葡萄酒复杂性的神秘面纱
演讲者首先邀请听众闭上眼睛，想象品尝一杯葡萄酒的完整体验：倒酒、摇晃、闻香、啜饮。接着，当邻座评价“这是一款非常复杂的葡萄酒”时，引出了核心问题：“复杂性”究竟意味着什么？在葡萄酒的世界里，这是一个被广泛使用却定义模糊的术语。即使是专业的酒评家，在品鉴笔记中也可能用到“复杂性”一词，但若询问十个人，可能得到十种不同的答案。因此，本次演讲旨在探索“复杂性”这一概念，并揭示在嗅觉和味觉领域，我们如何能够体验到通常只在艺术领域才能找到的复杂认知。在此之前，将首先进行一个简短的艺术史引入，然后回归到葡萄酒这一“好东西”。
Original English
So I would like everyone to close their eyes, just for a second, and imagine that you're holding a glass of wine. And really imagine that experience. Take the wine, swirl the wine, smell it, and take a sip. Now imagine that someone next to you leans over and says, "Mm, that was a really complex wine." OK, what did they mean by that? What exactly does complexity mean? It's a term that we use a lot in the world of wine. Wine critics use it, right? When I write a tasting note, I sometimes use complexity. But I will tell you an industry secret. If you ask 10 different people, "What does complexity mean," you might get 10 different answers. So today I want to explore this idea of complexity to show that actually, when it comes to smell and taste, we can get a lot of complex cognitive ideas that usually are reserved for the realm of art. So a very small art history introduction first, and then we'll get to the good stuff, which is the wine.化学复杂性：杯中之酒的科学
在心理学中，我们常遇到解释“复杂性”与“喜好度”之间关系的贝尔莱因曲线 (Berlyne curve)，它呈现一个倒U型。这意味着，过于简单的东西人们不喜欢，而过于复杂的对象同样难以获得好感，存在一个“理想复杂性”的甜蜜点。为了演示，演讲者制作了几个音乐片段，从极简到逐渐增加复杂度的不同风格，展示了在听觉上，当复杂性过高时，大多数人可能不再喜欢。
但这种关系在葡萄酒世界中是否成立？引用著名酒评家马特·克莱默 (Matt Kramer) 的观点：“评估葡萄酒品质的最高标准是复杂性。你回看一杯酒的次数越多，发现不同的东西，无论是香气还是味道，这款酒就越复杂。最伟大的葡萄酒并非压倒性，而是看似取之不尽。” 这句话充满诗意，但也给作为科学家的演讲者带来了挑战：如何量化并研究这个模糊的概念？
演讲者提出，理解葡萄酒的复杂性，可以从两个层面入手：一是化学复杂性 (chemical complexity)，即酒杯中实际含有的化学分子；二是心理复杂性 (psychological complexity)，即复杂性存在于饮酒者的心智中，而非酒本身。
第一个研究，考察了“杯中之酒”的化学复杂性。在牛津大学的一项研究中，实验者将三款单一葡萄品种、同一生产商的葡萄酒进行50/50的调配，制造出六款酒（包括原酒和调配酒）。参与者在不知情的情况下品尝这六款酒，并进行评分，包括猜测是否为混酿、命名风味，以及评估喜好度、熟悉度、复杂性、风味强度、品质和支付意愿。
出人意料的结果是：总体而言，参与者无法在盲品中准确区分混酿与单一品种酒。更令人惊讶的是，那些对葡萄酒知之甚少的初学者，反而比有认证的品酒者更能“高于随机概率”地猜对混酿。研究还发现，混酿葡萄酒的复杂性评分并不比单一品种酒更高。这表明，杯中的化学复杂性并不一定等同于人们感知的复杂性。
然而，研究确实发现了一个能预测人们复杂性评分的因素：橡木桶的使用。当品尝者在风味描述中提及“香草 (vanilla)”、 “辛香 (spice)”或“肉桂 (cinnamon)”时，这会显著预测更高的复杂性评分和支付意愿。因此，尽管化学复杂性不直接等于感知复杂性，但通过橡木桶陈酿，似乎能让人们觉得酒更复杂，并可能更喜欢它。
Original English
So in psychology, we often see this curve, this Berlyne curve, that explains the relationship between complexity and liking. And the idea here is that it's an upside-down U curve, so if something is too simple, people don't like it. On the other hand, if something is too complex, people also don't like it. So there's a sweet spot in the middle, like an ideal complexity. I made some audio clips to demonstrate this idea. So here's a simple one. (Music note) So maybe if you're into very modern music, you might think that was really good. Otherwise you might find that a bit boring. So here's another one with a little bit more complexity. (Musical tune) These are my original compositions, by the way. And here's a third example where there's a lot of complexity. (Musical tune) So how many people like that one? OK, we have a few in the audience. Very nice, very nice. But I wanted to demonstrate the point, right? When you have too much complexity, often people don't like it. But is that true in the world of wine? I chose one quote from a famous wine critic, Matt Kramer, and let's see what he said about complexity. "The greatest standard used in assessing the quality of wine is complexity. The more times you can return to a glass of wine and find something different, either the bouquet or the taste, the more complex the wine. The greatest wines are not so much overpowering as they are seemingly limitless." So this is a very beautiful quote, right? Very artistic. And you start thinking, OK, so what exactly is