Calilasseia wrote:Wilberforce1860 wrote:There haves been a variety of references to this research in this thread, as if everything is all resolved and well understood. Yet when a specific paper has been presented, there have always been `.
Lies. Let's take a look at this shall we, by reference to some of those papers I've been talking about?
…. <some dialog omitted>Miwayaki et al, 2008 wrote:More interesting are attempts to reconstruct subjective states that are elicited without sensory stimulation, such as visual imagery, illusions, and dreams. Several studies have suggested that these subjective percepts occur in the early visual cortex (Kosslyn et al., 1995), consistent with the retinotopy map (Meng et al., 2005; Sasaki and Watanabe, 2004; Thirion et al., 2006). Of particular interest is to examine if such subjective percepts share the same representation as stimulus-evoked percepts (Kamitani and Tong, 2005, 2006; Haynes and Rees, 2005). One could address this issue by attempting to reconstruct a subjective state using a reconstruction model trained with physical stimuli. The combination of elemental decoders could even reveal subjective states that have never been experienced with sensory stimulation. Reconstruction performance can also be compared among cortical areas and reconstruction models. Thus, our approach could provide valuable insights into the complexity of perceptual experience and its neural substrates.
What was that you said above again, about "tremendous unknowns, and things yet to be determined, and unanswered question after unanswered question"?
Let’s try to boil this down a little. The thrust of this research appears to be to use certain techniques to detect when someone is visualizing a predefined image. It appears that these images are still pictures, but it is easy to get lost in the technical jargon. And that there has been some success in doing this. 100% accuracy under certain conditions, and 10% accuracy or better under other conditions, although the specifics are not clear to me from the excerpts presented.
So far, this appears no different than what we have discussed previously. And I have said, this is impressive, but only the barest of beginnings. Now, in the above paragraph, the authors project where their research might take them (“Our approach could provide valuable insights, etc.”). Insights into what? For example, ”More interesting are attempts to reconstruct subjective states that are elicited without sensory stimulation, such as visual imagery, illusions, and dreams.” That would be pretty interesting. I see no claims of success yet in these areas.
Calilasseia wrote:
What was that you said above again, about "tremendous unknowns, and things yet to be determined, and unanswered question after unanswered question"?
Aren’t you just making my case? This literature documents what has been accomplished, and what has not. Don’t “subjective states elicited without sensory stimulation, such as visual imagery, illusions and dreams” remain tremendous unknowns? What am I missing?
What game? And how is it over? I see tremendous unknowns. The authors of the above paper see tremendous unknowns. Don’t you see tremendous unknowns?
A simple question to ask at this point, though, is this: If you want to know what someone is thinking, why don’t you just ask them?
Calilasseia wrote:Reconstructed images from all trials of the figure image session are illustrated in Figure 2A. They were reconstructed using the model trained with all data from the random image session. Reconstruction was performed on single-trial, block-averaged data (average of 12 s or six-volume fMRI signals). Note that no postprocessing was applied. Even though the geometric and alphabet shapes were not used for the training of the reconstruction model, the reconstructed images reveal essential features of the original shapes. The spatial correlation between the presented and reconstructed images was 0.68 ± 0.16 (mean ± s.d.) for subject S1 and 0.62 ± 0.09 for S2. We also found that reconstruction was possible even from 2 s single-volume data without block averaging (Figure 2B). The results show the temporal evolution of volume-by-volume reconstruction including the rest periods. All reconstruction sequences are presented in Movie S1.
Oh, so even though they didn't use geometric and alphabetic shapes during the neural net training of their computational model, the model still recognised them.
This might be impressive, if it was clear what was being discussed. Perhaps you would be willing to cut through the technical jargon and explain this in plain English.
Calilasseia wrote:
Once again, I see no "tremendous unknowns, and things yet to be determined, and unanswered question after unanswered question" in the above.Miwayaki et al, 2008 wrote:Image Identification via Reconstruction
To further quantify reconstruction performance, we conducted image identification analysis (Kay et al., 2008; Thirion et al., 2006) in which the presented image was identified among a number of candidate images using an fMRI activity pattern (Figure 3A). We generated a candidate image set consisting of an image presented in the random image session and a specified number of random images selected from 2100 possible images (combinations of 10 × 10 binary contrasts). Given an fMRI activity pattern, image identification was performed by selecting the image with the smallest mean square difference from the reconstructed image. The rate of correct identification was calculated for a varied number of candidate random images.
