Audio & Alternative Medicine More Letters
In awe of George
Editor: I am in awe of George Reisch's response to the idea of blind testing ("Undercurrents," September 2000). Let me see if I can summarize his argument:
Blind testing will never work because there are too many uncontrolled variables: bad power, lousy mood, I'm not holding my mouth right, whatever...
The whole purpose of blind testing is to remove these variables to the extent possible. However, the problem with Mr. Reisch's position is that he has this inexhaustible list of variables that he can add to the pile at will and then say, "See, I told you: too hard, too complex." The whole argument gets a little silly in that, by implication, any measure of human performance, according to Mr. Reisch, is ultimately not measurable. Just too many damn variables.
Let's, for the sake of argument, say that Mr. Reisch is correct. He forgets that repeated measurements over time, or measurements over multiple people, control for this very randomness he dislikes so much. That's how science works.
Mr. Reisch also fails to see certain implications about his position that should be of concern to him. For example, I hope Mr. Reisch, following his own argument, would agree that the anticipation of sex is as much of a mind thing as sex is itself. As Mr. Reisch clearly values his music, can we infer that the anticipation of a new cable is almost climactic? (Sorry, couldn't resist.) And, just maybe, that this anticipation changes his expectations, and therefore perceptions, just a tad more than random variations of the electrical grid system?
Let's go another step. If Mr. Reisch is correct, according to his own logic, he must have real trouble evaluating the contribution of any new component to his system. Is it the component, the electrical grid, his mood, the distracting smell of his lover's perfume? Should we therefore assume that Mr. Reisch can't evaluate anything and should be fired from his job as evaluator?
Mr. Reisch argues that mistakes might be made. Well, duh, yeah. If mistakes are made in a controlled environment, how many mistakes are made in a totally uncontrolled environment? And which process, over time, is self-correcting?
There is nothing wrong with applying metrics to the human experience. As a matter of fact, this application of metrics has done lots of good (and, yes, bad) things for us. If Mr. Reisch thinks that listening to music is too complex to measure, how about measuring the effects of new drugs? These effects are measured all the time using rigorous, scientific methodology. And double-blind tests.—Steve Devan, Newport News, VA, firstname.lastname@example.org
In awe of Ross
Editor: I loved Ross Salinger's letter to the editor in the September Stereophile (pp.9-10). I not only agree 100% with Mr. Salinger, but his letter was almost therapeutic. I read too much that leaves me so skeptical, the hairs on the back of my neck stand up. In a clear and thoughtful way, Mr. Salinger stated things that I've felt for a long time.
Will people believe anything that sounds reasonable? This dread of objective listening tests damages my trust in what I read, but because I love music and delight in great sound, there's nothing I want more than finding a way to improve my music. I crave a box attached to a laptop that randomly switches between two components and asks, "Do you prefer A or B?" a dozen times or so, after which it would tell me if I really do prefer one over the other, or if it's all a trick of anticipation.
I gave a wine-tasting party a few years ago, for which I brought home a dozen empty bottles that had held nice wine from a local restaurant. I cleaned the bottles and filled them with Yosemite Road Burgundy. The unsuspecting tasters were between 30 and 40 years old and of varying levels of sophistication. What happened surprised me. I'd bring out another bottle, and after everyone had a taste, someone would say something like, "Oh, this one is much more tart than the last." Someone else would take another taste and say, "You're right, it is!" Another person would mention how much better they liked this wine than the last.
I never said a word about the wine, outside of reading the labels aloud. I asked myself if people were going along with others' comments so as not to look ignorant, but the closer I watched, the more I was convinced that they were sincere. The exact same wine was experienced by the group as sweeter, more tart, bitter, dryer, etc. About half the people loved at least one of the "wines," and about a quarter of them disliked one.
I can't keep from remembering this experiment party when I read some of the ads and articles in audiophile magazines like Stereophile, on the Web, and in the occasional book on high-end stereo. It seems undeniable to me that there is at least some of this placebo effect happening. The fanatical resistance to anything that might compensate for it leaves me frustrated and, as I said before, incapable of trusting what I'm being told.
I could go on about wanting reproducibility, wanting improvement in the face of doubt, and wanting truth independent of the reporter's state of mind. If I have to know what I'm hearing beforehand, or to believe in order to get the benefit, then I'll keep my money and my disbelief. I crave a reporter with a healthy amount of doubt and skepticism coupled with an honest love of music, and that's exactly how Ross Salinger's letter sounded to me.—Tony Earl, email@example.com