i suspect this violates iswc's reprint policy, but anyway, the paper should be here.
this paper doesn't answer your question directly. but current research does, and the answer is simple: the cost of misinformation. if the cue is correct, then leaving it up improves performance more than a subliminal cue can. but computers are imperfect (news flash!). so the memory glasses are also imperfect; sometimes they will get it wrong. (you should already know this about face-recognition technology, which is our primary current working domain.)
anyway: if you give somebody an incorrect cue in a readable way, they tend to then give the wrong answer. however, if you give an incorrect cue subliminally, you can at least in some cases still improve their chances of getting the *correct* answer anyway. and subliminal misinformation doesn't interfere with processing the way overt misinformation does.
for gorier details, i refer you to the paper.
it was peer-reviewed. iswc (the ieee wearables conference) has a peer-review process, and we went through a round of revisions requested by the proverbial peers before publication. i suspect we will try for a journal publication (probably in a cognition-and-perception context) after we get some more data, and that will involve its own round of peer review as well.
(i'm one of the authors, but i'm the one from harvard medical, not one of the ones from mit:)
i'm one of the authors on the paper, and you should check out the section on "miscues" for why this doesn't work.
subliminal cueing works like this: let's say you teach somebody some name-and-face pairs -- "anne" and "becky". then you show them anne's face and subliminally cue with "anne", and you can improve the person's likelihood of remembering that name.
but let's say you "miscue" -- you show them anne's face but subliminally cue with the name "becky". they are *not* likelier to then type "becky" -- but they *are* likelier to correctly type "anne"! this is the really weird and interesting part of our findings.
we hypothesize that there is some of what psychologists call "spreading activation" taking place: the miscue helps you remember other things you learned in the context of the experiment, but doesn't interfere with the actual production of the correct answer.
anyway, this is why subliminal advertising doesn't work. if you see the word "coke" but what you want is "lemonade", maybe you are likelier to think about getting a drink, but you'll likely get yourself a lemonade rather than a coke.
we have some preliminary data showing that *overt* cues don't work that way. if we show the name "becky" with anne's face in a non-subliminal way, then subjects appear to type "becky" a lot of the time. this is probably why overt advertising actually does work, too.
i've been an embarrassingly devoted tolkien fan of a particularly narrow subgroup -- those of us devoted to his created languages.
in general, i think the linguistics work shown in the movie was pretty high-quality, but i was sad to see some of the (relatively little) elvish from the book left out. we have "mellon" and "noro lim, asfaloth!" and that's great.
but how are they going to make it through the rest of the trilogy without the plot-point of "a elbereth gilthoniel" (a prayer, more or less)?
i thought it was also particularly sad that nobody in the movie said "elen sila lumenn' omentielvo" ("a star shines on the hour of our meeting"; a formal elvish greeting). in particular, in the books, that does a lot to set up the elves' reactions to frodo (which i paraphrase as "look, it's so little and cute, and it can *talk!*":). and it might have been a pleasant addition to the (out-of-book-but-i-agree-reasonable-for-cinema) meeting of aragorn and arwen.
tolkien is often said to have written lotr so that he would have created people to speak the languages he invented. i loved the movie, but it makes me wistful to see the people and hear the voices, but to be missing the words that tolkien especially loved.
"freedom of the press" has always belonged to "them that owns the presses". as far as the internet goes, the commercialism thing cuts two ways:
1, are we willing to pay to own our own "presses", and to do the work to provide meaningful and interesting content with them? i have to wonder if the increasingly broad net community doesn't consist largely of people trained by other forms of traditional mass media to be "audience", to listen in silence, not to publish or converse or even talk back. the web provides the technology for people to do otherwise, but it's still up to the individual to decide whether or not to *do* it.
2, are we, as web users, deciding to just go see what cnn says today, or do we go to any lengths to look around a little more for broader perspectives? do we decide to click on the big shiny button, or go to the search engines and dig up the sites about our own weird little interests? it's ten seconds' effort to go to a search engine and thence to a page on modern poetry written in latin -- but i have to decide i'd rather read latin today, instead of yet another osama bin laden rant and an ad for cellphone service. caveat lector!
one more note:
i couldn't figure out from the article whether the total quantity of "non-commercial" web traffic was on the increase. it wouldn't surprise me if it were, and was only decreasing in terms of percentages, rather than actual quantity. i know that my websites' traffic continues to go up, and i suspect that the topics i talk about (mushrooms and religion, formal poetry, folk-goth music, my cats) aren't seeing much competition from the corporate world.
