This statement “But you can’t do that with an image of the word” seems to be saying it’s just not possible.
Before I wrote that post I tested with the following words:
fish
water
probably
The longer the word, the more time and energy it took to create the mental image but once created the random ordering of the letters was accurate and fast.
I understand that the “image” is not the same as drawing on a white board and is most likely tied to underlying computation and may just be a representation of whats going on, but that quote states that my mental image of “probably” doesn’t allow me to randomly order the letters, which is wrong.
I just tried it with “psychotherapy”, it definitely takes effort to load it up, once there I easily pulled letters at random, no dups and hitting each letter.
It does seem that the imagery is a key part of the process, but I understand what you are saying that there is probably really some underlying computation going on.
They learn written English, just as you or I would learn sign language.
The sentence structure is different in the two languages, and email can become confusing.
Try it with “floccinaucinihilipilification” (yes, it’s a word!) and let us know how it goes.
What I was objecting to was your comment that the paper you quoted “makes a completely incorrect assumption … stunned someone actually wrote it down” as if it was some random hunch rather than a widely accepted principle in cognitive science – all based on your self-assessment of what you think is going on. Much as someone’s claim that when making decisions he carries on an internal monologue in words, or in images, or in abstractions that are neither, these subjective perceptions vary from case to case and from person to person and offer relatively little objective insight into what’s really going on.
The comment that “you can’t do that with a mental image” is just saying that the process that operates on a mental image in this class of task is qualitatively different than one which operates on a perceived image of a physical object, and obviously not trying to say that you can’t do it mentally at all since of course you can. But with a degree of difficulty that increases commensurately with the complexity of the word, while it remains trivial with a physically perceived image.
The quote may be a slight overstatement, and may be less true of some people than of others, but it is most certainly the case, and has been demonstrated in many experiments, that people (all people) find it more difficult, and often quite impossible, to place an interpretation upon their mental images that is different from the original or “standard” one that they first gave it (although some images, may be easier to reinterpret than others, and some conditions may be more or less conducive to enabling reinterpretation). The experimental and theoretical literature on the subject is large and fractious, but I think it fully supports my statement.
I will take your word for it (if that is what you are saying) that you do not find it very difficult to visualize a word and read off the letters backwards (personally, I do find it difficult, though not necessarily impossible, for words of more than about three letters, and I strongly suspect that most people do), but I would be very surprised if, were this ability to be tested under experimental conditions using moderately uncommon and moderately lengthy words, you would prove to be a lot slower at spelling them out backwards, by reading them off your image (assuming that you could manage the feat at all), than you would at spelling them out forwards, also by reading off. On the other hand, there would be little difference in how quickly you could read off the letters backward or forward if you were seeing the word written in front of you.
I am not sure if this precise experiment has been done, but enough similar ones have been done to make me confident of the result.
Nobody is denying that, if you were asked to spell out a word backward, you would almost certainly try to do it by visualizing the word and reading off the letters “backwards”, right to left. It may well indeed be the “only way” that people are able to do the task at all. Pylyshyn (the author of your quote, and the paper at your link) is not denying that at all. If that is what you think he is “completely wrong” about, you are misunderstanding the quote. What he is denying (and all the experimental evidence, as well as my own introspections, strongly support him) is that you would find it anything like as easy to read the letters backwards off your mental image, as either to read them backwards from a blackboard, or forwards from your mental image. If you are an exception to that, you are very unusual.
What do you mean by “pulling letters at random”? How are you selecting which to “pull”? If this is all in your own head, you are almost certainly just confirming to yourself that letters you already know are there really are visualized: “Have I visualized the E? Yes, there it is.” That is easy, and not to the point. A real test would be either what I suggested in my previous post (spelling it out backwards via (subjectively) reading the letters off your image), or having someone else calling out random positions in the word such as “6th letter” and having you answer with the 6th letter. I am confident that, if you could manage this at all, you would find it much more difficult than finding the 6th letter in “psychotherapy” when you can see it written down.
