How Google Recipe Search ruins food: hoisin sauce edition.

Because I’m a food idiot, I like making sauces from scratch. Why not hoisin sauce? With a general knowledge of what’s in hoisin sauce (it’s soybean paste-based and has historically followed Cantonese migrants around South East Asia), I Googled “hoisin sauce recipe”.

Here’s what was served up: 6 out of 9 based on peanut butter, one based on black bean paste, 2 with a soybean paste base.

hoisin sauce is not peanut butter

A special mention/middle finger to Bon Appetit, because here’s their hoisin horrorshow which seems to define hoisin sauce as something vaguely brown and ahistoric. White supremacist food is putting sriracha in something and calling it by whatever Asian name you feel like.

hoisin sauce is not peanut butter

If you didn’t know what you were doing in advance and had never read the ingredients on your bottle of Lee Kum Kee, you’d assume from Google that hoisin sauce was a sweetened, lightly spiced peanut butter. In recipes, good SEO beats food knowledge and Google has set up a perverse incentive for big sites to write bad recipes. If you want a site to be on the first page of Google for “hoisin sauce recipe” you’re probably going to look at the top five existing recipes and copy them rather than questioning whether any of them are good. As Google builds “better” machine intelligence and semantic search, it will assume hoisin sauce is made from peanut butter, because most of the websites say it is, even “reputable” ones like BBC or Bon Appetit.

It becomes a virtuous cycle of shit.

Maybe Anglo hoisin (spiced peanut butter) is metastasizing into different sauce from Cantonese hoisin (sweetened soybean paste) and Vietnamese hoisin/tương đen (sweetened soybean paste with MSG and caramel color (150c)) and Google reflects both semantic drift between languages and changing national tastes. Cookbook author and blogger Andrea Nguyen made the observation in 2010: that tương đen/tương an phở and Cantonese hoisin are clearly different sauces, even when they’re made by the same manufacturer (Lee Kum Kee) for presumably, the same market.

Lee Kum Kee, probably the world’s biggest manufacturer of hoisin, says the origins of the sauce are Hakka (Kaixuan sauce?/”凯旋酱”?) rather than Cantonese, so it is probably a sauce with a lot of genocidal history to unpack.

Not an economic analysis of food trends

Thomas the Think Engine takes on an economic analysis of food trends and the growth in American barbecue in Melbourne, and it’s really quite wrong.

The whole city is suddenly buzzing with American cuisine – and just a few short years ago, that would have seemed like an oxymoron.

The reason is one restaurateurs almost grasp.

“Alabama-born, Dallas-raised Jeremy Sutphin, chef at Le Bon Ton, attributes it to adventure and awareness. ”I’ve been here eight years and the palates are searching for something different – and people are becoming more aware.” “

He’s right about that awareness. Australia’s knowledge of America is now a lot deeper and wider – we’ve now been to America enough that we’ve ventured beyond LA and New York.

He draws a link between travel to different countries and the perception of increased interest in their food. The problem is that the food trends that get written about in the Australian food press from Broadsheet to Epicure bear absolutely no relationship to how the vast majority of Australians eat in restaurants. They bear something of a relationship to how a minority of inner city urbanites eat in the short term, but even then, they’re a terrible guide. Claire from Melbourne Gastronome and I have had a running joke that every year since 2004 someone in Epicure has announced that this will be the year of Peruvian food, but that never happens. I’m still waiting for my plate of delicioso cuy con papas.

Actual food trends are long term and driven by a huge number of factors. If it was as easy as tracking overseas departures, I’d be rich after my investment in an L&P distribution deal. New Zealand is Australia’s biggest destination for short term departures but it’s still pretty tough to get a paua fritter in Melbourne. There probably is a link between Australian travel and interest in foreign food but it isn’t a sufficient condition for it to become popular in Australia.

Here’s a better representation of Australian restaurant trends in Google search data: searches in Australia for different national cuisines in the Restaurants category of Google.

Italian is still dominant with Thai breaking away from Indian and Chinese in mid-2005. Interest in American food has stayed relatively static with some growth in interest since 2011, but not nearly as much as the hype suggests.

For another confirmation of the difference in scale, Urbanspoon lists 1228 Italian restaurants in Melbourne and 131 American restaurants excluding McDonalds, Hungry Jacks, KFC, Subway and Pizza Hut (which should probably also go in the Italian column). Including the chain restaurants, there’s 233. American food is really quite marginal.

When food writers talk about food trends, they’re really talking about a game of cultural capital to distinguish themselves and their readers from others, rather than what most people eat or will be eating in the future. Food writers are talking about American food because it distinguishes them from the mass of people who still love a creamy carbonara and Hawaiian pizza from their local Italian joint. The easiest way to predict what food writers will call a trend next is to see which restaurants open within walking distance from their house or office.

