and the clinical questions that come with them are still relevant to this discussion because they show what doesn’t solve cognition. Failed clinical examples and incorrect approaches still push us towards discovering what exactly cognition is by showing us what it is not.
As Fodor said it, he is not a neuroscientist, but a philosopher. He may not realize the implications of being able to localize brain functions. Although we might not always be right when stipulating that region X of the brain does function Y, this error-and-trial process is part of science in my opinion. Furthermore, when Fodor says, “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?”, I think he is not understanding neuroimaging correctly. What scientists are doing is not precisely finding a one to one match for every brain function. Rather, neuroscientists look for patterns of activation that taken together represent a function. Fodor was afraid that we might never find a region for thinking of specific things, but I believe we are already farther than this. Here is an example of a study that was able to reconstruct images using the patterns of activation from the visual cortex of participants: https://www.sciencedirect.com/science/article/pii/S0896627308009586. To some extent, this shows that we can be confident that activation patterns represent brain functions.
ReplyDeleteIf where and when an activity occurs will not explain how it works, how do patterns of activation explain it?
DeleteThis “diary” definitely elicited my scepticism of neuroresearch. We all just assume research is innovative, progressive, pertinent etc etc. Attach the words “neuro” and related or even unrelated jargon, much to the nescience/awe of the public – it siphons the clicks, conversation, excitement. Much of it seems pseudo, but much of it an automated response. Most people have internalised that anything brain research is “cool”. But is it? Have we become less critical of the pertinence of contemporary brain experiments because “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.”? So much funding is funneled on finding which part of the brain gets more glucose when I think of a tennis racquet, amazing, ok so what?
ReplyDeleteThe other side though is that it is valuable knowing such things due to it’s use in surgery or treatment. Like a surgeon knowing to avoid broca’s area unless he want’s language to be disrupted (even then it’s not certain, it’s never that simple when it comes to the brain) or stimulating the anterior cingulate to generate more deep recovery sleep in elders because that’s where it is believed to propagate from. Emphasis on believe because we don’t know for sure, will we ever? Maybe that’s what neuroscientists are chasing? Nonetheless, because the brain is so complex and our functions are increasingly found to be networks of mechanisms rather than an output from a certified input center (again, why does the center matter? Does it help better understand the input?) we orbit back to the very little usefulness in knowing that a certain area is more active during a particular function. An example is cerveau isole: transection at the midbrain that isolates the forebrain and/or mid-pontine lesions. Despite these areas being the most active for sleep/arousal circuits and thus would result in disrupted circadian rhythms, the brain recalibrates and self generates new circadian rhythms. So we have confirmed that “yeah this part is kinda important but not absolute”. Thus, as the writer points out, the serendipitous “you can never tell, it might pan out” strategy in research hasn’t affirmed or enlightened much at all.
However, I think his reductionism of neuroimaging is not made valid just because neuroimaging hasn’t confirmed much yet. It is still a new field and each finding builds a more concrete picture of mental functions, even if only affirming the complexity of brain functions. All the more reason to keep looking right? Knowing where and how certain functions take precedence helps with diagnosis and treatment if not just understanding how the brain works, those are immense motivations. Knowing that it’s the dopaminergic neurons in the basal ganglia that need to be regenerated to elevate Parkinson’s symptoms is critical to finding how it can be done. The very reason why neuroimaging is so pertinent is because of the monumental uncertainty about the brain. Everything we learn about it is useful, and where the carburetor matters to the mechanic not to the average person. Just as where neural function is matters to the neuroscientist, not the average person or a philosopher (because those are totally not the same).
Cognitive, not clinical.
DeleteI find Fodor's ideas interesting, is it really necessary to know where in the brain the teapot is? The answer is no certainly not. However I think when this was first discovered people were so incredibly interested in the technology and thought that investigation could lead to something more, maybe if you can map the location of all of our behaviors and observations it will lead to further understanding of the mechanisms of the brain. And Fodor raises good points that this is certainly a leap and seems like a random treasure hunt. However in the last 20 years neuroscientists themselves have come to many of the same conclusions as Fodor and this phrenology investigation has halted.
ReplyDeleteI don't really agree on how Fodor thought that knowing where in the brain the teapot is, is unnecessary and how the technology used to discover this is expensive. Yes, okay, maybe it is expensive but it is definitely necessary to explore the different parts of the brain and its implications in the name of research. Without the exploration of these technologies and the brain, how would we know how one area affects a certain behaviour. How would we go about approaching certain disorders?
DeleteThe question is not clinical but cognitive.
DeleteTaking this article in context of its publication in 1999, Fodor’s opinion may have been reasonable (I’m not aware of the neuroscience climate of the 90s) but has now become inappropriate. In the past two decades, brain imaging tech has become more functional and affordable, and has facilitated many non-trivial scientific breakthroughs such as the use of fMRI to communicate with a comatose patient [1] which was achieved precisely through using knowledge of where particular brain functions occurred. This use of fMRI shows that at least some comatose patients are conscious[2]! Plus it has allowed for improved medical knowledge for recovering consciousness in humans that suffered brain damage [3].