There is an analogy with touch typing. When I try to type the word “casino”, it always comes out “casion” and I have to go back and correct it. I am a touch typist, and my muscle memory types “ion” so often, that it takes a special conscious effort to type a word that has the same letter in a new and novel order. Similarly, “network” comes out “newtork” because I type “new” so often. I no longer type words by spelling them in my mind, unless I have used the word so infrequently that I have to think a letter at a time to type them. When I think the word “new” my fingers go to those keys in that order, the same way I put up my hand when I think you have tossed me a ball.
If you are a hunt and peck two-finger typist, you might have difficulty visualizing how I type, just as a hearing person has difficulty visualing now a deaf person has learned hot to efficiently form linear ideas. Remember, deaf people can read and write, so the idea of subject-verb-object is not completely alien to them.
Yeah, I doubt there is much, if any, validity to that. It’s just an idiomatic saying. I don’t think I ever say “I hear what you’re saying” or any variation thereof, but I say “I see,” and “I see what you mean,” etc., even though I almost always think through things with language/internal monologue. (Plus there’s the criticism mentioned above about imagery types.) It’s just the phrase that I’m most used to hearing, so that’s the one I’m most used to using. It’s not meant to be literal.
I’m not saying it’s comparable to reading from a blackboard, but I am saying, for manageable sized words, it’s substantially similar.
The word “probably” is manageable, the word “psychotherapy” is still barely in the manageable group but much harder than “probably”.
Yes, maybe I am misunderstanding. The wording seemed to indicate it just doesn’t happen to any degree, if the point is that the brain is more limited in that capacity than viewing a scene then of course that is true.
What I mean is to randomly choose letters from the image of the word in the same way you would randomly pull letters from the word on a blackboard.
In both cases your mind is (kind of) randomly choosing the next position to pull a letter from and then doing it and keeping track that the letter in that position has already been taken and continuing on until complete.
For me, for manageable sized words, the process appears to be basically the same.
If you prefer reversing the letters that’s fine, that’s even easier, and again, once the image of the word has been loaded up, the retrieval process is quick and easy (for manageable sized words).
It’s possible the imagery is just along for the ride during the computation, I can’t say for sure of course, but it does seem like the imagery is actually adding value and making the process easier.
My mother was born deaf , and left my father and me when I was 5. She came back for a visit when I was 12, and again when I was 17. I don’t know about the thought processes but to be able to “talk to each other” we had to write everything. I remember her wanting to write a word that she wasn’t sure of the spelling, she sat pen in one hand the other hand was spelling out the word in sign language letters. Was that thinking, or was she doing what I sometimes do when I write a word out to see if it looks right?
It might seem similar subjectively, but, unless you are truly unusual, I am confident that if you were actually tested and timed (on spelling backwards versus forwards), there would be a big difference. Your observation about the difference between longer and shorter words is true, but more significant than you realize. Longer words will rapidly get harder (and so slower) as length increases (e.g., a ten letter word will take much more than twice as long as a 5 letter one), and soon will get to be impossible. By contrast, the length of the words on the blackboard will not affect how rapidly they can be spelled out backwards.
The point (Pylyshyn’s point, anyway) is that, despite common assumptions to the contrary, experiencing a mental image of something is not the same as, or even usefully analogous to, seeing a picture of that thing. (In this case, of course, there is no real distinction to be made between a written word and a picture of that written word, but there is still an important, indeed radical, difference between a written word and the mental image of a written word.
Yes, if you get to choose which letters you “pull”, and in what order, this is not a meaningful test of anything, and not of any theoretical relevance. Furthermore, if there is one thing that experimental psychology has learned, and substantiated repeatedly, over the past century or so, it is that how a mental process “seems” to the person undergoing it is an extremely unreliable indicator of what is actually going on.