complexity? I work as a researcher, I'm a scientist. In my work, I need to try to define what one critic says, right? This is a very vague idea. So how do I exactly go in and investigate this concept? And the approach I took, I will demonstrate a little bit today. There are actually two ways of thinking about complexity when it comes to wine or food. There's, first, chemical complexity, right? That's the stuff that's in the glass. What are the chemical molecules in the wine? And then there's a concept called psychological complexity, where complexity is not really in the wine, but it's in the mind of the drinker. And I will show two different studies. The first one: we investigate the idea of chemical complexity -- it's complexity in the glass. And in the second study, we'll look at psychological complexity. It's complexity in the mind of the drinker. So let's look at the first one. This was a study I conducted at the University of Oxford some years ago. And the idea was, OK, if I have two wines and I make a 50/50 blend, surely the blend should chemically be more complex, right, than the individuals. So here we took three wines, single grape variety from the same producer, and then I made 50/50 blends of these three wines for a total of six. It looks something like this. When someone came to the lab to do the study, they were presented with six wines. They were not told which one was which, right? We didn't tell them at all about which are blends, which are single variety. They were just told, here are six red wines, I would like you to taste them. And people made some ratings. For example, they had to guess whether the wine was a blend or not. They had to name the flavours and then they evaluated the wine based on things like liking, familiarity, complexity, flavour, intensity, quality and willingness to pay. So I want you all to just take a moment and think. Do you think that people could actually tell, in a blind tasting, which were blends and which were single varieties? So maybe hands up, how many people think yes, they can tell the difference? OK, maybe one or two or three. Most people think no. So I'm going to tell you something that was a very surprising finding. Because I had around 80 people in this study, there were some people who knew nothing about wine, they just worked in the department and they wanted to come for free wine. And we had some other people who had different wine certifications. I was running the Oxford Blind Tasting Society, so I kind of also got people in society to do the study. So the surprising finding was, if you looked at everybody overall, people couldn't tell blends from single varieties. But the fun thing was that actually it was the beginners, the people who didn't know that much about wine, who actually could guess blends or not higher than chance. Very surprising. It was the beginners who did better. I don't know why, but that was one of the surprising findings. Another finding we found was that the blends were not rated as more complex than the single varieties. This was maybe also a little bit surprising. What it showed is that chemical complexity in the glass does not necessarily equal perceived complexity when people are rating the wine. A final surprising finding from this study was that there was actually one thing that could explain people's complexity rating. It wasn't whether it was a blend or not. It was how much oak they could taste in the wine. So people who wrote things like vanilla or spice or cinnamon, writing those words in the flavor was actually predictive of higher complexity and also more willingness to pay. So at the end of the story, it seems like chemical complexity doesn't really equal perceived complexity, but if you put your wine in oak, people are going to think it’s more complex, and they might like it more.心理复杂性：品鉴者的心智与时间维度
