Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Multiple Intelligences

  1. Aug 22, 2003 #1


    User Avatar
    Staff Emeritus
    Gold Member

    Howard Gardner, a Harvard Professor, has written quite a bit on this psychological theory which claims that the human intelligence is more then just math or book smarts...

    Multiple Intelligence list

    I have to completely agree with his theory because IQ tests are very selective in the intelligences it measures. Read this quick page, and I would like to hear any comments...
  2. jcsd
  3. Aug 22, 2003 #2
    This list provides absolutely nothing new at all. Except one catagory which is absurd.

    "Musical". This person has seriously overstepped the boundaries of what intelligence is. Intelligence is a physical process clearly understood by specialists in the field. This person is not in the field. It sounds like he might be that type of person that use the term "emotional intelligence" which is scientifically absurd. There is only one mental process which is designated as being intelligence.

    This person is WAY out of their league and way out of their education. This is the type of broken boundaries which damage the research of us who take this pursuit seriously. Horrible horrible work on this "professors" part.
  4. Aug 22, 2003 #3


    User Avatar
    Staff Emeritus
    Gold Member

    how would you define intelligence? i believe musical talent is a form of intelligence...
  5. Aug 22, 2003 #4

    Probably over 99% of the population uses the word intelligence as the wrong type of word.

    For instance you cannot say "Hi Kerrie, I've got some jumping to share with you." It's the wrong type of word.

    Intelligence is a processing ability that is defined in only mathematical and physic(al) equations.

    You can't say intelligence is the ability to.........

    Just like you can't say Newton's second law is the ability to......

    Do you see what I mean? Humans think intelligence is a human ability to "do something". Intelligence is a process like respiration, digestion, photosynthesis.

    With that said it becomes obvious music talent can't be a form of intelligence just like a piece of apple pie can't be a forum of newton's law. It isn't a sensical sentence. I know it's difficult for some people to relearn the fact that intelligence is a physical process and not an ability...

    This can be made more evident in attempts to createo artificial intelligence.The goal is merely to use computer programms to mimic the process - just like a computer program that can mimic an ecological system to determin what excess fishing will do to the reef etc...

    It's a hard thing for people to understand and to accept. I find this in classes I teach - but in a student environment they are much more likely to feel good about knowing what intelligence really is.

    For instance Kerrie, let's say I smoke and you don't. Your respiratory capacity would probably be much stronger than mine.

    In just the same way, someone who was fetal alcohol syndrome might have a less strong intelligence.

    I can say more on this if anyone wants - don't want to bore you all :)
  6. Aug 22, 2003 #5


    User Avatar
    Science Advisor

    What exactly do you think music is? Do you play any instruments?

    I mean, if intelligence is as you define it the ability to process information, well, music is nothing but math.

    That being said, some people have a particular knack for music. I know a guy that is dumb as a brick, would probably struggle with a simple problem like 15x10. Infact, the only thing I've ever known him to be incredibly good at was playing guitar. Let him listen to a song he's never heard before, maybe twice, and the third time around he'll play right along with the music.

    Now, seeing as how music is nothing more then a different form of mathematics, and there are more ppl in the world then my friend with that described ability, what exactly would you call it?

    Not arguing your point about peoples misuse of words though. Musical ability is a form of intelligence though. You do process info, either by hearing it, or reading it from a sheet. So, you have input of info, and the music is the results of the processing.
  7. Aug 22, 2003 #6
    MegaShawn - I have played music for 23 years, and am currently the CEO of a record label in California. I've taught music for 6 years, and written 4 scores for orchestra.

    And here you conclude from 4 false premises that music is a form of intelligence.

    I thought I stated above the scientific definition of intelligence.

    Apparently you missed one (well many) important points.

    The definition I gave is the scientific definition. I was not displaying a personal opinion.

    This is what scientists have designated as being intelligence - it's not up for debate unless you have an enormously overwhelming amount of evidence.

    I'll be focusing on others responses......
  8. Aug 22, 2003 #7


    User Avatar
    Staff Emeritus
    Gold Member


    You offer an interesting insight into the definition of intelligence...I began this thread because I believe IQ tests only measure a minor amount of one's abilities...as stated in the link:

    I get annoyed when the IQ of a person is glorified but other abilities are not, such as musical ability...

    consider this blurb regarding autistic savants from the same website:
    how do we explain situations like these with an IQ test?
  9. Aug 22, 2003 #8
    Kerrie - gonna grab a bite to eat, anxious to answer your question with some science info brb!!!
  10. Aug 22, 2003 #9
    Thanks! It comes from lots o' research. Intelligence is one of those topics, the information of which seems to not leak out into the public.

