I spent the holiday week with a few friends at a ranch outside Telluride and took the opportunity to explore the Manhattan with them. The Manhattan is the root of more cocktails than any other drink. By my count, there are 169 direct variants on Kindred Cocktails, and the tree expands considerably when you trace the Martini (and all its variants) via the Martinez back to the Manhattan (Difford’s Guide for Discerning Drinkers has a nice history of the Martini). The cool thing is that all these drinks are often one or two ingredient substitutions away from the original Manhattan. If you are working with a limited home bar, the density of variations and the tastiness of most of them is very encouraging. The trick is getting the ratios right, and, as I discovered on my holiday, “right” is a personal thing.
The Manhattan is a three ingredient drink comprised of rye, sweet vermouth and bitters. As with the Old Fashioned, I prefer to use overproof rye for Manhattans for the same reason, that the additional alcohol serves to amplify the other flavors. Complex sweet vermouths make Manhattans more interesting; normally I select Carpano Antica but used Sacred Spirits English Spiced Vermouth for our holiday drinks. Both are bittered sweet vermouths, so you do not need to use additional bitters if you do not like bitter drinks. I added a dash of Angostura Aromatic bitters. Because I was not sure about my friends’ preferences, I also brought a bottle of Elijah Craig Small Batch Bourbon as a means of diluting the rye.
Since I was traveling light and wanted to experiment for this post, we tried different proportions of rye and bourbon over four evenings: all rye, 3/4 rye, 1/2 rye, 1/4 rye. The panel is truncated because I was not interested in the all bourbon variation, and I made the drinks. Unsurprisingly, people’s preferences varied. Surprisingly, the 3/4 rye and 1/4 rye versions were the most popular. The experiment is not scientific, but the outcome reinforces how idiosyncratic preferences are.
If you do much reading about cocktails, you probably have come across some discussion about modern palates being drier than historically. Drier drinks contain less sugar—think Winston Churchill’s Martini made solely of gin and stirred while glancing at the bottle of vermouth—and perhaps people used to like their cocktails sweeter. Based on our holiday experiment (The Holiday Manhattan Test) and my limited experience making drinks for others, I am a bit skeptical that modern palates are drier. I suspect the preference for dry drinks is learned, as is the preference for bitter drinks. Right on cue, Difford’s Guide recently added the High Ryser, which is decidedly not dry, with equal parts rye and sweet vermouth.
In the Holiday Manhattan Test, I discovered that I liked the 3/4 rye version best because the bourbon substitution exposed the flavor of the Angostura Aromatic bitters from the mid taste with the spices from the rye and vermouth. Interestingly, the vermouth is also bitter, but I found the vermouth bitters more integrated into the mid taste. The addition of 1/4 bourbon created a distinct, bitter aftertaste for me. The addition of more bourbon served to push the bitter further out, creating too much separation between the mid taste and after taste and making the drinks less coherent (spherical, see “How I Think About Cocktails”). The Angostura bitters were superfluous without the slight bourbon dilution.
Does the bourbon dilution result in a “new” cocktail? Possibly, but cocktail creation seems to be more of a continuum. Because ratios are subject to personal preference, very fine gradations would mean an explosion of drinks. More coarse gradations seem to make more sense. Clearly, adding an additional ingredient like an Amaro or a fruit juice would make a different drink, as would substituting another liquor for either the rye or the sweet vermouth, but does replacing some rye with bourbon constitute a different cocktail? I think not. The Perfect Manhattan, which substitutes dry vermouth for half the sweet vermouth (back to the modern palate question), is usually mentioned as a footnote in Manhattan recipes, not as a novel drink itself.
Reducing complex subjects into a simple model enables us to consider things such a boundaries more clearly. I find the Taylor Series from mathematics a useful reduction analogy for thinking about cocktails, both in answering the not particularly useful question “Is my creation novel?” and more useful questions such as “Do I need larger ice cubes?” The Taylor Series Expansion stems from the observation that many continuous functions can be expressed as an infinite sum of the function’s derivatives at a certain point. That’s not important for cocktails, but the analogy becomes useful because of the observation that, generally, the first derivative in the series contributes the most to the sum, the second the next most and often you do not need to worry about things beyond the third or fourth derivative.
With this analogy, we can consider the craft of making drinks. We can equate the ingredients and preparation actions to terms in the Taylor Series. For the Manhattan, the rye is the first term because it has the greatest influence on the experience of the drink. The sweet vermouth is the second term. Chilling is the third order term. Based on the Holiday Manhattan Test, I estimate that the bourbon dilution is the fourth term, and the bitters are the fifth (you could omit them and possibly not notice the omission). Dilution during chilling (by stirring duration and ice cube size selection) is the sixth. All these terms (choices) influence the final product, but each successive one contributes less than the previous term. Ice cube size is a noticeable choice, but it’s a fairly minor influence on your enjoyment of a Manhattan.
Is the Holiday favorite Manhattan with the bourbon dilution a new drink? At fourth order, I think it is on the borderline. After all, the Perfect Manhattan with its fifty percent dry vermouth substitution introduces a new term in the Taylor series that is on par with the second order term of sweet vermouth selection. The bourbon dilution is not as substantial a change, and the Perfect Manhattan usually only rates a mention in Manhattan recipes. As with so many things regarding cocktails, the novelty comes down to personal preference. I would mix the drink differently depending on the audience so I consider the dilution a pleasing tweak and not a novel drink.
Cocktails appeal to me because you are in control of many degrees of freedom. A beer brewer, assuming standard production skills, finds their craft in the selection of water, grains, malts, roasts and hops. These degrees of freedom are approximate equivalents to the Taylor Series terms. I am uncertain which input approximates the first term; it’s likely hops because hop use dramatically changes the characteristics of beer. The second term is probably roasting the barley. Water and the barley itself lend less distinction to the beer so these are higher order terms.
Distillation adds a few more degrees of freedom (alcohol percentage, barrel aging, botanicals). Again, considering core spirits like bourbon and gin where you could start with a beer (tequila is somewhat different), the craft is around barrel aging or botanical selection. Depending on the spirit, barrel aging or botanicals are the first term and alcohol percentage the second. There is an art to distillation columns, but most distillers purchase these from someone else who does the crafting.
If you made it this far, the Belden Anti-hero is a very nice cocktail that prominently features an India Pale Ale. In this drink, the IPA is the dominant first order term. If your bar does not include the other ingredients, the subsequent terms are minor enough that you can comfortably substitute rye for Applejack (to stay on theme) and replace the ginger liquor with a ginger flavored simple syrup (Sasha Petraske’s Regarding Cocktails has a good recipe). Amaro Averna works well in place of Amaro Lucano. Amaros and Amaro substitutions are topics for a future post.