inding fraud is unlikely but strong evidence against the claim that people can tell the difference—because it crucially depends on the priming effects [...] dominating.
Wait, can you expand on how fraud in the trade of a non-alcoholic drink is “strong but unlikely” evidence that people cannot tell the differences between wines?
Such fraud might be evidence that people cannot tell the difference between tastes more generally, but that seems like a higher hurdle to clear.
Wait, can you expand on how fraud in the trade of a non-alcoholic drink is “strong but unlikely” evidence that people cannot tell the differences between wines?
Where did I say that? The claim you quoted from me was about “what would be evidence of people’s ability to distinguish that non-alcoholic drink”, not wines.
You’d check for people’s ability to distinguish wines by fraud in wines, and people’s ability to distinguish specific non-wine drinks by fraud in specific non-wine drinks.
I thought I made that very clear, and if not, common sense and the principle of charity should have sufficed for you not to infer that I would get the “wires crossed” like that.
(ETA: By the way: is there a shorter way to refer to a person’s ability to distinguish foods/drinks? I’ve tried shorter expressions, but they make other posters go batty about the imprecision without even suggesting an alternative. Paul Birch suggests “taste entropy of choice”, but that’s obscure.)
And, to preempt a possible point you may be trying to make: yes, fruit drinks may be fraudulently labeled as real fruit juice, but that’s not a parallel case unless people claim to be able to distinguish by taste the presence of real juice and purchase on that basis.
So we’ve averted misunderstanding, good. My question remains: what does fraud (or non-fraud) in non-alcoholic drinks tell us about whether people “really can tell the difference” between wines?
Just to be sure what you’re claiming, btw: if I did a “triangle test”, blinded, on two arbitrary bottles of wine from my supermarket, and I could tell them apart, would you retract that claim? Or is your claim restricted to some specific varietals?
How could I have explained my position better so that you would not have inferred the point about fruit drinks?
My question remains: what does fraud (or non-fraud) in non-alcoholic drinks tell us about whether people “really can tell the difference” between wines?
It doesn’t remain. If people can tell the difference, you don’t gain from fraud. If they can’t, you could gain from fraud. Where’s the confusion?
Just to be sure what you’re claiming, btw: if I did a “triangle test”, blinded, on two arbitrary bottles of wine from my supermarket, and I could tell them apart, would you retract that claim? Or is your claim restricted to some specific varietals?
I accept that you could tell red from white, so it couldn’t be completely random. I’d want a test over the two varieties they said were swapped in the story, or within the same variety, but significant cost difference.
How could I have explained my position better so that you would not have inferred the point about fruit drinks?
I have inferred nothing about fruit drinks. In this comment you replied to me with an allusion to “other drinks”. Later in the same comment you referred to milk. In other words, you primed me to think about non-alcoholic drinks. Later in the exchange, you ruled out the possibility that fraud in non-alcoholic drinks will provide any evidence relevant to your original claim. So we’re better off dropping this line of inquiry altogether.
I’d want a test over the two varieties they said were swapped in the story
The story mentioned three varietals: merlot, shiraz and pinot noir. It seems likely that in the next few days I will be able to procure half-bottles of a Merlot (vin de pays, 2005) at €6.5/l and a Pinot Noir (Burgundy, 2005) at 14€/l. The experiment will set me back about 10€.
Do you consider the experiment valid if I buy the bottles myself, and have a third person prepare the glasses? Would you care to stipulate any particular controls? Are you willing to trust my word when I report back with the results? Do I have to correctly identify the cheaper and the more expensive wine, or just to show I can tell the difference? What will you bet on the outcome?
There’s a difficulty that affects tests like these: you can be much more sensitive when you know you’re being tested, so it wouldn’t tell as much about people who can be fooled when given wine that they didn’t expect something was wrong with. The test would have to be made slightly more difficult in order to account for your preparation.