    Yeah - Having a high IQ can be a good quality for sure, but it doesn't really seem to lead to any special circumstances in life. Poor people have high IQs, so do rich people. Although it does correlate to some things.

    Some of them IQ does test, like calculating. Some of them it does not test.

    The big problem we have here is Gardners misuse of the word intelligence. It means what I stated earlier, in the field of science which deal with the part of the animal body that contains this process.

    And so in using this word he's attempting to say that these abilities are some how more important on the hierarchy of all a human can do.

    They are not. They are merely things a person can do, just like singing, dancing, writing a story etc....

    I'm speaking particularly of his mention of musical ability.

    To be frank, I've played music at all ends, and taught and wrote, for 23 years.

    Claiming someone can replicate piano pitches of a song from hearing it once has 100% nothing to do with musical ability.

    Music is taught - it is not innate. That's like saying reading is innate. An ability which may assist this process may be innate, but musical knowledge is learned.

    So we have alot of issues here - colleagues in my fields have big issues with this kind of imporper terminology and generalizations.
  11. Aug 22, 2003 #10
    You got me on a roll here, this is a topic I often speak on so it's on the tip of my tongue.

    Once we replace this misuse of the word intelligence in Gardners work with another word, we notice that, for example if we use the word "ability", that all he is doing is talking about a few of the literally billions of abilities humans have.

    (mis)Using the intelligence makes it seem like he is talking about some kind of hierarchical dictator of humans ultimate abilities - and that someone who can caculuate quickly, play perfect pitch songs with one listen, and has excellent spatial skills is anymore "intelligent" or anymoreo "anything" for that matter, than a person who can balance a spoon on there nose, belch out jingle bells, and or dribble four basketballs at once.

    See what I mean?

    I understand that his attempt appears, based on the webpage, to be to single out the "true" abilities that designate what he calls intelligence, or desginate the tip of the hierarchy from which all abilities progress downward from.

    It's just that he is not doing that, but it would appear he is since he uses this word intelligence.

    (You might be able to see that we in these fields take a bit of an issue with this!).

    Keep this going, ask me more :)
    Last edited by a moderator: Aug 22, 2003
  12. Aug 22, 2003 #11
    I have to agree with Biology on this. Music can be "taught" people can improve thier music ability by practicing. It's true some people learn music more rapidly than others. But this could be due to other improved abilities such as accute pitch, etc that help them master this quicker. Intelligence isn't something that can be improved through any amount of practice. It's static(allbeit on a curve through early adulthood). Mentat and I had an extensive discussion about this before. Intelligence and musical ability through practice are two different things. I play guitar and piano. When I started I sucked.I was no virtuoso. But I've grudually improved over the yeaars. I can't say that I'm more intelligent that I was when I began. Perhaps more knowledgable, but as it has been stated and confirmed before, intelligence and knowledge are two separate things.

    As For Autism, Asperger's Syndrome and other disorders associated with "rainman syndrome". First you should know that only 10 percent of autistic patients exhibit savant syndrome. It's been theorized that these types people are compensating for damage- in much the same way a blind man tends to have an increase sense of smell and hearing ability.

    You can read the entire article
  13. Aug 22, 2003 #12

    Les Sleeth

    User Avatar
    Gold Member

    The problem for me is, I don't think you took enough time to understood Kerrie's position because your refutation is so uncompromisingly and instantly negating.

    I think she sees something even if she hasn't expressed it precisely . . . do you know what it is?
  14. Aug 22, 2003 #13
    I didn't see anything where Kerrie was displaying much of an opinion. Everything she said was quoted from other sources....and she asked me "what I thought". I wasn't agreeing or disagreeing with her - I didn't think she made a claim of her own....
  15. Aug 22, 2003 #14
    I don't know what Kerrie sees, but you seem to know. Very interested to hear your take on this.
  16. Aug 22, 2003 #15

    Les Sleeth

    User Avatar
    Gold Member

    Well, in a way this is related to the my recent thread on "rightness."

    Who gets to define what is valuable? Let's say you have the responsiblity for developing the IQ test all the world will use, and you are firmly convinced that intelligence is best when it produces products/services useful to society.

    Guess what sorts of problems are going to show up on the test?