化学复杂性研究之后，演讲者转向了“心理复杂性”——即“美在于观赏者眼中”，或者说“复杂性在于品鉴者的口中”。这引入了一个新的概念：时间。在品尝时，复杂性可以从两个维度定义：一是静态复杂性 (static complexity)，即在任何特定时刻感知到的风味数量；二是动态复杂性 (dynamic complexity)，即品尝者在品饮过程中，感受到的风味如何随时间在口中演变和变化。为了形象地说明这一点，演讲者使用了音乐片段作为类比。
为了研究葡萄酒的心理复杂性，尤其关注其动态演变，演讲者选择了一类极其适合的酒品：马德拉酒 (Madeira wines)。因为在葡萄酒界普遍认为，马德拉酒陈酿时间越长，复杂性越高。本次研究与Justino’s Winery合作，使用了三种不同陈酿年份（3年、10年和20年）的马德拉酒。为了消除视觉干扰（如酒的颜色随陈酿而变化），研究采用了黑色的品酒杯。
研究参与者包括约70位新手和30位专家。首先，他们被要求盲品并从1到9分评价酒的复杂性。结果显示，新手对陈酿年份与复杂性的关联性感知不明显（增长趋势不显著），而专家则能一致地将更陈年的葡萄酒评定为更复杂。这证实了专家确实能感知到陈年带来的复杂性增长。
更关键的是，研究采用了TCATA（Temporal Check All That Apply）方法，即参与者在品尝过程中，每秒钟需要勾选他们感知到的八种风味描述符（如焦糖、橙皮、苦味等），并持续记录风味在口中的演变，甚至包括吞咽后的余味。这项“疯狂的游戏”旨在精确测量风味随时间如何变化。
研究展示了一段视频，直观呈现了三种不同年份（3年、10年、20年）Tinta Negra葡萄酿造的马德拉酒的风味轨迹。对于专家而言，20年陈酿的酒（用深色线条表示）在风味空间中的演变轨迹明显不同于另外两种，显示出更复杂的风味层次和变化。然而，新手在观看同一视频时，三条曲线几乎重叠，表明他们难以区分和感知风味随时间的变化。这再次印证了，复杂性不仅在于酒本身，更在于饮酒者心智的差异。
Original English
So that was the first study. For the second one, I'm going to go back to this idea of psychological complexity. So, you know, it's beauty in the eye of the beholder. Or is complexity in the mouth, I guess, of the taster? And here, I wanted to introduce a different concept, which is this idea of time. Complexity, when you're tasting something, there can be two different ways of defining it. You can have static complexity. This is at any moment in time, you know, how many different flavors do you perceive? Or there's an idea of dynamic complexity, which is, imagine, go back to your imagination. You're tasting the wine in your mind. If you pay attention to how the wine is evolving in the mouth over time, you can actually follow how it changes. So I also made some audio clips to demonstrate this idea. The first one is a static complexity. (Music note) So standard major triad, nothing complex. And -- it's complex. The second one is dynamic complexity. (Musical tune) So when I run these studies, I use music a lot as a way to demonstrate these concepts because everyone knows that music is something that evolves over time, right? So you can listen to a change and the same thing can be applied to flavor. So for this study, I really needed to find some wines that are very complex. Because we're in Portugal, but also because I love Madeira, it's my favorite fortified wine, we ran a study with Madeira wines at different aging points, because it's a well-known fact, at least in the wine world, that the older the wine, the older the Madeira, specifically, the more complex it gets. This is what everyone in the wine trade believes, but we wanted to see if this is actually true. So for this study, I worked with Justino’s Winery, and we had Madeira samples that are three, 10 or 20 years old, aged in barrel. Again, the study looked something like this. People got six glasses, and you might think, OK, these glasses look different from the other study. Why are they black? So the reason why they're black is, if anyone has had Madeira or port before, I'm in Portugal after all. As the wine ages, the color will change. So if we used clear glasses here, people could just notice by the color how old the wine might be. So to eliminate that factor, we used black glasses here. You can also see that they have randomized three-digit numbers. So we really made this study where people had no idea exactly what they were tasting. They were told it was Madeira tasting. And for the study, we had both novices and experts. This time I had about 70 novices and 30 experts come do the study, so I could also compare between them. Now, the first thing they had to do is just taste the wine and rate how complex they think the