ReplyDeleteWhile this does not causally show “whether mental functions are neurally localised in the brain,” (and here I agree with Fodor that identifying where mental processes are localised would not answer the former question), the study of the arrangement and brain could provide key insights to that lead to the answer to that question while probably helping many medical patients in the process.
Also, relating to the discussion on the associative learning hypothesis of mirror neurons, it would be interesting to continue the study of MNs on comatose patients who no longer have the sensorimotor capabilities for the associative learning described in the Cook article.
[1] https://www.sciencedirect.com/science/article/pii/S147444220400852X
[2] https://www.nature.com/news/neuroscience-the-mind-reader-1.10816
[3] https://www.sciencedirect.com/science/article/pii/S1053811911014431
Your example is clinical. Fodor agrees. What about reverse-engineering cognition (T3)?
Delete"I don’t know who’s right about all that, but it’s easy to see that whether mental functions are neurally localised is likely to be relevant. If the brain does different tasks at different places, that rather suggests that it may do them in different ways. Whereas, if anything that the brain can do it can do just about anywhere, that rather suggests that different kinds of thinking may recruit quite similar neural mechanisms."
ReplyDelete- I understand Fodor's point here, and this seems the nucleus of his argument so I will attempt a rebuttal for those who have spent countless hours of hard work in the lab localizing functional areas in the brain... I find that Fodor undervalues the work of brain localization in examining the functional connections between brain areas. Localization matters - this can easily be appreciated in areas such as the amygdala, which takes advantage of its location proximal to memory structures like the hippocampus to create emotional associations with the memories we store. Further, the location of the parietal cortex which is postulated to include the locus of our human "lexicon" is in a prime location for combining information processed in primary auditory, visual and somatosensory cortices in neighboring areas. The physical layout of the brain is equally important to understand in the case of a brain tumor or medical complication. Localizing the most crucial brain areas can help save the functional capabilities of those who must resort to neural surgery to mitigate health complications. The research of brain localizers can be viewed through the lens of empiricism vs rationalism but it is also important to understand the geography of our brains to establish the building blocks for further research and in ensuring better solutions for neuro-degeneration.
In this article, Fodor implies that we shouldn't 'waste' our resources on localizing the areas of the brain that carry out specific mental functions. He doesn't see a point in knowing "how far north" the mind happens, as long as we know it happens "somewhere north of the neck". Despite the fact that I can acknowledge that we won't always find a direct cause and effect between our mental functions and our brain structure, I believe that Fodor's argument fails to take into account the importance of such information upon the development of our capacity as intelligent creatures, and the surpassing of multiple conditions that affect the brain in a debilitating manner. In a very simple example, knowing that the comprehension of speech occurs mainly in Wernicke's Area, in the temporal lobe on the left side of the brain, has significantly clarified the surrounding issues and helped us understand the functioning of language, aiding those who suffer from such conditions. Consequently, despite being more geared towards a philosophical approach myself, I have to disagree with Fodor's absolute rejection of brain imaging and the development of neuroscientific methods to uncover the mysterious functions of the brain, as it has proved to be very useful and has already expanded our understanding of it to a significant degree.
ReplyDeleteI did take this argument on the clinical side, so from a cognitive perspective I can see how it isn't as useful a practice. However, I still believe that important information can arise about our cognition--the way information is transmitted, stored, and so on. My response still stands, but maybe to a lesser extent.
DeleteFodor touches on very important points in the critique of functional neuroanatomy. Chief among them I would argue would be the focus on the result rather than the full explanation of why they occur “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.” This comes from a base in which the available technology drives what science is done, since neuroimaging is the most advanced technology available that shows consumable results scientists may lose focus on the process and focus on the picture that they can provide for the consumption of science. This leads to “knowledge” about where functions might be located based on the paintings of neural activity, not really progressing to our actual understanding of how things are processed (cognition and other things). There needs to be either a refocusing into a more comprehensive form of research that aims to start from the ground up of understanding more, or a unseen technological advancement that will fill the gap in a manner that I can’t conceive of.
ReplyDeleteI think that Fodor makes some really interesting points about localization of function and the distribution of taxpayer dollars in scientific research. First, I don't think it necessarily matters where exactly a certain function takes place in the brain but it does matter that there IS a place where a function occurs, and I think that there should be a more varied distribution in where research dollars go. So, the fact that language is seen to activate regions in the frontal lobe, to me, does not feel like a big deal. But the fact that it does have a specialized location does provide input as to how the brain works especially since language has a different place in the brain when it's your second language rather than your first (noting that it must be an L2 rather than a second L1, and the extent to which the location is different is dependent on the age at which you learn the given language). That said, all humans are subject to their own different experiences in the world which effect their brains in different ways such that neural plasticity molds their neuronal connections differently to those of other people. So different processes or stimuli may provide a more robust neuronal response (in brain imaging) in one person than another because of this neuronal plasticity, but these different kinds of responses further help us to understand how the brain works simply because they are different. If we see that people who are simultaneous bilinguals activate Broca’s area while using both their L1s and we see that sequential bilinguals or late bilinguals use an area that is different or larger than the region used for L2, we may be able to determine what kinds of mental resources are used for learning a language at different stages in life and determine how some people are able to achieve native-like competency while others are not based on the brain regions involved in their language processing.