    Dominance prevails and IQ values reflect the values of those dominating. I realize the argument can be made that if something is dominating, it must be superior. But in actuality that is not always or even mostly true.

    If one dumps cayenne in the soup, it will dominate, but does it make the best soup? Quality is very often (maybe MOST often) found in subtlty. As a music listener, my tastes have steadily moved toward subtlty; and the same is true for me with food, wine, sex, art . . .

    So the big guys who are in power can define "best" how they please. Might makes right, doesn't it? And who, speaking from the more subtle place, dares question that?
  17. Aug 22, 2003 #16
    Conspiracy theory? Not biting
  18. Aug 23, 2003 #17


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member


    I don't know if what you're saying is entirely consistent, so I would appreciate it if you clarified your stance a little. You say that the 'ability to' do something does not constitute intelligence-- but you also seem to think that mathematical aptitude is a genuine form of intelligence. So how precisely can the ability to do quick calculations in your head be a form of intelligence, while the ability to pick up a song after a couple of lessons not be?

    You also say that intelligence is best seen as a process. Exactly what kind of process do you mean? It must be something more specific than 'neural information processing,' because then any activity in the brain can be seen as a form of intelligence. So what exactly is it about the process of intelligence, as you define it, that delineates it from other neural processes in the brain?

    You also seem to imply the criterion that that which can be taught does not constitute a form of intelligence. But if mathematical aptitude constitutes a form of intelligence, then clearly this intelligence depends on some kind of formal teaching. For instance, if I had never taken math beyond the grade school level, I would have much greater difficulty performing calculations in my head. Now there is a distinction to be made; is this the case because of the math I was taught, or was the continual experience of doing calculations honing some innate mathematical ability of mine? I think you would say the latter, and I would (mostly) agree. But can't we then also look at the process of formally learning music as, on some level, honing some innate musical ability? Would it then be incorrect to denote this innate musical ability as musical intelligence?

    As an aside, I think you overstate your case when you talk about the firm grasp the scientific community has on the concept of intelligence. If we understood intelligence that well, the field of artificial intelligence might have lived up to some of its bold predictions by now. Even some of the more successful applications, such as Deep Blue, rely more on a priori knowledge and brute force than actual intelligence. Deep Blue has beaten the best chess player in the world because it was programmed to give each chess piece a particular value as determined by human chess masters and because the breadth and depth of its analysis of future moves was much, much greater than is humanly possible. The ratio of the quality of strategy to computational resources is still much higher in the human than in the computer; I would argue that this efficiency of information processing is part of what constitutes true intelligence. Classical AI has run into countless theoretical objections, and more importantly, its progress as a field has continually run into one brick wall after another. Simulated neural networks are a better approach, but even the results here have been relatively modest thus far.
  19. Aug 23, 2003 #18


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    How did this post constitute a conspiracy theory? I think it is a valid point. The scientific community may have formulated specific criteria for what constitutes intelligence, but that does not make it an open and shut case. It may be that, for the scientific purposes, the definition is overly narrow and does not capture what we usually mean when we say 'intelligence' (still waiting for BiologyForums to get back to me on this one)-- in which case I would argue that it is the scientific community that is misusing the word by essentially redefining it. Now this would not be a malicious action on their part, so it wouldn't constitute a 'conspiracy'-- it's not like the scientists are all getting together plotting ways to keep the lowly people down-- but the simple fact is that the scientific stance on the matter will have a sociological impact, especially if it is used to guage the value of human beings.
  20. Aug 23, 2003 #19
    Not to sidetrack, but you bring up some very good points. How close are we to formulating algorythms that closely parallel human thought? Obviously this would go beyond simply learning from your mistakes. or am I looking too deep and it is as simple as that? Was deep blue indeed capable of learning from trial and error? I would certainly think so. Otherwise we did nothing more than immensely scale up the EMACS. I guess a more humanistic approach would be "reading your opponent"? Sensing his posture, attitude, body signs, etc, and figuring that into the equation? Then you would have to figure in emotions such as anxiety, arrogance, fear,or any other psychological aspects that might figure into it. unfortunately those things can't at this point be fully integrated into a machine, so replication of realistic AI is still decades away. At least it seems to me. Maybe someone could talk about this more ?
  21. Aug 23, 2003 #20
    I'm smarter than all of you.
  22. Aug 23, 2003 #21
    The problem is that you're categorizing ALL intelligence as subject to the whims of our fancy. Does that mean we should throw up the entire system? You say that the factors for intelligence may have been determined by the needs of society, but if that is the case, it has failed miserably. Mensa is a testament to that. Look at all the high IQ people who are unsucessful. Success isn't something that can be measured as a function of intelligence. Certainly it would have a bearing, but there's not accounting for sheer willpower and determination. That's the test we need to develop in order to single out sucessful people. Not spatial and mathmatical reasoning. Sucessful self-made persons generally all have one single characteristic, and it's not logical facilities, or creativity, or social skills as much as it is Tenacity. It's the single characteristic that can compensate for the lack of other abilities sucessfully. You can't mkake yourself smarter, but you can make yourself more knowledgeable, more experienced, and more determined than the next person. Quite simply, raw intelligence rating still doesn't account for good old free human will. I know some pretty intelligent people, who are also some of the laziest people you could ever hope to meet. If there's an easier way to do it, they'll take it, if only to make thier lives easier.

