The first part of my reply could have been better. I should have said "unbiased" rather than "accurate". The training part is very much about learning to iden

The first part of my reply could have been better. I should have said "unbiased" rather than "accurate". The training part is very much about learning to identify specific chemicals, or groups of chemicals. That absolutely is raw detection, as you say. Let me provide some examples. For the "level 1" training, the most basic stuff, samples would be things like: plain water, acid (very dilute sulphuric acid), alkali (very dilute sodium hydroxide), metal (copper and nickel metal coins in a bottle of water), rubber (rubber tubing cut up in a bottle of water), salt, sugar and a few others I've forgotten (this was 22 years ago), maybe some other major categories like aldehydes. Just for the record, you don't actually drink any of that stuff down, it's tasting only and its very dilute! When the solutions are all clear, being able to identify each blind from some anonymised samples is surprisingly difficult, especially when diluted so it's more subtle. But it's absolutely possible with practice. The higher levels are all more subtle things like many different flavour compounds. For wine and beer, almost exclusively aromatic hydrocarbons, along with esters and other flavour compounds. I never did this but it's the same process, but it's much more difficult. Instead of under 10, you have an industry-standard reference set of IIRC around 60 compounds, and you have to be able to identify each and every one blind when diluted in isolation. And obviously in a drink you have to be able to identify them all individually in combination as well. This is the part that takes the most time and effort to master. And it's these people that I wouldn't question too much regarding their skills, because they actually had to pass objective assessments to demonstrate those skills. This is really hard! This is why I'm a little disappointed that the original article mentioned using students. They would be unlikely to have achieved that level of competence. It takes many months, if not years, of regular practice to get to that level. I spent a year doing the basic training and then continual daily testing and monthly assessments, and while I still class myself as being at a very basic level, I'd still likely have more practical experience than them. Regarding subjectivity and the "experience" of taste, that's absolutely taken on board as well. But it's a separate question. For where I worked, every product had a spec sheet, mostly physical and chemical properties but also a taste profile. The professional tasters could make sure it met that profile exactly. The rest of us just made sure it looked and tasted as a consumer would expect it to. For those who are wondering about why those specific "level 1" tastes, they aren't random and they are actually serving an additional purpose. Those are to also pick up on product contamination. Metal and rubber from storage tanks and pipework. Acid and alkali from cleaning agents and coolant. I actually had to check for the latter once when we had a potential set of hairline cracks in a tank suspected of leaching coolant [in beverage production, they use very concentrated KCl since it's non-toxic if it leaks, but at pH14 it's nasty stuff].