>A: It does not contain the texts of books or recipes.
Seems to be a bit useless without it.
Yes, it's "useless" if one wants the The Sifter to be a source of recipes. Instead, the Sifter's goal appears to be a database of meta information[1] (e.g. ingredients, techniques, etc) of recipes and not the full text of recipes themselves.
An analogy might be a website of baseball statistics[2]. Yes, one can also complain that site is "useless" because it doesn't include videos of the actual games. But others can use it to see when black players first appeared, trends of batting averages, trends of pitcher rotations, etc. A researcher wouldn't need the actual films of gameplay to answer those types of questions and a database of meta info about baseball actually enables faster lookup.
[1] excerpt:
>The Sifter isn’t a collection of recipes, or a repository of entire texts. Instead, it’s a multilingual database, currently 130,000-items strong, of the ingredients, techniques, authors, and section titles included in more than 5,000 European and U.S. cookbooks. It provides a bird’s-eye view of long-term trends in European and American cuisines, from shifting trade routes and dining habits to culinary fads. Search “cupcakes,” for example, and you’ll find the term may have first popped up in Mrs. Putnam’s Receipt Book And Young Housekeeper’s Assistant,
With that attitude, you’re missing out on the cultural and historical significance of food and ingredients.
Food tells so many stories. Pepper usage could tell you things if you’re combining it with other data. There’s a lot there, with such a simple data point — class, race, ethnicity, technology, famine, publishing trends, economic output, agricultural history, climate, trends in taste — so much!
Food history is not just for academic researchers.
This is a funny argument because I can certainly imagine myself giving baseball statistics as a prime example of huge effort going into collecting “useless” data! (Unless you manage a baseball team.) I don’t see how your question is less interesting than “who had the 43rd highest batting average in 1934?”, which I’m sure there are multiple places to look up.
A database of baseball stats includes game results, so you can tell which teams won, how they did in each particular inning, or against a particular pitcher on a particular day.
Not including recipes themselves is like showing all the individual stats for every player without giving any information about who won the games. There's value there, but it would be much more valuable if the full information was available - not just to the general public, but to researchers as well.
I'm personally interested in what old fast food and junk food used to taste like, for example the recipe to the original Coke drink: https://www.thisamericanlife.org/extras/the-recipe. I wish I could go back and taste a 50's In-N-Out meal.
Lard burgers and fries are really good. I highly recommend checking them out. It may not be the exact taste of a brand from the past, but it will be closer to the taste of a mom and pop burger stand pre 1970s.
You might be interested in the restaurant Next in Chicago. From the Wikipedia:
>Rather than stick with one type of cuisine, Next completely changes its style every few months, focusing on a different time period, parts of the world, or various abstract themes for each "season" of its menu.
Next's very first menu was "Paris 1906" which was based on Escoffier's dishes from that year.
I recently read and was amazed that ”The Art of Cookery made Plain and Easy“ by Hannah Glasse, published in 1747 was a best seller around the English speaking world for over a hundred years.
“We had emancipated ourselves from the sceptre of King George, but that of Hannah Glasse was extended without challenge over our fire-sides and dinner-tables, with a sway far more imperative and absolute"
The more I try to cook, the deeper respect I have for historical cooks. The things that we have for granted now, like thermometers, stoves where one could set the level of heat were unavailable and cooking almost anything besides very simple dishes was quite hard.
If you cook a lot, you'll soon realize that cooking without all the technological advances is actually easier. After some time one develops an instinct of how things should be be (based on color, taste, size, even sound !). For example, if one wants too cook onions, choosing between butter and oil, checking the colour, checking if it's cooked or not, deciding to add some water at some point, or additional stuff to induce better Maillard's reaction, etc. All of that you'll be able to do it without resorting to technology.
That's what I find so gratifying about cooking. After 20 years, I can now make things that taste much better than in the past I'm not always sure of why :-) But lots of practice sure make a difference and it's rewarded. Plus additional bonus : you make many better moments in your life that don't cost much !
But granted, technological advances makes somethings better. For example cooking under vacuum allows you a incredible level of control on what you do plus much healthier result.
>> If you cook a lot, you'll soon realize that cooking without all the technological advances is actually easier.
Another thing it's important to realise is that a cooking recipe, unlike a computer program, is not a set of precise instructions that are necessary and sufficient to produce the same result always, but a template that can be modified at will. For the majority of cases you don't need to create precise chamical conditions to reproduce a certain dish with perfect accuracy -and that's just boring anyway.
