I’ve destroyed vast swaths of whitespace and needlessly abused countless thousands of words in my endeavors to describe and explain Viking-age brewing.
I’m a scientist. Screw this “word” stuff. Let’s make something.
Pictured are the fruits of my labor so far – my interpretation of Viking-era “malt,” based on the research that I’ve done to-date. Let’s talk about how I got here.
As I’ve previously explained, I’ve drawn connections between the method for the processing of “polenta” described by Pliny the Elder, the method for producing “zythos” or “zythorum” described by Zosimos of Panopolis, and the analyses of actual bread finds from pre-1000 CE Scandinavia. I’ve also drawn inspiration from a recipe for “Ethiopian beer” documented by Olaus Magnus in 1555, which bears a striking resemblance to all of the other processing methods I’ve documented – and to the method presumably described in “A Hymn to Ninkasi.”
So, my method has borrowed from each source, in an attempt to extrapolate a speculative processing method.
Part I: The Grain Bill
First, this is how Pliny describes the ingredients of “polenta:”
“But whatever the mode of preparation adopted, the proportions are always twenty pounds of barley to three (pounds) of linseed,4 half a pound of coriander, and fifteen drachmæ5* of salt: the ingredients are first parched, and then ground in the mill.“
In Latin: “quocumque autem genere praeparato [vicenis hordei libris] [ternas seminis lini] et [coriandri selibram] [salisque acetabulum], torrentes omnia ante, miscent in mola“
Note that the “drachmae” in the recipe is an interpretation of the original Latin “acetabulum;” according to Wikipedia, the “acetabulum” is a liquid or dry unit of measure with a capacity of 68 mL (1/8 sextarius). 1 tablespoon of salt (15 ml) is roughly 20 grams, which means we’re talking about roughly 90 grams of salt.
The “pound” to which they refer is the “libra” in Latin, which is the equivalent of 328 grams – or roughly 72% of a conventional modern English pound. Thus:
20 libra of barley = ~14.5 pounds = 6.56 kg = ~84%
3 libra linseed = ~2.15 pounds = 984 g = ~12.5%
0.5 libra coriander = ~0.30 pounds = 168 g = ~2.25%
Salt = 90 g = ~1.25%
Total mass: 7820 g
I decided to alter the recipe a bit, to make it a little easier to grasp (and to calculate ingredient amounts), and to standardize it a bit better so that I have a more solid platform for experimenting.
10% oil seed
Of course, the Viking bread was not all-barley. The above-linked finds show that breads could contain barley, oats, and legumes – peas were the particular find.
In order to replicate such a bread, this is the final grain bill that I used:
Viking grain bill (proportions by weight) [500 g batch]
35% barley (un-malted, with husk) [175 g]
35% oats (steel-cut) [175 g]
15% peas (green, dried, whole) [75 g]
10% flax seeds [50 g]
2.5% herb (wild Icelandic thyme) [12.5 g – reduced to 4 g to account for dried herbs]
2.5% Atlantic sea salt [12.5 g]
What’s that? Wild Icelandic thyme?
This was a gift from my younger brother from his vacation in Iceland. The thyme here is dried; since the directions specifically state that the ingredients have to be “parched,” I assume they were starting with fresh herbs. I reduced the amount of thyme used to 1/3 of what I calculated, to account for the difference between fresh and dried herbs.
Part II: Processing the Grains
Zosimos of Panopolis provides a fairly clear method for the processing of the grains, and subsequent conversion to the bread-like substance pictured at the beginning:
Take good pure barley and water, and soak it for a day. Spread it out and put it in a windy place for another day. Again soak it for 5 hours, then collect it in a sieve with handles, and soak it again after it has drained until it becomes puffy.
When this is done, dry it in the sun, until it deflates: The husk is indeed bitter.
Now mill (it), and make a bread-dough, adding leaven* as in bread-making, and bake it very well. Then boil it well, and separate the sweet water, straining it through a sieve.
Some heat toasted bread in a pan with water, and cook it a bit, but neither must he boil it nor heat too long, and taking it from the fire, transfer to other vessels, and again heat and reserve (the liquid).
*Note: According to Pliny, leaven was either made of must and grain, or fermented porridge, or a bit leftover from a previous batch – in other words, sourdough starters.
I am adapting this method 1) to account for the Viking-era grain bill I’ve identified and 2) to account for climatic differences between ancient Rome and very northern Europe.
First, we take all of the ingredients and steep them in water overnight:
Next, we “spread it out and put it in a windy place for another day.” In this case, I spread the soaked stuff onto a baking sheet and put it on my table with the ceiling fan running.
After this, we soak it again for 5 hours, and then drain it in a special vessel, and soak it some more. The archaeological record of the Vikings does not seem to have a “sieve with handles” in the way that Zosimos describes, so I just sort of sprinkled more water on the grain and left it out while I was at work (~7 hours).
The next stage is to “dry it in the sun,” until it “deflates.” Ultimately, this is a method for peeling the grain – soaking and drying will cause the husk to shrivel away from the grain, making separation easier. Now, northern Europe (especially Iceland and northern Scotland) is a cold, wet place. Drying in the sun is unlikely to work. That’s probably why there are so many corn-drying kilns in northern Scotland – they needed a way to dry out their wet grain. Keeping that in mind, I put the baking sheet in my oven at 275 F, until the grain was dried out.
Wow, that looks an awful lot like a high-kilned malt, with a bit of crystallized appearance. Hardly surprising, given the moisture content. Now, we need to “mill” the grain, add “leaven,” and bake the crap out of it.
I did not have a proper Viking quern (hand-cranked two-stone rotary style – very laborious, but it makes flour), so I had to make due with my Barley Crusher malt mill. I ran the grain through 3 times to try to get it finely crushed, but it wound up being a fairly coarse meal.
I used the “boil some meal into a porridge and let it ferment” method of leaven. However, as I was on a time budget, I also added a pinch of baker’s yeast and a small dollop of the liquid from some plain yogurt. A sourdough is, after all, a symbiotic system of lactic acid bacteria and yeast – and Pliny’s methods of leaven would very likely result in a sourdough. Ideally, the starter would have been a bit of the dough leftover from a previous batch – but as I had no previous batch, this was not possible.
The “starter” was fermented overnight, then mixed in with the coarsely-ground meal and some water until it achieved a dough-ish consistency, and fermented overnight again. After that, I spread the dough mixture out into rounds ~6 cm in diameter and ~0.5 cm thick, and baked them at 300 F until they were rock-hard. See that first picture.
That gives us the Viking “malt.”
Part III: The Brewing
Zosimos similarly describes a method by which the bread can be processed into a beverage. However, I’ve identified two different speculative processing streams – one to make “brauð,” and the other to make the wine-like beverage “öl”. I’ve drawn my inspiration for the wine-like beverage from the work of Olaus Magnus, who described an “Ethiopian” beer made from sourdough bread mixed with water and honey.
For both products, I have used a 1:4 ratio of solid:liquid (by volume) in constructing my recipe. This is essentially the ratio documented by Magnus (effectively 1 part bread, 3 parts water, and 1 part honey), and is the ratio very commonly used in the ancient world for the production of mead. 16th century accountings of “ordinary” beer, such as the one described by William Harrison, also use a ratio of roughly 1:4 grain:water by volume (after accounting for differences in units in use at the time). Given that “ordinary” beer was intended as a common drink, I suspect this ratio may have echoes in a far earlier era, where processing methods had not yet evolved into totally separate specialized activities.
One of the biscuits I’ve baked occupies roughly 1/4 cup when crushed up.
For brauð: 8 biscuits were crushed and mixed with 2 quarts of cold tap water in a pot. The liquid in the pot is slowly brought to a boil (took about 1.5 hours) and boiled for ~5 minutes. The liquid is strained into another container, allowed to cool, and then poured into a jug.
For öl: The same essential method is used, though ratio varies a bit. 8 biscuits were crushed and steeped in 6 cups (1.5 quarts) of cold tap water – that’s a ratio of 1:3 biscuit:water by volume. As above, the mixture is heated slowly (~1.5 hours), brought to a boil, and boiled for ~5 minutes (until the protein foam subsides). This mixture was allowed to cool in the pot for ~30 minutes before being strained as above; the warm-to-the-touch liquid was poured into a different jug, and 2 cups of local raw honey were poured into the jug. The jug was shaken to ensure that everything was dissolved.
Why raw honey? Because while the Vikings had honey and apiary technology, they did not have the high-pressure filtration methods we have today. Any honey they used would have been full of pollen and wax. This particular honey has the comb removed, but still contains pollen – and also wild yeast and/or bacterial spores. Raw honey will ferment at about 17% moisture, so this will be an excellent vehicle for promoting wine production.
Both jugs have been left on my counter with the tops open, to promote a sort of open, wild fermentation. The saga of St. Olaf talks about ale being ladeled from an open cauldron into cups – indicating that fermentation was probably carried out in open containers. In the case of brauð, they may have simply left the liquid in the pot in which it was first cooked, or they may have transferred it to another vessel as Zosimos recommends.
These will be fermented (well, hopefully they’ll ferment!) until Saturday, where I’m teaching this whole thing (plus the entire Brewing with Egil series) as a class at the East Kingdom Brewing University this Saturday.
Summary of Process
1) Assemble the grain bill: 35% barley, 35% oats, 15% peas, 10% flax seeds, 2.5% herbs, 2.5% salt. Steep ingredients in water for ~24 hours.
2) Spread out grain and place it in a breezy location for ~24 hours.
3) Re-water the grain and allow it to stand for ~8 hours.
4) Dry the grain by low direct heat (an oven set to 275 F, for example).
5) Grind the grain into a coarse flour/fine meal.
6) Mix the meal with a sourdough starter (ideally a bit left from a previous batch) and some water, and allow to ferment overnight.
7) Form the fermented dough into cakes ~6 cm in diameter and ~0.5 cm thick. Bake at ~300 F (again, relatively low temperature) until they are dried and hard.
8) For brauð, use 1 biscuit in 1 cup of water. For öl, use 1 biscuit in 3/4 cup water. Crush the biscuit(s) into the water and slowly bring to a boil over a gentle heat.
9) Strain the liquid into an appropriate container. For öl, add 1/4 cup of honey per biscuit to the liquid once it’s cooled (but still warm enough to dissolve the honey).
10) Ferment for ~3 days, and enjoy!
Commentary, Limitations, and References
There are a few notable limitations in this method. First, of course, is that this is all still speculation on my part. There is no written method for the production of these beverages, and it’s unlikely that we’ll ever find one.
My 4 primary process limitations are: 1) lack of a proper rotary quern to produce flour, 2) lack of a proper period corn-drying kiln fired by appropriate fuel, 3) lack of a proper period baking setup to replicate the heat profile, and 4) lack of proper period-traceable ingredients.
A minor limitation was my lack of fully-soured leaven, but this is partly related to having a proper flour, and is generally trivially rectified by simply giving myself more lead time to allow the starter to ferment properly. As for the rest, in order:
1) While I cannot obtain an artifact rotary quern, instructions for making a facsimile using poured cement and pie plate tins exist online. While still not a truly period material, this would provide a flour with a more proper consistency. This is a project for the future.
2) The simplest corn-drying kilns are little more than fancy holes in the ground, dug in a two-bowl style. I am in the process of planning a reproduction of such a kiln; this will enable me to dry the grain using a proper fuel. Research indicates that, in addition to local hardwoods, Icelanders used sheep dung as a fuel source. This would produce a very smoky fire, which would impart a smoky taste to the dried grain.
3) Some archaeological evidence suggests that Viking bread may have been baked on iron pans or directly on burning coals. Given the small size of the extant finds, this seems plausible. Other evidence points to earthern ovens being used at the time. Both methods will be attempted and the results compared side-by-side. This is another experiment which will be attempted in the future.
4) Traceable ingredients are difficult to obtain. A variety of barley called Bere has been examined and traced to the Viking age (and earlier); however, Bere is native to northern Scotland, and importing it is difficult. A small group in western Canada also grows Bere, but the cost of exporting a sufficient amount is prohibitive. I will, in the future, either grow Bere or suck it up and shell out for it. Native oats are easier to obtain, and green peas are mostly unchanged.
It’s worth noting that salt in the Viking age was very very likely produced by being poured over burning wood (a method documented by Pliny as being practiced by the Germanic tribes), which would produce an alkaline, smoky product. My salt research is a completely separate topic, but will definitely have an impact here.
So, I am increasingly confident in my conclusion about Viking-era beer brewing. Now that I have established a baseline method, I can begin experimenting with different elements of the process, in an effort to make them more “period.” However, I believe that my current method is a reasonable representation of a product that likely existed in the Viking age.
Pliny the Elder. Naturalis Historia. Perseus Digital Library. http://www.perseus.tufts.edu/hopper/text?doc=Plin.+Nat.+toc
Magnus, O. A Description of the Northern Peoples. trans. Peter Fisher and Humphrey Higgins, ed. Peter Foote, 1996 Hakluyt Society. (originally published 1555 in Rome.)
Zosimos of Panopolis. De Zythorum… trans Gruner CG. 1814. http://archive.org/stream/zosimipanopolita00zosi#page/n3/mode/2up
- “Ancient Roman units of measurement.” Wikipedia article. http://en.wikipedia.org/wiki/Ancient_Roman_units_of_measurement (I know, I know. Wikipedia. I double-checked the sources and they seem legitimate.)
- Scott, Sir L. “Corn-Drying Kilns.” Antiquity. Vol. 25. Num. 100. pp 196 – 208. Antiquity Publications Ltd, 1951.
Harrison, W. Elizabethan England. From A Description of England. Ed. Lothrop Withington. Project Gutenberg. Released 30 May 2010. EBook #32593. London: Walter Scott. http://www.gutenberg.org/files/32593/32593-h/32593-h.htm
I figured I should add in some pictures of the stuff fermenting. I would up adding a pinch of Munton’s dry ale yeast, just to get the stuff going. Turns out, my apartment doesn’t contain enough wild yeast to start a fermentation. Next time, I’ll just leave the stuff outside.
On the left, we have brauð – the basic beverage used as a food. This one has “fermented” for about 5 days now – longer than was likely typical. However, it’s still a light fermentation, and a stable beverage; the flavor profiles at day 3 and day 5 are the same. It tastes something like a small beer crossed with a broth/stock flavor; it also has a very pleasant citrus-like brightness to it. Almost lemony, actually. This may be a result of the interaction between the wild thyme and the fermentation.
On the right, we have öl. This has some alcohol content at this point, and is the only one of the two with some carbonation. It is, however, quite sweet – it tastes like mostly un-fermented mead. This one may become more alcoholic (and balanced – it’s really damn sweet) with time. I definitely see how something like this could be glossed with “hydromel.”