In both subjects, image identification performance was far above the chance level, even up to an image set size of 10 million (Figure 3B). The identification performance can be further extrapolated by fitting the sigmoid function. The extrapolation suggests that performance above 10% correct could be achieved even with image sets of 1010.8 for S1 and of 107.4 for S2, using block-averaged single-trial data. The identification performance with 2 s single-volume data was lower than that of block-averaged data, but was still above the chance level for a large number of candidate images (above 10% correct with image sets of 108.5 for S1 and of 105.8 for S2).
In the following sections, we examine how multivoxel patterns and multiscale image representation, critical components of our reconstruction model, contributed to the high reconstruction performance.
So their model was able to isolate correctly the requisite image seen, out of a possible data set of tens of millions of images.
Once again, I see no "tremendous unknowns, and things yet to be determined, and unanswered question after unanswered question" in the above.
I will attempt to translate into plain English what I make up is being said here. There are a lot of images that they have on file (up to 10 million). They present an image to the subject. They use their scanning technique to look at what is going on in the brain of the subject. They then use the data gathered to record what the subject saw. They do a database lookup based on the brain scan, and viola, they pick out the picture they showed to the subject! And they are right “far above chance” level. And they think they could be right 10% of the time using some other parameters, but I will freely admit I am not sure what those were. Perhaps you could clarify.
So again, we have picture experiments. I am not sure what “far above chance level” is, but they are on to something. Doesn’t detecting thoughts which are NOT pictures remain a tremendous unknown? It seems so to me…
And isn’t asking people what they are seeing or thinking far simpler (and more useful)? That also seems so to me.
Calilasseia wrote:
Once again, I see no "tremendous unknowns, and things yet to be determined, and unanswered question after unanswered question" in the above.
Are we reading the same paper?
Calilasseia wrote:
Oh, so the authors not only demonstrated that their model correctly determined the nature of images seen, extracted the correct images from the fMRI data, and matched what was seen against a database of several million candidate images, but conducted a detailed analysis of the error profiles of different parts of their model, and demonstrated that the integrated whole minimised the errors arising at each scale in the final, combined result.
Once again, I see no "tremendous unknowns, and things yet to be determined, and unanswered question after unanswered question" in the above.
Perhaps you do you not see any unknowns because you aren’t looking for any? And I don’t think they claimed 100% accuracy for a database above 100 images. The precise claim for the 10 million case was unclear. Somewhere between 10% and “far above average”. There is plenty of room for the unknown there, it seems to me.
Calilasseia wrote:
The authors saved the best for the end, however:Miwayaki et al, 2008 wrote:More interesting are attempts to reconstruct subjective states that are elicited without sensory stimulation, such as visual imagery, illusions, and dreams. Several studies have suggested that these subjective percepts occur in the early visual cortex (Kosslyn et al., 1995), consistent with the retinotopy map (Meng et al., 2005; Sasaki and Watanabe, 2004; Thirion et al., 2006). Of particular interest is to examine if such subjective percepts share the same representation as stimulus-evoked percepts (Kamitani and Tong, 2005, 2006; Haynes and Rees, 2005). One could address this issue by attempting to reconstruct a subjective state using a reconstruction model trained with physical stimuli. The combination of elemental decoders could even reveal subjective states that have never been experienced with sensory stimulation. Reconstruction performance can also be compared among cortical areas and reconstruction models. Thus, our approach could provide valuable insights into the complexity of perceptual experience and its neural substrates.
What was that you said above again, about "tremendous unknowns, and things yet to be determined, and unanswered question after unanswered question"?
This is not a result, but a research topic, is it not? To boldly go where no man has gone before? And hopefully get someone to fund it? You seem to imply it is a foregone conclusion that success is sure. To me, this is simply arguing from the future.
At best it is “game on”. Go for it, my friend!
Calilasseia wrote:
Now let's move on to this paper:
Reconstructing Visual Experiences From Brain Activity Evoked By Natural Movies by Shinji Nishimoto, An T. Vu, Thomas Naselaris, Yuval Benjamini, Bin Yu, and Jack L. Gallant, current Biology, 21: 1641-1646 (11th October 2011)Nishimoto et al, 2011 wrote:Summary
Quantitative modeling of human brain activity can provide crucial insights about cortical representations [1, 2] and can form the basis for brain decoding devices [3–5]. Recent functional magnetic resonance imaging (fMRI) studies have modeled brain activity elicited by static visual patterns and have reconstructed these patterns from brain activity [6–8]. However, blood oxygen level-dependent (BOLD) signals measured via fMRI are very slow [9], so it has been difficult to model brain activity elicited by dynamic stimuli such as natural movies. Here we present a new motion-energy [10, 11] encoding model that largely overcomes this limitation. The model describes fast visual information and slow hemodynamics by separate components. We recorded BOLD signals in occipitotemporal visual cortex of human subjects who watched natural movies and fit the model separately to individual voxels. Visualization of the fit models reveals how early visual areas represent the information in movies. To demonstrate the power of our approach, we also constructed a Bayesian decoder [8] by combining estimated encoding models with a sampled natural movie prior. The decoder provides remarkable reconstructions of the viewed movies. These results demonstrate that dynamic brain activity measured under naturalistic conditions can be decoded using current fMRI technology.