Also, despite appearances, people who publish recipes (as books or websites, or what have you) are usually not trying to communicate with you how to cook a dish, because that is something you can easily find out for free (e.g. on wikipedia). Instead, they 're trying to, well, add value - by using an uncommon ingredient or revealing some "secret" of the dish, etc. So there's no reason to try and follow most recipes to the letter, because they're not really meant to be followed at all. The majority are only there to make you buy a book or click on some link etc.
For me the best way to cook is to know what you like to eat. The rest comes on its own. It takes some time and you have to start out with simple, boring stuff like boiling rice or frying eggs, but eventually you get to the point where you taste a dish once and you know what to do to cook it yourself. Without any need for a thermometer or a precision scales!
I think he's talking about more basic technology. For example, cooking eggs in a non-stick pan is significantly easier, and even regular pans have uniform surfaces that heat evenly.
Things like silicone cooking tools, food professors, even just ovens where you can set a temperature are all miracle breakthroughs compared to how cooking was done in the past.
I remember both my grandmothers cooking and I don't remember either of them speaking about food processors or say pressure cookers in terms of "miracles". Both grew up and learned to cook at a time when such niceties where not available. I myself cook most of my food in simple pots and pans with simple utensils no different than what would have been available even a thousand years ago. Ovens are another matter- but even there, the variation between different modern ovens (e.g. how quickly they reach a certain temperature) means that experience with a particular oven is much more important than precision.
So I think "miracle breakthroughs" is an exaggeration. Those are things that are more "nice to have" than "absolutely indispensible".
If I wanted to admit to an exception- I've made mayonnaise by hand and by hand mixer. The former is a pain in the ass and takes a lot longer. Even so, having made it by hand I understand how the ingredients work together so I know what to do with the hand mixer. Which is not something you can know if you've always made it with a mixer. So you're actually losing something with too much technology in cooking, so I think anyway.
> So you're actually losing something with too much technology in cooking, so I think anyway.
Somewhat OT:
You're talking about cooking, but the same is true for many other activities and systems. Our technology level permits us to ignore many details and so we fail to learn the underlying mechanisms, processes, etc. In cooking: learning details of the cooking process, why you can make it with a hand or stand mixer versus by hand and what settings are appropriate in those cases. In driving: automatic transmissions hide the details of the different gears, so people driving in mountains and certain other conditions for the first time don't understand why they'd use a lower gear, whereas someone driving a manual transmission all the time would understand more easily.
Examples can be found from any field, but they all demonstrate a common theme: Sometimes you have to do things by hand to understand what's happening before you can apply the higher technology tooling effectively.
Agreed. My other example is animation. I studied classical animation (in another life, another career...) and then I learned how to do 3d computer animation. I was never very good at 3d animation (I preferred modelling) so I can't brag about it but it's very clear to see who hasn't learned to do it the traditional way first, e.g. when you look at character animations in games vs. animated movies.
>>> I've made mayonnaise by hand and by hand mixer.
Strongly disagree !!! :-) Although I'm a pretty good cook, I fail the mayonnaise 9 times out of 10. So my wife does it. And I can assure you she does that in a short and super quiet way, all by hand. Just softness. It's been like that for years and I'm stall amazed.
Oh and when she bakes, I have a lot of pride in mixing everything by hand : Chantilly, egg yolks and sugar, anything and butter, etc. The harder the better. Dunno why I like to do that. I guess it's a kind of micro sport :-)
I agree that, once you figure it out, the sense based instincts is an easier way to cook. I would argue though, that modern cooking tools make learning those instincts much easier because you can use the tools to fly by instruments, so to speak, so you can get a feel for what looks, smells, sounds, and feels right.
On a smaller timescale this has happened over the last 4-6 decades as well. My parents and grandparents didn't learn to cook with an instant read thermometer, digital scale, microwave, instant pot, sous vide machine, etc. I am not a talented cook but I am decent at RTFM and taking measurements so I can make decent meals some times. It makes cooking in a consistent way very easy compared to the learning curves they would had to have gone through.
One of my "if I ever become insanely rich I'm doing this" ideas is to have a culinary museum. Imagine exhibits of cooking utensils and traditions from around the world, with guest experts who can cook some of the items for you. So it becomes a walk/learn/sample kind of experience.
If I do cook german cuisine, I will usually use Henriette Davidis book, it is detailed yet simple, and unpretentious, which I don't often find in good cookbooks.
Apparently the db even has her book in English, so if people here are interested in cooking german food, I think it is a really good recommendation.
One of my favorite aspects is how she will usually give recommendations about using everything that you might usually think is trash. If you're cooking on a budget, these old cookbooks are actually incredibly helpful.
If you ever want to cook Austrian cuisine, the definitive book is "Die Gute Küche" by Plachutta. If you can find a used old copy, it's pretty much in plain text so it's very easy to run it through OCR translation.
