by Lady Anubh de Mona

Recipe as starting point

The winning entry of the “Pre-1000 CE” Tempore competition at 12th Night this year was a recreation of a brew found in the burial of the Bronze Age Egtved Girl. For this entry, I adapted a general “Nordic Grog” Bronze Age recipe to focus more narrowly on just the brew from the specific find of the Egtved Girl. Research into the archaeological find of the Egtved Girl burial revealed that the grog recipe listed some ingredients that were not actually found in the Egtved bucket residue analysis. In addition, basic knowledge of the science and traditional techniques of brewing made other procedural changes advisable. Sometimes in food and drink projects, entries are judged based on how closely the entry adheres to a recipe. But there are no written  recipe documents from prehistoric times, so reenactors must make guesses based on archaeological evidence. In addition, here the “original” recipe was not focused tightly enough on a single find, but was instead inspired by several different finds. So in this case, a published recipe was the starting point for more informed recipe development.

The Egtved Girl 

The Egtved Girl was thought to have been a teenager, aged 16-18 years, when she was buried in an oak log coffin in Jutland, southeastern Denmark.The coffin’s rings allowed for dendrochronological dating to 1370 BCE, during the Bronze Age. She wore a wool crop-top and string skirt, bronze bracelets, and at her waist was a bronze belt-disc with a spike on the front. At her head was a birch bark box containing bronze pins, an awl, and a hair net. The cremated bones of a 5 or 6 year old child were wrapped in cloth by her feet, and she was wrapped in a wool blanket and cowhide. Also at her feet was a birch bark bucket containing a beverage. It is this beverage I sought to recreate for the Tempore entry.

Photo from news website Sjaellandske Nyheder https://www.sn.dk/art3460013/danmark/egtvedpigen-er-ikke-fra-danmark/

The Brew

Residue analysis determined the beverage in the bucket was probably fermented with sugars from wheat, honey, and fruit. Patrick McGovern is a chemist-anthropologist widely referred to as the father of chemical residue analysis in archaeology, and he calls prehistoric combination drinks like this “extreme beverages,” or “grogs”. In his book Ancient Brews, Rediscovered and Re-Created (2017), McGovern included a recipe created by Doug Griffith called Kvasir. I used this as the base recipe for my Tempore entry. McGovern also collaborated with Dogfish Head brewery on a series of Ancient Ales, including one called Kvasir which is similar to the recipe in his book. The Dogfish Head Kvasir was marketed as being based specifically on the Egtved Girl find– there was even a picture of the Egtved Girl on the label! However, Griffith’s recipe diverged from the Egtved Girl’s brew in some important ways, and teasing out those differences was a big part of the work for this project.

Label from a Dogfish Head brew created in collaboration with Patrick McGovern. 

https://www.mazzaart.com/designs/y5htccph4eorhf8746c81k0pucwe6g

Residue Analysis and Ingredient adjustments

A dark crust in the bottom of the Egtved bark bucket was first analyzed at the time of the discovery by Prof. Bille Gram in the 1920s. His analysis was referenced by T. Thomsen in the initial publication presenting the Egtved Girl finds, where Gram is quoted as stating: “The examination identifies cowberry or cranberry, some cereal grain (bread wheat (Triticum)), glandular hairs of bog myrtle (Myrica gale) along with a large amount of pollen, including small-leafed lime (Tilia)” (Thomsen, 1929). In 1977, a re-analysis of pollens from the residue by J. Troels-Smith, using more modern microscopes, both fine-tuned this ingredients list and noted an important difference from what Gram had found. 

In particular, Troels-Smith did not find any Bog Myrtle (myrica gale) when he re-analyzed the same residue, despite looking at thousands of pollen grains in the Egtved residue. He did find Bog Myrtle in residue from a bronze bucket found at Juellinge, also in Denmark. Griffith did not include Bog Myrtle in his recipe, so that was not a change I needed to make–but others doing recreations of the Egtved brew, including Snoremark brewery in Denmark and Dogfish Head in the US, do include Bog Myrtle. Faced with the two conflicting reports, I prefer to trust the more recent analysis and leave the Bog Myrtle out. 

