Search tips
Search criteria 


Logo of acscentsciACS Central Science
ACS Cent Sci. 2016 June 22; 2(6): 361–362.
Published online 2016 June 8. doi:  10.1021/acscentsci.6b00159
PMCID: PMC4919770

A Conversation with Hildegarde Heymann

When it comes to the complex flavors and fragrances in a wine, chemical analysis is no match for the human nose and mouth. Sensory scientist Hildegarde Heymann develops and tests methods to quantify what people smell and taste in wine, so that winemakers can tailor their craft to control the flavors in their products. The benches in her lab at the University of California, Davis, are covered with beer, juice, flower petals, boxes of tea, and other ingredients used to make reference standards for flavors and fragrances. Heymann talked to Katherine Bourzac about how she interprets the messy data produced by human senses.

How do you quantify people’s sensory experiences?

In all cases, we have a group we’ve trained to evaluate products using standardized references. Pinot people are doing Pinot, one at a time, for example.

An external file that holds a picture, illustration, etc.
Object name is oc-2016-00159x_0001.jpg

First they sit around and evaluate the products blind over a period of time and come up with words to describe them. Let’s say you have rosés, and they’re all the same pink. Then pink is not a good descriptor because it doesn’t discriminate. But if you have some rosés that are orange and some that are pink, then pink is a good term. Eventually they come to a consensus as to the number of terms, what the terms are, and what the reference standards are. We never ask them what they like, because it’s not the point. They’re trained to pull the wine apart.

We’ll have, say, 10 judges and 30 wines which they taste in triplicate and evaluate for 20 attributes. The numbers become insane. We do a lot of multivariate statistics to try to simplify the data or get it to where we can make some sense out of it. On top of that, you’re dealing with a very noisy measuring device: a human being. People have moods, they have off days and on days, so we have to do lots of replication to make sure they’re consistent.

Do people use different words to describe the same thing?

Oh, frequently. They will use different words to describe the same thing, they will use the same word to describe different things, and they don’t realize it until you start putting reference standards in front of them, and they say, “That’s not what I meant.”

Somebody could say a wine smells like violets when everyone else says it smells like blackberries. Then the panel leader makes reference standards that smell like violets and blackberries. It may turn out that one person’s violet was actually blackberry. Two people might say a wine smells woody, but one means the smell of oak and the other one means the smell of forest-floor debris. We use references to make sure everyone understands the same word the same way.

Sounds like it would be easier to eliminate the people. Why not just use gas chromotography (GC) and mass spectrometry?

Those methods identify things very well—including a lot of things the human brain doesn’t interpret as a smell or a taste. I like to say, “GCs don’t drink wine.” It might be picking up on things that are not relevant. Or a human may be more sensitive than the GC is. Often we can smell something that the GC can’t pick up, even at nanogram and lower levels. That was very true 20 years ago. Now it’s getting less so, but every so often it happens that we know something is there because people smell it, and then five years later they finally find it with chemical analysis.

The GC pulls things apart. To people, combinations of volatiles smell different than volatiles by themselves. Volatiles in a different amount of alcohol smell different. It’s a combination of what happens in the olfactory receptors and what happens in the brain. That’s why you have to have the sensory studies. You can’t just take the chemistry and know how it’s going to taste.

How do winemakers use sensory data?

The winemaker has choices. You can look at where the grapes are grown—we know that makes a difference. We can choose different yeast strains. The winemaker can choose to age the wine or not on its lees, which is the dead yeast, or to use oak barrels or not. Every one of these small changes affects the style of the wine, and every one has sensory and chemical consequences. We don’t tell wine makers what to do, we just tell them that when you do A, B, C, and D, then you get this flavor.

You’ve shown quantitatively that chemical and sensory differences exist between wines from different regions and winemakers. Does this work support some of the mythos about wine and the way people talk about it?

When people talk about smelling violets in wine, I can understand that, because my panelists have reference standards for violets under controlled conditions. A lot of wine writers are just writing things they think, because they don’t use reference standards. I think some of that more flowery writing is somewhat counterproductive when it comes to consumers. They don’t really care whether it smells like violets. All they want to know is, am I going to like this or not?

How do you taste a wine to enjoy it, compared with when you’re studying it?

The number one difference is spitting. When I am drinking wine for enjoyment, the fact that I’m not spitting turns my brain off, and I don’t go into pulling things apart. When I walk in a sensory booth and start spitting, it’s work.

Katherine Bourzac is a freelance contributor toChemical & Engineering News, the weekly newsmagazine of the American Chemical Society. Center Stage interviews are edited for length and clarity.

In collaboration with C&EN.

Articles from ACS Central Science are provided here courtesy of American Chemical Society