ReplyDeleteFodor is not trying to say that scientific research is useless or that tax payer dollars are being wasted. Instead, he is saying that we are trying to answer the "how" question while using "where" methods. In other words, instead of trying to see how the brain works we are finding the specific areas that light up during different activities which doesn't help us with the question on how they light up to begin with. Thus, looking at the brain in order to answer the how question is what he would consider futile
ReplyDeleteI agree with Fodor’s point that functional localization by neuroimaging teaches us very little in terms of cognitive science. Indeed, we can see areas light up, but we get no information about what they are connected to and how the networks operate. However, I think that Fodor is underestimating the clinical utility of these studies. I concede that knowing where the “thinking about teapots” area is is not clinically relevant, but knowing with great precision where language areas are, for example, can be critical in performing surgeries without debilitating consequences for patients. Even though functional localization studies cannot teach us how the brain functions, the fact that they can contribute to a better understanding of the structure of the brain and thus increase neurosurgical precision and safety is reason enough to keep funding them.
ReplyDeleteFodor compares the mind/brain with an engine, asking what part of how your engine words have you failed to understand if you don't know where in the engine the carburettor is. I think it is an interesting comparison which definitely drives home his point, but also I think it seems a bit far fetched to compare something so complex as the human mind to a car engine. Where in the engine its different parts are isn't relevant to the functioning of the car because each part has one function irrespective of where in the engine it is. In the brain, thousands of neurons have functional connectivity to other thousands of neurons either nearby or far away, and the geographical principles/topologies of the neurons are related to their function based on the surrounding neurons and their functions. I also agree with Myriam's point above, that he also doesn't understand the clinical utility of these brain imaging and brain mapping studies. We know that Broca's area is consistently in an area of the frontal lobe, and during surgery, knowing where this is (like Fodor says) will cause the surgeon to avoid it. So obviously knowing where certain neural correlates, some that might be more important and more critical to avoid than others, are in the brain.
ReplyDeleteFrom the comments here it’s been made apparent that although the benefits of neuroimaging are abundant for clinical purposes, they’re not clear in regards to cognitive insight. The question that keeps coming up is how does knowing where/when things happen in the brain help explain why/how they happen? It seems to me that Fodor isn’t offering more to the discussion than to point out that scientists might be looking in the wrong place. It’s certainly clear that we don’t know how/why things like mental states occur, but are we any better off knowing now, thanks to Fodor, that PHILOSOPHICALLY we don’t know? I’m not familiar with much of Fodor’s work (perhaps this journal entry is auxiliary to his philosophies), but does he offer an alternative mode of research to develop cognitive theories? From my surface introduction to his work, even the Language of Thought and “categories are innate” theories don’t do much to explain HOW they come to be. My issue with the argument "but this doesn't tell us much about how/why we can do what we do" is that it can be applied to most of cognitive science... If we base the value of some research on this criteria, we might as well abolish all work until someone comes up with the perfect idea. From my understanding of scientific progress, this isn’t the way to do it. You kind of need to work your way through all the wrong answers on your way to the right one (— is there a “right” answer in this case?)
ReplyDeleteIt is true that when it comes to understanding cognition we’re not forward engineering the car, but if we needed to develop a mode of transportation, and we happened to have these things we’ve been using, left over from our ancestors, that fulfill that purpose (but we don’t know how), wouldn’t it be useful to take a peek under the hood? From what I understand, neural networks draw inspiration on its biological counterparts. Couldn’t neuroimaging give insight into the coordination that occurs across various regions of the brain during specific tasks? An example: let’s say we learn that when you’re trying to infer what others are thinking, you’re using parts of the brain that correspond with this this and this process, so when trying to reverse engineer that capacity, it might give indication as to where to start developing a solution, or where to connect certain features. It might not provide a solution to the how/why question yet, but the answer might be realized eventually through an iterative approach that draws from a varied background of disciplines. Is that too naive a belief? Maybe I’ve been drawn in by the mode of thought (especially in the age of big data) that all data is good data. In response to Fodor’s criticism of serendipity, what IS cognitive science BUT “hey it might pan out let’s give it a shot?” Aren’t we all kind of naive when it comes to the how/why question about the brain?
I am conflicted about the nature of Fodor’s argument in his article. On one hand, I agree that perhaps too much money is being spent on neuroimaging and research that does not contribute much in terms of truly understanding the brain and what it does, and serves only to make correlations. It seems to me that Fodor does not think it's a completely useless endeavor due to the many medical applications that have been developed out of the research done in Neuroscience. Regardless, I do agree with him that in terms of revealing the depths of why and how our brain does the things is more important towards actually understanding cognition and our brains cognitive processes as a whole. Meanwhile, neuroscience research seems to have a much more practical approach towards medical application, where the end goal is to apply our knowledge of the “when and where” of the brain towards solving medical problems.
ReplyDelete