Here’s what I would consider a fair test: I pick three wines, (or imitations of wines) and then do a modified version of the triangle test: I can make all three the same, all three different, or just two the same, and not tell you which. To pass, you have to correctly rank their prices and identify which, if any, are the same. (You’d be allowed to count two different ones as effectively the same prices if they differed only by a few Euros.)
And of course, the test would have to be triple blind, etc.
You can be much more sensitive when you know you’re being tested
Why thank you. So, you are expecting that in some situations people actually can tell wines apart?
Here’s what I would consider a fair test
Sounds complex. I might play, but you’d have a) to pick three wines I can actually pick up in my local supermarket chain, Monoprix; they have an online catalog; b) to write the instructions with enough precision that a local confederate can carry them out. I’m willing to do the work myself for the simple case (have done so in the grandparent), but not if the rules get too baroque.
There’s one thing I object to. Why do I have to rank them by price? I am not claiming that the more expensive wines are systematically better, taste-wise, than the cheaper ones. I do claim that the very good stuff is more likely to be found in the more expensive bottles. There is no reason to expect that price will be systematically correlated with my idiosyncratic, relatively untrainted tastes.
If I have to rank them by price via recognition, I’d have to have tried them out beforehand. If your hypothesis is correct that shouldn’t affect the results, since you’re claiming my discrimination of “better” comes from priming effects.
Why thank you. So, you are expecting that in some situations people actually can tell wines apart?
Sure, just like how, if you put enough effort in, you can tell any two drinks apart. The difference is that people make these claims about wine without have done the exercises necessary for precise discernment, yet claim these subtleties matter to them.
Sounds complex. I might play, but you’d have a) to pick three wines I can actually pick up in my local supermarket chain, Monoprix; they have an online catalog; b) to write the instructions with enough precision that a local confederate can carry them out. I’m willing to do the work myself for the simple case (have done so in the grandparent), but not if the rules get too baroque.
You’re really not grasping the concept of biasproofing, are you? “A confederate” taints the experiment. Your purchasing the wines taints the experiment.
There’s one thing I object to. Why do I have to rank them by price? I am not claiming that the more expensive wines are systematically better, taste-wise, than the cheaper ones.
Okay, well, the wine cheerleaders do, so this is one way you break from them and agree with me.
If I have to rank them by price via recognition, I’d have to have tried them out beforehand. If your hypothesis is correct that shouldn’t affect the results, since you’re claiming my discrimination of “better” comes from priming effects.
Does not follow: if you’re told in advance which has the label “good”, you can spit that information back out latter. My claim is that a judgment of good dependent on being told in advance that it’s good is not a genuine judgment of good.
You’re really not grasping the concept of biasproofing, are you? “A confederate” taints the experiment.
Not performing the experiment taints it even more. But perhaps by now we’ve learned enough of each other’s claims. I leave it to you to find a way forward.
Wait, can you expand on how fraud in the trade of a non-alcoholic drink is “strong but unlikely” evidence that people cannot tell the differences between wines?
Such fraud might be evidence that people cannot tell the difference between tastes more generally, but that seems like a higher hurdle to clear.
Where did I say that? The claim you quoted from me was about “what would be evidence of people’s ability to distinguish that non-alcoholic drink”, not wines.
You’d check for people’s ability to distinguish wines by fraud in wines, and people’s ability to distinguish specific non-wine drinks by fraud in specific non-wine drinks.
I thought I made that very clear, and if not, common sense and the principle of charity should have sufficed for you not to infer that I would get the “wires crossed” like that.
(ETA: By the way: is there a shorter way to refer to a person’s ability to distinguish foods/drinks? I’ve tried shorter expressions, but they make other posters go batty about the imprecision without even suggesting an alternative. Paul Birch suggests “taste entropy of choice”, but that’s obscure.)
And, to preempt a possible point you may be trying to make: yes, fruit drinks may be fraudulently labeled as real fruit juice, but that’s not a parallel case unless people claim to be able to distinguish by taste the presence of real juice and purchase on that basis.