Residue analyses revealed indicators for wheat malt, honey, and meadowsweet, and those ingredients are unproblematic. However, several other ingredients, listed below, are included in Griffith’s recipe but were not present in the Egtved Girl bucket residue. Griffith may have added:

• Barley (hordeum vulgare) because beer equals barley in many people’s minds;
• Hops (humulus lupulus) because he didn’t feel safe without an antimicrobial and some countries have regulations about requiring hops in beer; 
• Yarrow (achillea millefolium) because yarrow flowers were placed on the top of the cow hide in the Egtved Girl’s coffin (yarrow was not found in the residue, however); 
• birch syrup or birch bark (betula pendula)  because the brew was placed in a birch bark bucket –Birch sap was found in residue from the Kostraaede strainer southwest of Copenhagen, dated to 1100-500 BCE, but no sap was detected in the Egtved bucket;  
• pectic enzymes to clarify, which would not have been necessary if he didn’t puree the lingonberries; see the section on methods below; and
• Dehydrated Malt Extract because he didn’t want to bother with the step of mashing the malt, which would have required keeping the pot of water and malt between 150-160° F for an hour.

I could be wrong about Griffith’s reasons for making the choices he did. But since my goal was to recreate the Egtved Girl brew, and none of those ingredients were present in the Egtved bucket residue analysis, I did not include them.

Next, the fruit used in the brew is reported to be either cranberries, or lingonberries (Vaccinium vitis-idaea, also referred to as cowberries), or both, depending on the publication. The Griffith recipe calls for both cranberry concentrate and dehydrated lingonberries. I prefer to work with raw or dried plants when doing recreations, so initially, instead of using cranberry concentrate, I thought it would be preferable to boil fresh cranberries and strain them, to obtain the flavor. But then I learned that European and American cranberries are different species of plants and don’t taste the same: the American cranberry is vaccinium macrocarpon, while the European cranberry is vaccinium oxycoccus. European cranberries are more tart and bitter than the American variety. European cranberries are difficult to source and both Gram and Troels-Smith, who did the actual analyses, indicate the Egtved brew contained the seeds of EITHER cranberries OR lingonberries, not both. So it seems reasonable to leave the cranberries out and use only dried lingonberries.

I also chose to add linden (tilia) as an additive, which was not included in Grifith’s recipe. Linden was one of the most common pollens identified in Troel-Smith’s analysis. There is no clear pollen threshold  that can be used to determine whether any pollens are from: a) honey, b) contamination from nearby growing plants, or c) intentionally added to a beverage. However, some pollens are less likely to be additives than others. I chose to use white clover honey in this brew, based on the high prevalence of clover pollen in the sample, combined with the understanding that clover is not commonly used as an additive in any brewing traditions of which I am aware. While linden is also commonly interpreted as an indicator for honey, it is drunk as tea, and its bitterness and antimicrobial potential make it a suitable replacement for hops and/or bog myrtle.

Finally, for the sake of full disclosure, the residue analysis also indicated high levels of  pollen from the crucifer family (Brassicaceae) –possibly charlock (Sinapis arvensis). This is a kind of mustard common as a weed; it has no tradition as an additive for beverages and its flavor would not be complementary in this recipe. Griffith did not include it, and neither did I. This pollen may have been from plants growing nearby at the time of burial.

Methods requiring adjustment

Griffith’s recipe is described as being inspired by the Egtved Brew, but it is designed for modern homebrewers, not Bronze Age reenactors. It includes modern shortcuts and methods that can easily be improved upon.

First, the treatment of grains in Griffith’s recipe should not make sense to any all-grain brewer. It seems like the author wanted to play with grain but didn’t trust himself, or his readers, to be able to actually get the sugar from the grain. He instructs us to bring the crushed malt and water to 170° F, just above the temperature at which it will start the starch conversion process, but then he immediately stops it before much diastatic action can really occur. If we simply kept that water-malt mixture between 150-160° F for a full hour, enzymes in the malt would convert the starches to sugar. After that we could rinse, strain and discard the grain, which at that point would be “spent,” meaning all the sugar was taken out of it. Instead, in Griffith’s recipe, as soon as the malt reaches 170°, the still-unspent grain, with all its delicious sweet potential, is removed and not mentioned again. Dried Malt Extract (DME) is then immediately added, to replace the sugars lost from lack of a proper mash time. This is a big waste of grain. DME was not available in the Bronze Age and it’s really not that hard to keep a pot covered and warm for an hour. It makes a lot more sense to give the malt an hour’s mash-time at temperature– that way, no DME is necessary.