This, unlike the random “pulling” of letters is a meaningful test, but I am confident that, if you were actually timed, under experimental conditions, you would find that even for relatively short words (longer than about 3 or 4 letters, unless you are a real outlier), it is a lot more difficult to spell them backwards “in your head” from an image, than it is to spell them forwards. (This is assuming that you have not previously practiced spelling the words backwards, of course. A better test would be to use random but pronounceable letter strings with which you were not previously familiar.)
One need not, however, leap from the claim that experiencing an image during a task like this is importantly different from actually looking at a picture (or actual thing) to the conclusion that “imagery is just along for the ride during the computation” and does not “add value”. “Mental images are not like pictures” in no way implies “mental images are of no cognitive significance or value”.
The claim isn’t that backwards is as fast as forward. Reading forward is a memorized action, reading backwards is not.
The claim in the PDF was that mental imagery doesn’t help to “read in any order”, but in my experience, for small enough words, it seems to.
Agreed, brain image is not equivalent to blackboard.
But I think that it’s too strong a point based on the evidence he gives “can’t read in any order” because the statement appears to be too broad/strongly worded.
Clearly they are not the same, but he seems to be dismissing the possibility there is shared/similar processing.
Of course, no argument there. I think that’s part of what makes it a fun problem, trying to figure out which kinds of tests might actually reveal something about the inner workings.
Again, the claim wasn’t that backwards is as fast as forward, the claim was that mental imagery doesn’t allow a person to achieve backwards (or random) order anywhere close to blackboard speeds.
So I just ran a test to make sure I’m not just fooling myself (not experimental conditions, but I think it demonstrates general capability), I called a coworker and asked him to randomly choose 4 and 5 letter words, don’t tell me in advance, speak a word to me over the phone and I will read the letters backwards back to him.
We did a handful and his comment was “wow, you are fast at that, I can’t do that”.
The only practicing I’ve done is the few words I listed in this thread.
Haven’t you just disproved your own claim? You were given information verbally, and transformed it into a mental symbolic representation for processing that had nothing to do with anything you had actually seen, but that you conceptualized as an image. How do you think a blind person, knowing only Braille, would conceptualize the same problem? Do you think he’s be incapable of demonstrating the same level of computational symbol-processing skill?
No, I don’t think I disproved my own claim. The claim is that mental imagery (seems) to assist with this process.
1 - Yes I was given info verbally
2 - Yes I constructed a mental representation (possibly symbolic)
3 - Yes I had never seen it before (that’s what a mental image is, something generated in the brain - this supports not refutes the claim)
4 - Yes I conceptualized as an image (at least this seems valid)
None of that refutes the claim, unless you think that the last part of conceptualizing as an image adds no value and is along for the ride.
If that’s what you think (I sure don’t know the answer) then it would seem a test like this in which the person uses mental imagery and then tries the same task without may tell us something.
No, the claim that I am making (and that Pylyshyn is making in your PDF) is that whereas, when you are reading off a blackboard, the order in which you attempt to read off the letters makes no difference (or very little), it makes a great deal of difference when you are trying to read them off a mental image. In the image, it is much easier to go forward than backward (and this effect is stronger the longer the word is); on the blackboard, it makes virtually no difference, either which direction you go or how long the word is.
Here is your quote, with some emphasis and an annotation added to make the intended meaning clear:
Despite what you claim, it does not say anything about imagery “not helping”. The imagery may well help, it may even be vital to any ability we have to do backwards spelling “in our heads”, but it does not help nearly as much (or in quite the same way) as having the word written out before us on the blackboard helps. What the quote does say is that having imagery of a visual scene is very different from actually looking at such a scene (and behind that is Pylyshyn’s real underlying point, that it is also different from looking at a picture even an “inner picture” if such a thing could exist - of such a scene), and one of the ways that it is different, one of the evidences of that difference (there are others), is that there are constraints on how we extract information from images that do not constrain us when we extract equivalent information from equivalent visual scenes (or pictures). In particular, there is less flexibility in how we can extract information from images. It is not as easy to do it in any arbitrary order. This is true, and well supported by numerous experiments.