So we’ve averted misunderstanding, good. My question remains: what does fraud (or non-fraud) in non-alcoholic drinks tell us about whether people “really can tell the difference” between wines?
Just to be sure what you’re claiming, btw: if I did a “triangle test”, blinded, on two arbitrary bottles of wine from my supermarket, and I could tell them apart, would you retract that claim? Or is your claim restricted to some specific varietals?
How could I have explained my position better so that you would not have inferred the point about fruit drinks?
It doesn’t remain. If people can tell the difference, you don’t gain from fraud. If they can’t, you could gain from fraud. Where’s the confusion?
I accept that you could tell red from white, so it couldn’t be completely random. I’d want a test over the two varieties they said were swapped in the story, or within the same variety, but significant cost difference.
I have inferred nothing about fruit drinks. In this comment you replied to me with an allusion to “other drinks”. Later in the same comment you referred to milk. In other words, you primed me to think about non-alcoholic drinks. Later in the exchange, you ruled out the possibility that fraud in non-alcoholic drinks will provide any evidence relevant to your original claim. So we’re better off dropping this line of inquiry altogether.
The story mentioned three varietals: merlot, shiraz and pinot noir. It seems likely that in the next few days I will be able to procure half-bottles of a Merlot (vin de pays, 2005) at €6.5/l and a Pinot Noir (Burgundy, 2005) at 14€/l. The experiment will set me back about 10€.
Do you consider the experiment valid if I buy the bottles myself, and have a third person prepare the glasses? Would you care to stipulate any particular controls? Are you willing to trust my word when I report back with the results? Do I have to correctly identify the cheaper and the more expensive wine, or just to show I can tell the difference? What will you bet on the outcome?
There’s a difficulty that affects tests like these: you can be much more sensitive when you know you’re being tested, so it wouldn’t tell as much about people who can be fooled when given wine that they didn’t expect something was wrong with. The test would have to be made slightly more difficult in order to account for your preparation.
Here’s what I would consider a fair test: I pick three wines, (or imitations of wines) and then do a modified version of the triangle test: I can make all three the same, all three different, or just two the same, and not tell you which. To pass, you have to correctly rank their prices and identify which, if any, are the same. (You’d be allowed to count two different ones as effectively the same prices if they differed only by a few Euros.)
And of course, the test would have to be triple blind, etc.
Why thank you. So, you are expecting that in some situations people actually can tell wines apart?
Sounds complex. I might play, but you’d have a) to pick three wines I can actually pick up in my local supermarket chain, Monoprix; they have an online catalog; b) to write the instructions with enough precision that a local confederate can carry them out. I’m willing to do the work myself for the simple case (have done so in the grandparent), but not if the rules get too baroque.
There’s one thing I object to. Why do I have to rank them by price? I am not claiming that the more expensive wines are systematically better, taste-wise, than the cheaper ones. I do claim that the very good stuff is more likely to be found in the more expensive bottles. There is no reason to expect that price will be systematically correlated with my idiosyncratic, relatively untrainted tastes.
If I have to rank them by price via recognition, I’d have to have tried them out beforehand. If your hypothesis is correct that shouldn’t affect the results, since you’re claiming my discrimination of “better” comes from priming effects.
Sure, just like how, if you put enough effort in, you can tell any two drinks apart. The difference is that people make these claims about wine without have done the exercises necessary for precise discernment, yet claim these subtleties matter to them.
You’re really not grasping the concept of biasproofing, are you? “A confederate” taints the experiment. Your purchasing the wines taints the experiment.
Okay, well, the wine cheerleaders do, so this is one way you break from them and agree with me.
Does not follow: if you’re told in advance which has the label “good”, you can spit that information back out latter. My claim is that a judgment of good dependent on being told in advance that it’s good is not a genuine judgment of good.
Not performing the experiment taints it even more. But perhaps by now we’ve learned enough of each other’s claims. I leave it to you to find a way forward.