    If the scientists were hoping to use the modern Intelligence Quotient as a measure of sucess, they sure fell way way short of the bar.
  23. Aug 23, 2003 #22


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Actually, I was saying something more like the inverse of that: the deigned concepts of what constitutes intelligence partially influence the values, motives, and actions of society at large as well as specific societal structures, such as the education system. Thus a person may be judged and valued by himself, by society at large, and by specific societal structures on the basis of these values, motives, and actions.
  24. Aug 23, 2003 #23


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    If I'm not mistaken, Deep Blue did include a learning algorithm, but it was only used for fidgeting with the values assigned to the various chess pieces. Other, more subtle changes in its strategy were hard-wired in by its programmers. For instance, there is something called the horizon effect-- initially, Deep Blue would only look forward on each branch of potential future moves by a predetermined maximum amount. This lead to an inevitable 'horizon' that it could not see past, and the results could be disasterous-- for instance, if the maximum search depth was 20 moves ahead, Deep Blue may have determined that its best chain of moves in a particular game results in taking the opponent's queen 20 moves from now, but it could not see that the 21st move would result in a checkmate and a loss. Thus the designers programmed in a caveat to maximum depth rule, allowing deeper searches for certain sequences of moves that were 'most promising.' Deep Blue would not have been able to learn this strategy unless the programmers explicitly programmed in some kind of learning algorithm functioning on the breadth and depth of its board searches. This is another weakness of classical AI-- the program cannot learn what it is not explicitly instructed to learn about. This obviously is not a very good grasp on the actual human learning process.

    As far as reading the opponent, I believe there has been some progress on recognizing facial characteristics using simulated neural networks-- for example, face recognition. I don't think that reading emotions through subtle changes in facial expression or body posture is something that is beyond the potential of neural networks. But more salient with regards to playing a good game of chess is how the game itself is actually approached. Deep Blue uses a brute force method-- it checks millions of sequences of moves and determines which is the best next move to make. A human obviously cannot check millions of different permutations in a reasonable amount of time-- the approach is fundamentally different, in that a good human chess player uses pattern recognition as his principle computational tool. Good pattern recognition is essential to what we consider to be human intelligence. In this respect, neural networks appear to be much more promising than classical AI approaches in constituting what we think of as intelligence, in that they computationally mimic the functional characteristics of the human brain and thus theoretically have the power to perform complex pattern recognition.
  25. Aug 23, 2003 #24
    Ok, let's of stuff to respond to here:

    I am unsure why I stated something seeming to say mathematical aptitude is a genuine form of intelligence. Show me, perhaps it doesn't communicate what I meant or I made a mistake.

    Intelligence is a process of a biological system. For instance the Respiratory System, is fo course a biological system. One process this system performs is called breathing: inhaling oxygen and exhaling CO2.

    The endorcrine system performs processes of releasing hormones into the body or onto the outter skin.

    The nervous system, or more specifically the brain system, performs a process called intelligence.

    One might define some important parameters of breathing as the capacity of the lungs, the possible exhaling and inhaling pressure, and how well oxygen is pumped in - and CO2 "sucked" out, of the internal to lung chamber barrier.

    The process of intelligence has two important parameters. Notice they are identical to those of the computer processor.

    1. Speed
    2. Bandwidth (in a computer)

    Speed can be inhibited, such of that in a mentally retarded person, or someone consuming inhibitory medications.

    Bandwidth would be determined by the number of neurons in a given area, or in the brain in general if comparins cross-species.

    Your wording here is slightly out of scew. You say that "activity in the brain is a form of intelligence". But a correct phrase would be more like "acitivity in the brain uses the process of intelligence."