This resource is great, but only focuses on US and European ingredients/techniques -- probably because oral traditions were much more common back in the day in Asia/Africa.
The fact that these western recipes/techniques are more well-documented than eastern techniques is one factor in why western cuisine has historically received a lot more attention in restaurants/cooking schools as well as popular media (we'll set aside colonial biases, etc for now.)
As we use structured data to draw inferences and then guide actions more and more (ML -> AI...) the fact that databases like this skew entirely western will lead to distorted suggestions in applications that, say, aim to generate cooking curricula programmatically to teach cooking (I'm sure y'all can think of better examples than that.)
I wonder if voice technology can help capture oral histories, oral traditions, and allow those to be structured in a way that only written traditions have been in the recent past. There is hope! We'll all be better for it, too, when we can access the histories and traditions (many of which have been lost, or almost lost, at this point) of the whole world, not just a small section of it.
The bulk of the Schlesinger collection was donated by Julia Child from her personal library, and yes, English-language cookbooks covering Asian and and especially African cuisine were scarce before 1940. And Child was definitely a Francophile when it came to food, so her collection was skewed that way.
As someone who works on a program that procedurally generates fictional cultures, including their culinary side, this is a godsend.
It allows me to inspect patterns and the evolution of cuisine over time. This will greatly enhance the accuracy and believeability of the generated cultures.
Very cool that you are generating fictional cultures! This resource only structures Western food cultures (probably because a lot of ancient Eastern & African histories relied on oral traditions rather than written, which has been harder to capture and structure pre-voice tech.) As a result, the patterns you detect and stories you generate will be limited to those Western sources; maybe that could be interesting, but I imagine combining those with other traditions around the world would lead to far more novel fictions.
Have you been able to find good structured data about Asian/African food? If not, it would be cool to figure out a way to start building that sort of database so that your stories draw from a fuller set of information and traditions! I think voice tech and the improvement in transcriptions + gpt3-like tools will help us start capturing and structuring oral histories much more effectively and quickly than ever before.
What a treasure it would have been, recipes from several thousand cookbooks since the turn of the first millennium or older, but alas - there are no recipes here. This may be at best useful for laymen like me as an index for ingredients, prior to getting the actual book from a library or similar. Amazing example of cataloging effort though.
As a person who doesn't cook very well, the value I find in most old cookbooks (and a surprising number of modern cookbooks) is just that: historical. And sometimes the sadness that no one makes certain dishes anymore.
But I can't for the life of me figure something like "Stew until ready. Not long before it's ready add finely chopped garlic" [1] or "Prepare the duck, stuff it and stew until done" [2]
Give me times, and temperatures, and sizes, and weights! And don't start me on the weights :D Cups, and spoons, and whatever else are not valid systems of measure :)
>> Give me times, and temperatures, and sizes, and weights!
The thing is, it's pointless to give such precise instructions because they will only ever work for one time and one time only, the time the recipe was (presumably) tested. After that, everything you do as you carry out the instructions in the recipe with absolute precision, like a little robot chef, will still produce a different result not least because your ingredients are no longer the ingredients used in the original recipe.
So for example, if someone tells you to stew a pound of beef "until ready"- the imprecision matches reality. Your pound of beef will not be the same as the pound of beef of the chef who wrote the recipe. It won't be the same as my pound of beef. It won't be the same as anyone else's pound of beef. And it won't be ready in the same exact time as everyone else's pound of beef. Not even if we're all eating from the same, very large, animal.
Like I say in another comment, a recipe is not a precise set of instructions that leads to a reproducible result, like a computer program or an algorithm. It's a template, a mnemonic, an informal description of how to achieve a result in the ballpark of a certain target.
I think this is one reason why many people have trouble with cooking, even when they're very good at following instructions: because they treat recipes as instructions to be followed, rather than a discussion, a debate if you like, between two cooks, or a source of inspiration.
> Like I say in another comment, a recipe is not a precise set of instructions that leads to a reproducible result, like a computer program or an algorithm. It's a template, a mnemonic, an informal description of how to achieve a result in the ballpark of a certain target.
Which is provably false for the vast majority of recipes. And you need a starting point for a recipe. You can't "grab ingredients in the quantities you desire, and cook them" if you don't know the quantities, or the temperature, or what to look for in a finished product.
> I think this is one reason why many people have trouble with cooking, even when they're very good at following instructions: because they treat recipes as instructions to be followed
The reason people have trouble cooking is because people are not born with an innate ability to cook. And on the contrary I think many more people cook now precisely because the recipes are not vague incantations of ingredients.
Music and painting are not precise either. And yet, you don't expect anyone to become a master improviser from day one.