You are quoting, incidentally, from what is a brief entry in a (sort of) encyclopedia. Like all such works, it is intended as a brief overview of a subject, not a definitive account with all the nuances accounted for and spelled out, and every possible ambiguity eliminated. For that you need to read the original literature which it is summarizing. That is something I have done - know this literature very well - so I am confident in my interpretation.
No, the claim he is making is that order makes a difference when reading information off from mental images, whereas it does not make a difference when reading it from a blackboard.
Well that certainly suggests that you are better at the backward spelling task than most people. It does not show, however, that you are equally good at backward spelling (based on a mental image) as you are at forward spelling, based on the same image, and that is the point at issue. Furthermore, you have already admitted that you find the backward spelling task harder the longer the word is, and that, again, is not something that happens when reading off letters backwards from a blackboard.
It certainly seems to, and, very likely, really does. Indeed, I myself think it is very likely that the task of backward spelling of words of any length, without being able to see the written word in front of one, is impossible without visual imagery (or other spatial imagery: tactile or kinesthetic imagery might, conceivably work). That is not a claim that Pylyshyn is denying, however, in the passage that you quote. It is not what I (nor, I think, wolfpup) have been arguing with you about.
Pylyshyn (not me) may hold that the conscious experience of imagery during the backward spelling task (and other cognitive tasks normally evoke imagery) is merely a normal accompaniment of the underlying cognitive processes involved in the task, but does not actually make a contribution to doing it (much as driving your car is inevitably accompanied by the engine making a noise, but the noise does not actually help to make the car go). Certainly Pylyshyn is often, and quite plausibly, accused of holding such a view, although I do not think he has actually ever clearly committed himself to it. In any case, however, it is not the view that you are, wrongly, attributing to him.
You may be correct, but the words he wrote clearly state something different:
“If you were to write a word on the board you could easily read the letters in any order. But you can’t do that with an image of the word.”
If he really meant “only those words with greater than 10 letters” then I think he was careless with his wording.
For most people, the effect would be very apparent long before you got to words as long as ten letters. You may be an exception in this respect. However, I would lay odds that even you would be measurably slower in spelling even a five or six (maybe even four) letter word backwards from an image than you would be in reading it backwards from a blackboard, even if the difference were not obvious to casual observation in your case (whereas it probably would be obvious in most people’s cases).
I do not think he was careless in his wording to any culpable degree. You misread him, and imputed claims to him (certainly some of the claims you impute to him in your earlier posts) that are simply not there. (I wonder if you have been reading some of the works of certain of his theoretical opponents, who are given to insinuating, quite falsely and very misleadingly, that Pylyshyn believes that mental imagery does not really exists, or, if it does, that it does not serve any cognitive function. Pylyshyn has, in fact, never said or implied any such things, and more recently has made it quite explicit that he does not believe them.)
He could, I suppose, have rendered himself a little less open to misinterpretation if he had repeated the word “easily” in the second sentence. He probably thought, however, that it would have seemed clumsy and redundant, and that a sympathetic reader ought not to have too much trouble in divining his meaning without it.
[Perhaps I ought to add, as I seem to be defending him so hard, that I am not Pylyshyn, and, in fact, I strongly disagree with some of his most cherished and central views, both about mental imagery in particular and cognition in general. I have, however, been assured by him personally that my understanding of his view about mental imagery is substantially correct, and I do not believe that he makes the relatively elementary mistakes that you (in your earlier posts) and some of his more prominent theoretical opponents, attribute to him. If anything, the problem is that he is much smarter than many of his opponents, and many of his arguments go way over their heads (at least, that is the most charitable interpretation I can put on some of what they have said about him), despite the fact that, unlike you (presumably), they have been in close dialogue with him for many years, and really ought to be able to understand where he is coming from by now.]