Heh, not only are scientists reconstructing still images, they're now reconstructing movies played in the brain. Yet according to your tiresomely parroted assertions, Andrew, this is supposed to be impossible, and an instance of "ignorance and denial".
Andrew?
I believe I have pointed out, Calilasseia, that, detecting images of pictures or movies, from a pre-selected set (and using “education”, that is, calibration of the subject’s brain activity to help), while impressive, is not in the same class as detecting uncalibrated thoughts that do not involve imagery. So we are no further along in our discussion than we were before. The tremendous unknowns remain.
I am not sure who Andrew is, but would like to meet him!
Calilasseia wrote:
<other research along similar lines omitted…>
I think I've done enough here to demonstrate that the papers in question are completely devoid of "tremendous unknowns, and things yet to be determined, and unanswered question after unanswered question".
They are devoid of actually addressing the issue of detecting and interpreting uncallibrated symbolic thought.
This is a strong accusation when you haven’t contradicted my point at all. Perhaps you will one day. The tremendous unknowns and unreliability remain. Let me know when the research pulls a list of fellow resistance members out of a victim’s brain.
No weeping. No game not involving pictures has even begun. I do concede, and have conceded, that detecting the images in the brain, even with all the limitations and caveats and calibration, and small subsets, is still quite impressive. But very, very limited. You can use it as proof positive that thoughts are purely material if you wish. Because of your worldview, your assumptions, your pre-disposition. In spite of the tremendous unknowns. Because you can’t afford to let a divine foot in the door.
Calilasseia wrote:Wilberforce1860 wrote:Calilasseia wrote:Which on its own destroys your apologetic attempt to label thoughts "immaterial" by appealing to their purportedly "hidden" status. But we've been through this about 40 pages ago, and you still didn't understand the elementary concepts then. Are you starting to make progress here?
There is no progress to be made. Keeping something material hidden by whatever means, is entirely different than keeping a thought hidden.
No it isn't. Read those papers and fucking weep.
I await my list of resistance members. Your own example, by the way.
Calilasseia wrote:Wilberforce1860 wrote:The possibility exists that the material item can be found, by a variety of means. There is no possibility that the thought can be, unless the person in question voluntarily reveals it (coercion in all its forms aside, and there are many examples illustrating that this doesn’t work either).
Wrong. Those papers blow your assertions out of the water with a nuclear depth charge. Read them and fucking weep.
Your own example of the resistance members who have died without revealing their confederates proves my point. Your research doesn’t touch it. And by the way, this research has only been done with willing subjects, correct? Try it with the unwilling, and let me know how it turns out.
But to clarify, because I am interested in all this – are the people looking at the pictures while the experimenter are taking their brain scans or whatever. And then the experimenters are plucking a picture out of a database.
Or, are the people recalling a picture from memory, last seen many days ago, and then the experimenters are still successfully plucking that out of a database? I’m just wondering. I find some of the language a little tough to follow.
If you wish to make an argument based on science, you must use science relevant to the point. What science is going to reveal the list of resistance members?
Calilasseia wrote:Wilberforce1860 wrote:Calilasseia wrote:Not in the second instance. It's entirely possible, in the second instance, for the fridge not to contain a pie. But of course, definitive determination of this is relatively easy. However, just because the determination of the status of an entity happens to be difficult, and requires diligent skilled labour, doesn't in any way make magic entities necessary for that entity. Is the penny dropping here?
Finding the pie is not the point. We know if can be found (or not), given a bomb expert of sufficient skill. The point is the inability of a bomb expert to find a thought that someone does not wish to reveal.
Those neuroscientists are pissing all over your apologetic assertions. Read those papers and fucking weep.
The brave members of the Maquis rise up to make my point!
Calilasseia wrote:Wilberforce1860 wrote:It is the thoughts that I maintain have an immaterial component, as you well know.
But wait, it doesn't. Those papers point pretty conclusively to thoughts being entirely material processes. Along with the vast array of antecedent data about brain activity from the same scientific disciplines. Oh wait, how come is it that when key parts of your brain chemistry stop, so do your thoughts?