Slashdot Mirror
i suspect this violates iswc's reprint policy, but anyway, the paper should be here. this paper doesn't answer your question directly. but current research does, and the answer is simple: the cost of misinformation. if the cue is correct, then leaving it up improves performance more than a subliminal cue can. but computers are imperfect (news flash!). so the memory glasses are also imperfect; sometimes they will get it wrong. (you should already know this about face-recognition technology, which is our primary current working domain.) anyway: if you give somebody an incorrect cue in a readable way, they tend to then give the wrong answer. however, if you give an incorrect cue subliminally, you can at least in some cases still improve their chances of getting the *correct* answer anyway. and subliminal misinformation doesn't interfere with processing the way overt misinformation does. for gorier details, i refer you to the paper.
it was peer-reviewed. iswc (the ieee wearables conference) has a peer-review process, and we went through a round of revisions requested by the proverbial peers before publication. i suspect we will try for a journal publication (probably in a cognition-and-perception context) after we get some more data, and that will involve its own round of peer review as well.
:)
(i'm one of the authors, but i'm the one from harvard medical, not one of the ones from mit
i'm one of the authors on the paper, and you should check out the section on "miscues" for why this doesn't work.
subliminal cueing works like this: let's say you teach somebody some name-and-face pairs -- "anne" and "becky". then you show them anne's face and subliminally cue with "anne", and you can improve the person's likelihood of remembering that name.
but let's say you "miscue" -- you show them anne's face but subliminally cue with the name "becky". they are *not* likelier to then type "becky" -- but they *are* likelier to correctly type "anne"! this is the really weird and interesting part of our findings.
we hypothesize that there is some of what psychologists call "spreading activation" taking place: the miscue helps you remember other things you learned in the context of the experiment, but doesn't interfere with the actual production of the correct answer.
anyway, this is why subliminal advertising doesn't work. if you see the word "coke" but what you want is "lemonade", maybe you are likelier to think about getting a drink, but you'll likely get yourself a lemonade rather than a coke.
we have some preliminary data showing that *overt* cues don't work that way. if we show the name "becky" with anne's face in a non-subliminal way, then subjects appear to type "becky" a lot of the time. this is probably why overt advertising actually does work, too.
in general, i think the linguistics work shown in the movie was pretty high-quality, but i was sad to see some of the (relatively little) elvish from the book left out. we have "mellon" and "noro lim, asfaloth!" and that's great.
but how are they going to make it through the rest of the trilogy without the plot-point of "a elbereth gilthoniel" (a prayer, more or less)?
i thought it was also particularly sad that nobody in the movie said "elen sila lumenn' omentielvo" ("a star shines on the hour of our meeting"; a formal elvish greeting). in particular, in the books, that does a lot to set up the elves' reactions to frodo (which i paraphrase as "look, it's so little and cute, and it can *talk!*" :). and it might have been a pleasant addition to the (out-of-book-but-i-agree-reasonable-for-cinema) meeting of aragorn and arwen.
tolkien is often said to have written lotr so that he would have created people to speak the languages he invented. i loved the movie, but it makes me wistful to see the people and hear the voices, but to be missing the words that tolkien especially loved.
1, are we willing to pay to own our own "presses", and to do the work to provide meaningful and interesting content with them? i have to wonder if the increasingly broad net community doesn't consist largely of people trained by other forms of traditional mass media to be "audience", to listen in silence, not to publish or converse or even talk back. the web provides the technology for people to do otherwise, but it's still up to the individual to decide whether or not to *do* it.
2, are we, as web users, deciding to just go see what cnn says today, or do we go to any lengths to look around a little more for broader perspectives? do we decide to click on the big shiny button, or go to the search engines and dig up the sites about our own weird little interests? it's ten seconds' effort to go to a search engine and thence to a page on modern poetry written in latin -- but i have to decide i'd rather read latin today, instead of yet another osama bin laden rant and an ad for cellphone service. caveat lector!
one more note: i couldn't figure out from the article whether the total quantity of "non-commercial" web traffic was on the increase. it wouldn't surprise me if it were, and was only decreasing in terms of percentages, rather than actual quantity. i know that my websites' traffic continues to go up, and i suspect that the topics i talk about (mushrooms and religion, formal poetry, folk-goth music, my cats) aren't seeing much competition from the corporate world.