Next are issues with the treatment of the fruit. After making this recipe myself, I think the amount of dried lingonberries called for by Griffith is excessive, as the brew is extremely tart. In addition, I believe the acidity of the berry puree may have interfered with fermentation in my brew–it stopped fermenting when I added the puree. Griffith also called for the lingonberry-wort mixture to be pureed prior to adding it to the brew–but doing so caused a large amount of sediment and wasted brew.When made that way, fully half the brew wound up being sediment! Food processors were not available in the Bronze Age, and the identification of lingonberries in the bucket residue was possible because there were whole seeds present. Pureeing the berries pulverizes the seeds and destroys these remains. Instead, it makes more sense to steep a smaller quantity of whole dried berries without pureeing them. Then the berries can be strained out (I used a mesh strainer but a linen bag could have been used for this in period) and the liquid can be added to the brew. A reduced concentration of berries allows the acidity to be low enough that it does not affect fermentation. My recipe calls for the herb-berry tea to be added at the end of the boil, but the berries and herbs could probably also be added at the beginning of the boil, since the quantities of berries and antimicrobial herbs have been reduced.

Figuring out all of these adaptations was a slow and messy process, which required multiple iterations as I continued to brew and do book-research at the same time, learning new things along the way. My final recipe is below; I hope to display this final version at either L.E.A.F.S., Gem Joust, or Victory on the Vine this spring, so look for me at A&S tables there for a taste! 

	Griffith Recipe	Residue Analysis	Anubh Recipe
Wheat Malt	✔️	✔️	✔️
Barley	✔️	❌	❌
DME	✔️	❌	❌
Hops	✔️	❌	❌
Honey	✔️	✔️ clover pollen	✔️
Birch	✔️	❌	❌
Lingonberry	✔️	“lingonberry or cranberry”	✔️
Cranberry	✔️	“lingonberry or cranberry”	❌
Linden	❌	✔️ pollen	✔️
Meadowsweet	✔️	✔️ pollen	✔️
Yarrow	✔️	❌	❌

Anubh’s version – Egtved 3

Ingredients

3  gallons filtered water

5 pounds white wheat malt 

1 Tablespoon meadowsweet

1 Tablespoon linden flowers/leaves

1.5 pounds clover Honey

4 oz dried lingonberries

2 cups Yeast solution (I used elderberry yeast)

1. Bring 1.5 gallons water to 170F
2. Add malted wheat in a bag
3. Mash for 1 hour in oven preheated to 170F and turned off
4. Rinse with water at 170F, until you have 2.5 gallons of wort 
5. Bring the wort to boil.
6. To a saucepan add meadowsweet, linden, and lingonberries
7. Cover berries and herbs with boiling wort
8. Keep the berry botanical infusion hot but not boiling while the wort boils 30 minutes
9. At flame-out strain the berries and herbs from the fluid in the saucepan, then bring the strained infusion to boil briefly.
10. Add 1.5 lb honey and the berry-herb infusion to the wort and let it cool.
11. When wort is room temperature, add yeast solution. 
12. Allow to ferment for a week or more
13. Bottle or keg. Drink when you think it tastes good–this brew is delicious as a short mead but a few weeks aging in the bottle will allow the cloudiness to clear.