wine is. On a simple scale, one to nine, how complex they think this wine is. And the results were quite interesting. So on this graph, I tried to make it simple. On the horizontal axis, we have the three different aging groups. So three years, 10 years, 20 years. The green line represents the novice -- the people who don’t know that much about wine. You can see that there's a slight upwards curve, right? So people kind of think as the wine gets older, it gets more complex. Actually, this is not statistically significant. But if you look at the ... [experts] in the orange line, right, the curve is much more steep. And here we can actually see that for experts, in a blind tasting, with black glasses where they couldn't see the color, they actually do consistently rate older wines as more complex. So in other words, there is something to this idea that as the wine ages, it gets more complex, as we can see here. The second thing that we asked people to do, and this will take maybe a little bit more explanation, is that we asked them to do a test that we call TCATA, temporal check all that apply. What happened was that as people were tasting the wine, they had eight different descriptors -- that sounds like a crazy video game -- that they had to check for every descriptor they were perceiving at any moment in time. So imagine holding the wine in your mouth for 30 seconds. For each second, you had to tick all the descriptors like, do I taste caramel now? Do I taste orange zest now? So it was a constant ticking action. And the reason why we did this kind of crazy game is to really measure, as you're holding the wine in your mouth, how do the flavours change over time? And also people spit the wine out at some point, but we kept on measuring the aftertaste. So we had this measurement of both the in-mouth sensation and also what happened afterwards. And the reason why I'm explaining is because I'm going to show a crazy video trying to demonstrate what happened with the wines. So let me just talk you through it really quickly. What you're about to see is the temporal trajectory of how the wines change in the mouth over time. The simple way to interpret this graph is if you’re closer to, for example, orange zest in the middle, the closer the line gets to orange zest, the more you taste orange zest. The closer the line gets to bitter, the more you taste bitter. I'm going to illustrate the trajectory of three different wines. Tinta Negra is the grape variety, and they’re aged three, 10 and 20 years. Basically, the darker the line, the more aged it gets. So I'm going to start the video. Basically for every two seconds, the lines will move in this space, in this flavor space. Every two seconds equals 15 seconds in real time of evolution in the mouth. And this is the result of experts. So let me play the video, and then we can walk through it. (Video) So you can see that the three curves are different, right? Especially the dark one in the middle, for the 20 years. The flavor evolution is a little bit different from the other two. And now, if you compare that with how the novices do it, ... the curves don't really deviate from each other so much. They all kind of follow each other. And this says something about how novices perceive the flavor evolution. Whereas experts are able to clearly differentiate the three lines, for the novices, it's much more of a similar experience.复杂性、偏好与品鉴的艺术
综上所述，正如研究结果所示，葡萄酒的复杂性不仅仅在于杯中的客观化学成分，更深刻地体现在饮酒者的大脑中。新手与专家由于经验和认知的差异，对风味演变的感知截然不同。这引发了一个关键问题：在葡萄酒领域，更多的复杂性是否总是等同于更好的品质？这与音乐领域的情况可能不同，在音乐中，过度的复杂性往往会降低听众的偏好。然而，在葡萄酒品鉴中，目前尚无确凿证据表明过度的复杂性是有害的。或许，正如“美在于观赏者眼中”一样，“复杂性”的真正价值，最终取决于品鉴者的心智体验。
演讲者的最后寄语是：下次当你品尝任何食物或饮品时，不妨花些时间，放慢脚步，静下心来，仔细体会其中风味如何随着时间流转而演变。感谢各位。