Burger Joint Name Generator

Any word that ends with “-am” can be joined to the word “burger” to make the name of a burger joint.

Here are all of them:


When fried chicken went awry

He still puts on a white linen suit every morning, rides in a chauffeured white Cadillac, visits Kentucky Fried Chicken’s white column headquarters and plugs his “finger lickin’ good” chicken around the country.

But Harland D Sanders, everyone’s favourite Kentucky colonel, is disturbed about what has happened to his chicken and to America’s dining habits.

While it sounds like an alternate history from Adbusters archives, it’s from an interview with Sanders in The Milwaukee Journal, 1975. You know something has gone terribly wrong with food when the mascot starts decrying it.

Did McDonalds cause the decline of violence in America?

Violent crime has been on the decline in the US since 1990, and largely, the reasons for the decline have been inexplicable. Steven Leavitt (of Freakonomics fame) and John J. Donohue III argue that around 50% of the reduction in crime is the result of earlier introduction of legalised abortion (PDF). I’ve got a theory – and it is just a theory at this stage – that McDonalds in the US was also a causal factor in the decline of violence.

Over the last decade Cdr Joseph Hibbeln has been researching the link between violence and the consumption of omega 3 fatty acids. From the Guardian:

Over the last century most western countries have undergone a dramatic shift in the composition of their diets in which the omega-3 fatty acids that are essential to the brain have been flooded out by competing omega-6 fatty acids, mainly from industrial oils such as soya, corn, and sunflower. In the US, for example, soya oil accounted for only 0.02% of all calories available in 1909, but by 2000 it accounted for 20%. Americans have gone from eating a fraction of an ounce of soya oil a year to downing 25lbs (11.3kg) per person per year in that period. In the UK, omega-6 fats from oils such as soya, corn, and sunflower accounted for 1% of energy supply in the early 1960s, but by 2000 they were nearly 5%. These omega-6 fatty acids come mainly from industrial frying for takeaways, ready meals and snack foods such as crisps, chips, biscuits, ice-creams and from margarine. Alcohol, meanwhile, depletes omega-3s from the brain.

To test the hypothesis, Hibbeln and his colleagues have mapped the growth in consumption of omega-6 fatty acids from seed oils in 38 countries since the 1960s against the rise in murder rates over the same period. In all cases there is an unnerving match. As omega-6 goes up, so do homicides in a linear progression. Industrial societies where omega-3 consumption has remained high and omega-6 low because people eat fish, such as Japan, have low rates of murder and depression.

[note: link added by me]. Apart from flaxseed oil, canola oil has the lowest ratio of omega 6 to omega 3 of the vegetable oils at about 2:1. Beef tallow has a omega 6 to omega 3 ratio of 6:1. In 1990, when violent crime hit its peak in America, McDonalds stopped using beef tallow in its fryers and switched to (mostly) canola oil – and as far as I know – almost all American fast food chains followed suit. This certainly increased the total intake of omega 3s and the ratio of omega 6 to omega 3 amongst all Americans who eat a french fry-heavy Standard American Diet. Find out more right here kostenlosepornos.live.

Exactly the same trend followed in the UK. McDonalds replaced beef tallow in the mid-90s, and and since then, the UK has seen the number of victims of violent crimes halve.

They’re interesting correlations but I can’t find (or at least, don’t have the time to find) better data to come up with anything approaching causality.

Any economists want to pick up the baton from here? Anywhere that I can get good datasets on per capita canola oil consumption?

Teddy’s Bigger Burger, Hawaii

Teddy's Bigger Burger

My favourite trait in Americans is the lack of fear. It spawns an infectious entrepreneurialism. It tempts them to cook a patty of ground chuck to medium-rare over fire rather than safely char it to a risk-free tasteless puck. The above was hands down my favourite hamburger of 2010, from Teddy’s Bigger Burger in Waikiki, Hawaii.

Teddy’s is a short walk from “the Wall” surf break, just near the Zoo at Waikiki, a right-hand reef break that gets packed with local bodyboarders even in the smallest swells. It is a convenient break to bodyboard: you can just walk to the end of a pier and jump off straight into the midst of the action, catching waves that propel the fearless alongside the concrete jetty. Tourists line up to take photos. It’s the first place that I’ve ever been alongside someone on a paipo board, the wooden precursor to the modern foam bodyboards; a portly, grey-bearded Hawaiian who looked like he was carved from a brown leather banquette with an uncanny knack of picking the finest wave even from the poorest sets, riding a beautiful slice of polished timber.

Teddy’s is so close that my boardshorts were still moist. I could taste sea salt dripping from my holiday stubble.

Squishy bun, a patty that tastes of pure barely-cooked beef, pickle, sliced onion, an in-season tomato and a decorative frill of lettuce. There is no meal better.

Apologies about the photo. It’s rubbish.