Filling the birch bark bucket with Egtved Girl Ale

The Original Recipe by Griffith, from McGovern, 2017

Homebrew Interpretation of Kvasir, by Doug Griffith

Ingredients

5 gallons cool water

¾ pound Briess special roast malt, crushed

6 pounds Bavarian wheat dry malt extract

1 ounce Tettnang hop pellets

1 Tablespoon meadowsweet

1 Tablespoon yarrow

8 ounces birch syrup – OR 1 T ground birch bark and ½ lb corn sugar

3 pounds honey

8 ounces cranberry concentrate

2 packet Wyeast 1728

½ pound dried lingonberry

1 teaspoon pectic enzyme

Process

1. Fill a brewpot with 5 gallons of cool water.
2. Fill the grain bag with the Briess malt. Tie off the top and place the bag in the brewpot.
3. Heat the pot and stir the water and grain bag every 5 minutes
4. As the water reaches 170 F, pull out the grain bag using a large stirring spoon. Hold the bag above the brewpot for a minute, allowing most of the liquid to drain into the pot. Do not squeeze the grain bag. Continue heating the water.
5. As the water is beginning to boil, remove the pot from the heat.
6. Add the dry malt extract. Stir to prevent clumping and scorching on the bottom of the pot. Return the pot to the head.
7. Allow the wort to come to a boil
8. After boiling for 5 minutes, att eht Tettnang hop pellets and stir.
9. Start timing and 1-our boil at the point the the hops addition.  If using a defoamer to help prevent boilovers, add per instructions.
10. Put the lingonberries in a blender, cover with liquid from the brewpot, puree, cool, and add the pectic enzyme. Refrigerate for 1 day.
11. At 30 minutes before the end of the 1-our boil, in a separate small pot, remove 2 cups of the wort from the brewpot. Heat to just keep hot, not boiling.  Add the meadowsweet, yarrow, and one of the birch options.
    Option 1: Add the corn sugar and birch bark.
    Option 2: Add the birch syrup
    Stir well to make sure all is wet. Steep until the end of the boil.
12. At the 60-minute mark, turn off the heat. Add the honey, cranberry concentrate, and strained liquid from the steeping pot.  Stir the wort for 2 minutes to build up a whirlpool effect and dissolve the honey.  Stop stirring and allow the wort to sit for 10 minutes.
13. Chill the wort with a wort chiller or in a cold-water bath until is is under 75 F.
14. Transfer the wort into a fermenter; aerate (rock the baby) for 1 minute. 
15. Pitch the yeast into the fermenter.
16. Top up the fermenter to the 5-gallon mark with cool water. 
17. On the second day of fermentation, add the pureed lingonberries.
18. In about 14 days, the beer should be ready to bottle.  The beer can be siphoned to a carboy for further clearing, if desired, for about 7 days.
19. Before bottling, clean and sanitize the bottles and caps and create a priming solution of 1 cup boiling water and the priming sugar.
20. Siphon the beer into a sterilized bottling bucket, add the water-diluted priming solution, and gently stir.  Bottle and cap the beer.
21. Allow the beer to bottle-condition for another 10 days at 70-75F. It should then be ready to drink.

Target ABV 8.5%

References

Jaswa, Kyle and Jazwa-van den Berg, Kimberley (2017) Meadowsweet in Archaeology. Website https://brewingclassical.wordpress.com/2017/10/30/meadowsweet-in-archaeology/

Website written by a couple who are both archaeologists and brewers. 

McGovern, P.E. (2017) Ancient Brews Rediscovered & Re-created. W. W. Norton & Company, New York.

McGovern describes nine extreme fermented beverages of our ancestors, including the Midas Touch from Turkey and the 9000-year-old Chateau Jiahu from Neolithic China, the earliest chemically identified alcoholic drink yet discovered. Homebrew interpretations of the ancient drinks are provided, with matching meal recipes.

National Museum Denmark. The Egtved Girl’s Beer. Website https://en.natmus.dk/historical-knowledge/denmark/prehistoric-period-until-1050-ad/the-bronze-age/the-egtved-girl/the-egtved-girls-beer/

Museum website; picture-heavy and light on text.

Dineley, Merryn (2014). Beakers were for Beer! part two: a birch bark bucket with residues. Blog post

https://merryn.dineley.com/2014/04/beakers-were-for-beer-part-two-birch.html

Blog written by an archaeologist and her brewer husband.

Thomsen, T. (1929) Egekistfundelt fra Egtved fra den Aeldere bronze Alder. Nordiske fortidsminder / Utgivne af det Kgl. Nordiske oldskriftselskab 2:165-214.

Title Translates to: “The Oak Coffin Found in Egtved from the Early Bronze Age” This was the initial publication of the Egtved Girl finds.

Troels-Smith, Jørgen (2018)  Modern pollen analysis and prehistoric beer – A lecture by Jørgen Troels-Smith, March 1977.  Review of Palaeobotany and Palynology 259 (II) https://www.researchgate.net/publication/327629028_Modern_pollen_analysis_and_prehistoric_beer_-_A_lecture_by_Jorgen_Troels-Smith_March_1977

Lecture given after re-analyzing the Egtved Girl bucket residue.