    Your question of what delineates it from other processes in the brain... Mainly that it's normal called intelligence in areas of the brain that, throughout evolution, have grown to be what we refer to as "higher processing" areas. For instance neural processes in the medula, or the cerebellum, are not refered to as intelligence.

    This is because, general speaking, the changes in lineage of the brain stem sections do not greatly vary in size in a brain/body weight comparison - however the areas of the brain which touch the outter skull - all the lobes, are greatly enhanced in some creatures.

    So what is refered to as intelligence are the neural processes in areas which are considered variable in humans and in some other species.

    I think this is based on an earlier comment, I may have mistated something. But math aptitude is certainly a function of intelligence no doubt. But it's also a function of teaching skills, of attention span and ability, of interest etc. See below for some good info on this...

    Earlier I believe I stated that indeed even in the scientific community Intelligence is not well understood, as far as what is known, and how it is defined, in the specific fields. In that small niche of fields it is understood.

    Lemme give a comparison here. An evolutionary biologists understands evolution well, no doubt. When teaching to a class the biology may make a comment like the following:

    "There is a mechanism in predators known as the risk of ruin. It's an observation that a creature such as a lion, will attack a small rabbit with the same sneak and attack skills, the same speed, and the same srength as it does an Elk or a Zebra. It does this because it knows that it needs to be as sure as possible that each hunting attempt is a success, for itself and for it's children."

    Now this evolutionary biology knows that the lion does not "know" this. The lion isn't conscious about this at all. It's a personification to simplify to students.

    In reality what occurs is this:

    "The lions which attack all of their prey will full force are more likely to have greater number of successes, whcih provide the young with more food to grow stronger and thus they are selected FOR."

    But the teacher often chooses to use the personification to simplify the conversation.

    This occurs as well when scientists attempt to communicate information regarding intelligence - for the sake of making it easier on others.

    One might say - the ability to process math equations is a form of intelligence.

    But in reality what occurs is someone who has a greater "bandwidth" and "speed" of processes would likely be able to "deal with" any input information better than the average person.

    But again, a greater number of neural connection can be specific to one area of the brain, and be normal in another.

    Regarding the latter comments. You made the comment that intelligence is not well understood. This is kind of a yes and a no. What scientists have chosen to define as intelligence is well chosen - and the process itself, on the level of a neuron is very well understood.

    But what's the challenge is going from the single neuron to understanding entire embedded tissue layers of neurons, and bring this outward towards the entire processing human. It's an enormous step from small to big - and we are most certainly in an infancy, perhaps not even born yet!

    The AI shows this - some of the best ways to understand are to copy. AI has run into issues, and forced to be "rethought" as they say. AI started more as an attempt to fake intelligence rather than to copy it. Whereas neural networks attempt to copy it.

    It's definetely something that, if only showing us one thing, has shown us we have a hell of a long way to go.

    I hope that makes a bit of sense - it's early/late!
    Last edited by a moderator: Aug 23, 2003
  26. Aug 23, 2003 #25
    One of the main concerns, if I may call it that, with the entire genre of Intelligence is the publics popularity to complicate what is really much more simple than they would thin, or perhaps want to think.

    For some reason we find in many facets of culture - or society I should say, that when something becomes extremely complex, humans have a tendency to begin to think there is some extraordinary, or even some supernatural occurence.

    Apparently when some seems to be so monumental, either a massive natural event or the massive processes of the brain, humans dislike having to deal with it part by part (through the billions) and instead want to make it easier on themselves by attributing some other event or even entitity to this.

    We find this most certainly with "intelligence". Hypnagogue commented on the societal understanding of intelligence versus the scientific one.

    In fact they tend to be more similiar. It's nothing more than the compilations of millions of neurons in a particular facet of the brain (and billions in total) the create some sort of "intelligence" in the full on human experience.

    Biology's fundamental unit is the cell, of which a neuron is. Thus Biology looks at this issue from this level, and attempts to build from it. Here we have less accumuluation of error since we know the unit from which are working.

    The societal approach is to see the experience of inteligence in your friend or a classmate, and attempts - with no tools and probably no knowledge - to work downwards....

    ...what we come up with is the want to think that there is "something more" than neurons firing.

    It may seem like scientist is almost using the same word for something totally different - like society calls what science calls cars, cows - and science vice versa.

    But it's sciences approach to the issue "from the ground up".
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook