Shrewsbury Sauce: technology development, path dependence and innovation

Next, add the [pint of] wine, a little at a time, mixing with a wooden spoon after each addition.
— Delia Smith — ROAST LEG OF LAMB WITH SHREWSBURY SAUCE

This is a post about technology, path dependence and why it might not be the right time for your idea… yet. And how to think about that.

Before you go on: Subscribe to The Baseline| Major thinkers, tech experts, entrepreneurs and researchers unpicking how we can build better technology for society and how society is adjusting to technology.

Roast Lamb with Shrewsbury Sauce

My favourite dish to cook is Roast Lamb with Shrewsbury Sauce. It, like all good things in life, requires a pint of wine. Baste the lamb well and the combination of the rich sauce with the crispy skin is wonderful. I am not a food writer; I can’t describe it. It’s just a bloody lovely, rich and indulgent dish. This comes from the fat of the roast lamb, the sweetness of the wine and the red current jelly and the bitterness of the Worcester Sauce and mustard.

It’s common to wonder “how the hell did anyone invent this recipe in the first place” but with a complex dish it’s that much more confusing. Not only does someone have to invent that recipe but each of the ingredients in turn have to be invented. And the ingredients have to be transported. And before that the ingredients have to be cultivated and made to grow at agricultural scale.

Multiple people throughout history have to do “how the hell did anyone invent that” tasks, multiple times for long enough for the dish to be possible.

Where are you going with this, Dan?

This led me to use Shrewsbury sauce as a way of understanding technology history, and perhaps as a way of making me better able to think about technology future. This very quickly sounds like the history of all things: path dependence…

Path dependence is the idea that decisions we are faced with depend on past knowledge trajectory and decisions made, and are thus limited by the current competence base.

In other words, history matters for current decision-making situations and has a strong influence on strategic planning. Competences that have been built in the past define the option range for today’s moves. New business opportunities, in particular those based on technological progress, emerge gradually as a consequence of competencies acquired prior to new discoveries and over time

http://lexicon.ft.com/Term?term=path-dependence

Viewed like this, a (not so) humble recipe becomes the embodiment of hundreds of years of scientific, commercial and cultural advancement. There was a time before you could make Shrewsbury sauce. Part of that was cultural because people didn’t know to make it, but a larger part of it was practical because people couldn’t make it because the didn’t have access the ingredients.

We can only cook with the ingredients we have but because the ingredients we have change over time so the possible recipes can change over time. This is true of all technology (and therefore all things).

But before we generalise, let’s answer: When was the earliest time you could make Shrewsbury sauce? The exercise itself is useful in making a concrete (and delicious) example of technology development, path dependence and innovation.

What was the earliest time in history that I could make Shrewsbury sauce?

The ingredients of roast lamb and Shrewsbury sauce are: lamb (I prefer a shoulder), flour, mustard and few other herbs. Wine, redcurrant jelly, salt and pepper. All easy to buy things today, but that’s a fairly recent phenomenon of globalisation but also of the thousands of years of progress. Let’s find out how early in history we could have made this…

Lamb
The lamb is fairly easy. Any time after 4000 BC when Neolithic settlers brought their horned brown sheep to Britain.

Onions
We can’t be sure about the onion because their fine tissue doesn’t leave a trace and so is unlikely to be dug up and dated. But the onions would have been ready before the lamb, having been cultivated for around 5000 years and eaten wild before that.

Rosemary
Rosemary needed to be brought over as it is native to the coastal regions of North Africa and the Mediterranean. Rosemary was brought over with the Romans, so we can add that to our recipe in 55BC.

Flour
For the flour, wheat farming came to Britain around 3000 BC. Flour has been found off the Isle of Wight which predates this by about 2000 years, but this was thought to have been imported. Flour came with the cultivation of wheat but even if you had flour, I’m not sure it would work as well because it wouldn’t be as fine as the white flour we have these days. The recipe uses the flour to absorb the fat of the lamb and turn it into a rich paste. Courser flour would result in a clumpy paste which might not mix will with the wine and jelly. So we can go with the 3000BC date but it would be a bit crappy.

Black Pepper
The black pepper is native to south India and would need to be shipped over from Southeast Asia where it was grown before the 16th century. This required shipping, and this essay isn’t the place to get into what shipping required.

Mustard
Mustard came over to Britain with the romans. But it wasn’t until the 18th century when the seeds could be turned into the fine mustard flour used in the recipe. In 1720 the development of milling techniques allowed Mrs Clements of Durham to become the centre of mustard production and make a healthy lump of money by selling the fine powder. Mrs Clements is critical to our recipe.

Red Currant Jelly
The red currant jelly is more difficult to know. There are multiple claims to this. It was either invented in France 1364 or in England before the reign of Edward I.

Salt
The salt could have been shipped in, or come from (what is claimed to be) Britain’s oldest working mine. 1844 Winsford Rock Salt Mine.

Beaujolais (the wine)
Beaujolais had been produced from the Gamay grape in the Beaujolais region for years as an annual end-of-harvest wine, but until world war II it was only really used for local consumption. After World War II [how — who did this?], it was marketed more widely including having a special Beaujolais day on which it is released each year. So that moves us to 1945.

Worcestershire sauce
Going back a bit, Worcestershire sauce was invented in the 1830s by John Wheeley Lea and William Henry Perrins. But this introduces even more dependencies:

• Barley malt vinegar
• Spirit vinegar
• Molasses
• Sugar
• Salt
• Anchovies
• Tamarind extract
• Onions
• Garlic
• Spice
• Flavourings

The “flavourings” are believed to include soy sauce, lemons, pickles and peppers

To get just one, Molasses, requires not just the cultivation of sugar (let’s say in the west indies) but that in turn requires the discovery of sugar cane, the establishment of suitable farmland and (in the history which we humans built for ourselves) the use of slavery and shipping.

Recipe for making Shrewsbury sauce (time: c2 hours)

To make the Shrewsbury sauce you roast the lamb on chopped onions, basting it plenty of times throughout. When the lamb is done, you remove it from the roasting tray, place the trace on the hob and mix in the other sauce ingredients, taking the flavour of the lamb and the onions into the sauce.

The recipe for inventing Shrewsbury sauce (time: c5000+ years)

Embodied in the flavour of a single mouthful is thousands of years of human progress.

The entire cooking and preparation time is a few hours. In order for cooking and prep to take a few hours, eleven ingredients have had to be cultivated and traded over thousands of years. The methods of grinding flour and then mustard had to be discovered, honed and propagated throughout the world. Lea and Perrin had to experiment, nearly throw away and then rediscover their sauce. The recipe requires invasion, wars, invention, slavery and countless other human sacrifices and advances.

It requires humans, animals and plants from multiple continents to work together for thousands of years to make sure all the ingredients are available to someone in the west midland town of Shrewsbury, England.

Oh — As far as I can tell, none of these ingredients has anything to do with Shrewsbury aside from that’s where they were all brought together. This makes Shrewsbury the same as Lea and Perrin, grinding mustard: active progress. Or, as we normally call it: invention.

Recipe for the recipe inventing Shrewsbury sauce (Read that sentence twice)

So how did all that happen?

There were three types of progress in the invention of Shrewsbury sauce, all of which you can see in the invention of all things.

These are:

• Inevitable progress
• Active progress (invention)
• Periods of lack of both progress and lack of regress (i.e. survival)

Inevitable progress
Stuff that’s going to happen anyway: inevitable progress. This is stuff which will happen regardless of most factors and is much like a brute force approach. The spread of cultivating crops or sheep coming with early humans to Britain. These were likely to happen just because of the sheer numbers of humans doing the same thing and the fact that it was demonstrable a good idea.

Active progress or invention
Stuff that someone specific has to make happen: active progress or invention. Some advances in technology require someone to do it: Lea and Perrin needed to try recipe after recipe. Mrs Clements needed to apply the method of grinding flour to mustard seeds to make a fine mustard flour.

Lack of both progress and regress
Stuff that has to stick around for the next thing: Neither progress nor regress.

This third advancement is simply sticking around. If Rosemary had been brought over by the Romans in 55BC and hated by everyone so it was treated as a weed then it might not have survived over 800 years until Lea and Perrin rediscovered a recipe they had previously hated. The act of not disappearing is a kind of progress because it increases the chance of interaction in the future.

Large portions of the history of Shrewsbury sauce are just a lack of progress, but are thankfully a lack of regress. That stability is important: things cannot interact with other things if the two things cannot come into contact. If an old thing is created hundreds or thousands of years before a new thing — for example, flour was first cultivated nearly 5000 years before Lea and Perrin got together — someone needs to keep the old thing going until the new one comes along. Preventing regress can be as important to technological advancement as encouraging progress.

This is perhaps reassuring given the slump in productivity growth we’re in. Computers have brought amazing things, but it’s not clear if we are more productive. Even if the mismeasurement hypothesis (this is the idea that the productivity gains of computing can’t be seen in the traditional productivity figures) is true a lack of progress is not the worst outcome. Regress is the worst. Innovation hasn’t stopped forever but it’s important to preserve in some form the ideas and innovations of the past so when a traveller from another domain (place, subject, industry etc) visits us, our ideas can combine into a new recipe/technology/thing.

Lack of innovation is a short term problem; the loss of existing innovation (knowledge, ideas, capabilities) is a long term problem.

There is no end: Technologies, flavours and trends are all local maxima

Shrewsbury Sauce is a bit of a distraction from my main point. The fact that I chose a single recipe is a distraction from my main point. I could have told the same story about Lea and Perrin’s Worcester Sauce. Or about grinding mustard.

The story is about how any technology can come to exist, what it requires and how we might look at current activities to predict the future.

Let’s find a fact in all this waffle:

The fundamental fact of progress:
A new technology comes into being by combining old ideas in new ways, which means the old ideas have to end up in the same place and time where the new technology is invented.

This might be enabled by lack of regress (people keep using wheat, people use roman numerals etc) or inevitable spread (people adopt farming because it appears to be better than the hunter-gatherer lifestyle) or by active progress (someone invents it), but the idea has to exist and has to survive.

Once at the place of invention, someone has to actively recombine the old ideas (ingredients) in a new way (recipe) to make a new thing (food). Only then, after all those dependencies, can a new thing exist.

But the new thing isn’t the end. And Shrewsbury sauce isn’t the end. Nor were any of the ingredients before it the end. The story I told concludes with Shrewsbury Sauce making it feel like we’ve concluded and achieved everything that the ingredients were designed to achieve. But all technology progress is local maxima which makes us think that everything that has happened before was meant to happen and is the peak of our possible achievements. Neither is true.

The same story gets told about all technologies: steam, combustion engine, apps, IoT. Once we (humans) understand a technology we turn it into an industry and we double down on understanding the dynamics of the industry (consumer demands, cycles, supply chains etc). But this is a distraction because just as a new technology is turning into a new industry, so someone, somewhere is recombining another set of ingredients to create a totally new recipe and a totally new industry.

Every technology or flavour or trend is a local maxima. You tell the story of one technology, such as Shrewsbury sauce and history makes sense in terms of that thing but this is a trick of stories not a fact of the world; any narrative about the history of a particular technology (for example, a food) puts that technology at the centre and is misleading exactly because of that. This is the narrative fallacy: just because the facts weave together into a story doesn’t mean that there is causality, dependence of anything; a story is just a story.

The technology is a point in a chronological web of interactions and experimentation. Progress which is likely to happen will just happen — this is usually the spread of a technology rather than the invention of one, although spreading requires enabling technologies (shipping, etc).

The road to — and from — Shrewsbury Sauce

I have nothing to conclude about Shrewsbury sauce other than you should learn to make it well or you should come to my house (with the ingredients) and I’ll cook it for you.

I have a lot to conclude about the invention of new technologies.

Invention and progress is made by doing something differently, by fusing ideas and by not regressing. From these simple facts we can extract some rules for maintaining an innovative mind or culture:

• To do something differently — so be experimenting
• To be joining ideas — so have access to knowledge and ideas from other fields
• To be not regressing — therefore maintain knowledge and ideas

The first two are fairly well known, although both would be the subject of much longer essays. Experimenting takes a mindset, a specific business and financial approach and a cultural habit which is much easier to lose than to maintain. The entropy of life is away from invention, not towards it.

Something I didn’t expect to come out of investigating the technology of Shrewsbury sauce is the importance of lack of regress. Active progress — i.e. invention — is fairly obvious; we know we need inventors like Edison, Musk. We also know we need the spread of new and good ideas, like we’ve had the spread of language, maths, art and ideas.

But lack of regress? It’s obvious that nobody could have made anything using flour if wheat cultivation hadn’t survived to the 18th century. But it isn’t a given that wheat cultivation should have survived.

The broader idea here is that you — in your life, business, wider society — have to sustain ideas and knowledge because it may be useful. Ours is an age of easy access to information which can make it feel like all ideas will be around forever. But this isn’t inevitable. Digital rot means that software (ideas) which worked 20 years ago don’t work today. There may be obscure piles of data which are of no interest now, but huge importance in 30 years time.

Progress, however you measure it, is embodied in the artefacts we create but the route to the artefacts is impossible to reverse engineer. In computer science terms, it’s a one-way hashing function. In non-computer terms, you couldn’t reverse-engineer the history of Shrewsbury sauce it just from the taste.

But that’s what we all want to do: work out how to create a new amazing taste. The key is not necessarily to understand how one recipe was made, but to understand and cultivate an environment where ideas are created and sustained so they can recombine in future.

In your own life and work, you could see the lack of regress as maintaining skills and knowledge that you might have moved on from but could be useful in future. Keeping up with the swimming or running. Painting again. Writing again. Cooking Shrewsbury sauce again.

And if you haven’t cooked it before, you should because you’ll learn a lot.

Roast Lamb and Shrewsbury Sauce — Ingredients

From Delia Smith Online.

The lamb

• 1 leg of lamb, weighing 5 lb (2.25 kg)
• 1 small onion, peeled and sliced
• a few sprigs fresh rosemary, to garnish
• salt and freshly milled black pepper

For the Shrewsbury sauce

• 2 level tablespoons plain flour
• 1 heaped teaspoon mustard powder
• 1 pint (570 ml) Beaujolais or other light red wine
• 5 rounded tablespoons good-quality redcurrant jelly, such as Tiptree
• 3 tablespoons Worcestershire sauce
• juice 1 lemon
• salt and freshly milled black pepper