Viva La France!
Calilasseia wrote:
Indeed, I've a nice little experience of my own to bring to the table here, courtesy of my tonsillectomy in the year 2000. This is an operation that requires the patient to be under general anaesthesia, in order for the surgeon to perform it. Now, here's the fun part. I went through the usual procedure, lying on the operating table, and watching all those frighteningly professional theatre technicians prepping me for the op. Then, they deployed the standard "can you please count down from 10 to 1 backwards for us?" question as the anaesthetic was being injected. I made it to 4 before all the lights went out, and I was completely devoid of thoughts for the two hours of the operation. It was literally like having a switch thrown. One moment, I was awake and active, the next, I might as well have been dead, for all the activity I noticed for the two hour duration. Only when I came round, did I notice that, oh look, I'd been wheeled out of theatre into the post-op recovery room, and I was getting to grips with what was, in effect, a complete consciousness reboot.
Hmm. “A complete consciousness reboot”? The only part of your conscious thought that you don’t appear to remember was what happened when you were under the knife. It’s not like you couldn’t remember what happened before that, is it? You recall it well. This did not wipe your memory. It simply quiesced you. And your automatic brain functions continued, as they must, for you to live.
Calilasseia wrote:
Now if a simple chemical molecule such as etomidate, the intravenous anaesthetic used to propel me into deep anaesthesia, could have the effect of performing a complete consciousness shutdown, followed by a reboot once the effects wore off, I think it's pretty safe to conclude that your assertions about the "immaterial" nature of thoughts are complete hooey.
I keep wondering if we are in the same conversation. What does this have to do with anything? Yes, we live in material world, and yes, we are affected by a wide variety of drugs and chemicals, including the food we eat for lunch. To say that we are material is not something I disagree with. To say we are ONLY material – THAT I disagree with!
Viva La France!
You have made the references less vague. I appreciate that. They don’t negate your own example of the brave Maquis. Sorry.
Viva the pie eating, pastry loving Maquis! You may take their pastry, but you’ll never get them to reveal their comrades!
Which don’t negate your own example of the Maquis?
Calilasseia wrote:
So much for "assertions and vague references". But then I knew you would be too indolent to do your homework on this, and actually chase up the papers. Congratulations for walking into the prepared trap as expected.
It’s the same trap from many pages ago, with the same result. It’s a good try, but you haven’t negated your own example.
Calilasseia wrote:Wilberforce1860 wrote:Calilasseia wrote:Oh wait, what was it I said above about entities requiring diligent skilled labour to ascertain? QED.
You are holding the thread of the argument only if you have provided sufficient reason to believe that thoughts are material and can be detected by material means, without the consent of the thinker. I do not believe you have established this.
Those papers I've just presented just did that.
First of all, the research was all done with the consent of the subjects, their voluntary cooperation, and pre-calibration of their mental activity, not to mention a limited set of pictures. Not of much use to the Nazis trying to identify previously unidentified resistance members, is it?
Calilasseia wrote:Wilberforce1860 wrote:Calilasseia wrote:And once again, we're back to your woeful inability to comprehend processes.
Thoughts are processes. Which is why you won't detect them when those processes stop.
As I have observed in conversations with ADParker, I believe, thoughts are both processes, and the results of processes.
Ah, naive conflation of separate entities. A familiar part of the supernaturalist aetiology.
The end results of those processes are ideas. I think I told you this earlier in the thread.
Who is being naïve? Each thought in a train of thought is itself an idea. Separately storable and retrievable from memory. But you are not addressing the issue. You claim that thought is a “process”, and you appear to claim that thoughts stop and have no representation in your mind (or memory) when you stop thinking. I lose track of why you think that is important. But if that is what you claim, you need to explain your view on memory, and thoughts being stored in memory. Why does a person need to be thinking in order for you to examine their memory, for example? They can pull things from it easily enough, no? Can’t you, if these thoughts are strictly material?
Wilberforce1860 wrote: A train of thought goes in a direction and produces an end result. All of these are stored in memory and can be recalled, subject to the will of the thinker.
Calilasseia wrote:
And how is that processing and storage achieved? Oh wait, they're both products of brain chemistry. And come crashing to a halt when the requsite brain chemistry stops.
If you are making a point, I’ve lost it. What is your point?
Perhaps they can explain this better than you have, at least, from where I sit.
Another point, which, if there is one, I don’t follow.
Calilasseia wrote:Wilberforce1860 wrote:ADParker has asserted that memory is really only “thinking the thoughts over again”. But what is your explanation for thinking the same thoughts? Or recalling something we have observed, such as the color of a dress we aren’t currently looking at, or where we left our keys?
Data storage underpinned by brain chemistry. Next?
Well, at least we both agree you can store thoughts and recover them. What is your evidence that you can pull stored memories from people’s brains? All of your research quoted above requires actual brain activity in order to be able detect the picture being viewed, does it not? You are detecting the result of sensory input on the other side of the eye, so to speak, as it is processed by the brain. Now, can you identify this picture 10 hours later, while the viewer is unconscious? The Nazis want to know!
Calilasseia wrote:Wilberforce1860 wrote:This idea goes against common usage of language (e.g., recalling something from memory, memorizing a poem, etc.) as well as common every day experiences. Are you denying that thoughts are stored?
And here we have yet more naive conflation, between the process, the algorithm directing the process, and the generated data.
The latter two are stored. The first, by definition, not.
Tell me, do you know how neurons work? Because your above assertions suggest very strongly that you don't.
I know they exist. I do not know how they work. I maintain our understanding is extremely limited and just getting started. But tell me, and the Nazis, how to get the list of resistance members, and my knowledge of neurons and how they work will increase considerably.
Calilasseia wrote:Wilberforce1860 wrote:Are you claiming that you can only detect them if the person somehow recalls and thinks about them? I really am not sure what point you are trying to make.
The point I was making, was the elementary point that when certain parts of your brain chemistry come to a halt, so do your thoughts. That you were unable to discern this from my words in my post, is itself worryingly revealing.
Just tell me how to get the list of resistance members…
Calilasseia wrote:Wilberforce1860 wrote:Calilasseia wrote:Oh but wait, as long as the underlying brain chemistry is still active, we can detect the presence of thoughts. Indeed, scientists were able to do this before detailed resolution fMRI scanning became available. Ever heard of an electroencephalograph? It detects the tiny fluctuations of electrical current in your brain that arise from that chemistry. What's more, it tells us, every time, that when said chemistry stops, so does the detection of thoughts. All the electrical activity plummets to zero.
However, now that we have detailed resolution fMRI scanning technology, we can home in on individual thoughts, and work out what the person is thinking. I've presented papers here documenting the requisite experiments. The technology may be in it infancy, but it works. Which would be impossible if your assertions about thoughts being "immaterial" were true. Game over.
Where is my list?
Calilasseia wrote:Wilberforce1860 wrote: Are you claiming that you can determine the actual content of a person’s thoughts?
Like I said, the technology may be in its infancy, but it works, with respect to working out what images an individual has seen. It's only a matter of time before that technology will allow you to turn your dreams into watchable DVDs. Now there's a scary thought for you. Other researchers are working on determining language content of thoughts, such as the language content of our subvocalisations when we read text. Here's something for you to do some homework on while you're contemplating all of this: those self same neuroscientists have determined that the manner in which human brains process language, undergoes a rearrangement when those brains are affected by repeat episodes of temporal lobe epilepsy. There's a nice paper on this very subject in the European Journal of Neurology - let's see if you can find it shall we, without me doing your homework for you?
In fact, just to illustrate how much your apologetics constitute abject fail here, whilst checking the literature for some interesting papers to bring here next time you post fatuous and manifestly false assertions, I've just found something like two dozen other papers on apposite topics that I wasn't looking for originally. But which are going to come in very useful in this thread in the future. I love it when this happens. This should be telling you something very interesting, not least about the consilience of the scientific enterprise.
“The technology may be in its infancy,”
Really? I thought I was supposed to be weeping…
I’ll take that as a no. Sigh. No list of resistance members today…
Calilasseia wrote:Wilberforce1860 wrote:Is this any different than the research mentioned previously, where researchers were able to determine the movie someone was thinking about (out of a list of pre-selected movies)? If so, could you provide a reference?
And yet despite me having dropped this in your lap previously, you still think it's impossible to read someone else's thoughts. How cute.
The answer to my question is an obvious no. I’m not sure why you felt it important to re-iterate the same, very limited, research, that you already presented. (Don’t forget that I was impressed. But not enough. No list of resistance members…)
Calilasseia wrote:Wilberforce1860 wrote:You seem to be claiming that yes, we can determine the thoughts of others (assuming they consent to being wired up, or scanned, so that you can try), with complete accuracy? That seems to be what you are implying when you say “work out what the person is thinking”. Detecting the presence of thoughts is far different than detecting their content.
Oh wait, what part of "determining the images observed requires detecting the content by definition" do you not understand?
Where is my list?
End of part 1 of 2. To Be continued.