Another thing to consider is that prior to the late 19th/20th centuries, standardization of ingredients and quantities was not really a thing. Flour came in different densities and gluten contents, depending on what the mill felt like producing. Tomatoes came in different sizes and acidities depending on the soil and weather, and neither stoves nor ovens had effective temperature control. If they had told you to use 50g of something, that wouldn't have meant 50g of whatever the reader had, and it certainly wouldn't mean 50g of what you can get.
Instead, the reader is expected to "fill in the gaps", a much more reasonable expectation when everyone is cooking daily with the same few dozen ingredients because long term preservation isn't a thing yet outside pickling/curing/fermentation.
>> Which is provably false for the vast majority of recipes.
I'm not trying to pick a fight but I'm genuinely curious to see how you would disprove my statement. Like, a formal (counter) proof etc.
Just to clarify, I mean that a recipe might be written as if it was a precise set of instructions, but that's not the proper way to use it. And yes, to be honest, "grab ingredients in the quantities you desire and cook them" is how I cook all my meals. There are very few things I actually measure, for instance I have a little dosimetric spoon that I use for spices, mostly beause I like the spoon though, not because I really need it to figure out how much I need of each spice. Everything else I eyball, because I know what makes sense to cook. Everyone I know who knows how to cook more than basic stuff cooks that way.
That's cooking- not baking, btw. Apparently that requires precision, but I don't really bake.
> I'm not trying to pick a fight but I'm genuinely curious to see how you would disprove my statement. Like, a formal (counter) proof etc.
You've proven it yourself:
> That's cooking- not baking, btw. Apparently that requires precision, but I don't really bake.
How is baking different from cooking that baking requires precision whereas cooking "does not"? It's the same process: raw ingredients are subjected to chemical and thermochemical reactions to achieve a result. Their recipes are in the same cookbooks. They have a significant overlap in ingredients.
Where do you draw a line?
Is Beef Wellington coking or baking? Or do you go by feeling? "Today I'm baking, so I'm going to do a precise beef Wellington recipe"?
What about Shepherd's Pie? Baking is defined as "to cook by dry heat especially in an oven" [1] Exactly what's being done to the shepherd's pie (as is done to any other pies). Precision or "it's not a proper way to use it, you have to wing it"?
Or mince pie? If you have meat in it, then "it's pointless to give such precise instructions", but if you remove meat, then "apparently that requires precision", is that how it works?
> And yes, to be honest, "grab ingredients in the quantities you desire and cook them" is how I cook all my meals. There are very few things I actually measure
I am not you. And many people are not you. I presume that you learned how to cook as a kid from your parents? It means that you did receive quite precise instructions at one point until you learned how things work. And I bet that for any new recipe that you've never tried before you will look for and work with precise instructions.
I'm sorry but what you provide above is not a proof. It's what passes for a "proof" in internet forums. At most it can give you some points in a reddit argument or similar pointless exchange (which I'm glad to note we're not having) but it's really not the same thing as a proof at all. Though I'm always hopeful that on HN users will be a bit more careful with the terminology. To clarify, when you said "proof" I expected a formal proof and I was curious to see if you had one in mind. And I was curious because it would have been very difficult to provide one, given that I didn't make any formal claims.
In any case, I don't even agree that my distinction between baking and cooking is so obviously wrong. If you search the internet for "baking" you will find pages upon pages of results with confectionary items, such as cookies, cakes, bread occasionally and so on- but not dishes like beef Wellignton or sheperd's pie, which are "baked" but not generally refered to as "bakes" or "baking" (not in UK English anyway- we might be disagreeing on the grounds of different interpretations of the same English words).
>> I am not you. And many people are not you. I presume that you learned how to cook as a kid from your parents? It means that you did receive quite precise instructions at one point until you learned how things work. And I bet that for any new recipe that you've never tried before you will look for and work with precise instructions.
Actualy, no, on both counts. I never received "precise instructions" as a child. I sure didn't learn to cook from my parents- like the curse of lycanthropy, the ability to cook a decent meal skipped a generation in my family. I learned to cook by watching my grandmothers and by eating the food they cooked. "Watching" did not include taking careful note of the precise quantities of ingredients used in a dish, just noting that certain ingredients are used (without restrictions on what varieties of those ingredients can be used, or what other ingredients may also be used). Later I used my memory of how a certain dish tasted to reproduce it. Occasionally I helped my grandmothers cook so I had some hands-on experience with specific types of processing of ingredients, which I find is much more troublesome for beginner cooks than knowing precise quantities.
For example, I used to help one of my grandmothers make spanakopita, a Greek spinach pie with filo pastry. That was when I was very small, before I even went to school, so too small to know anything about quantities like kilos etc, let alone remember them now. So I would only really help in the part where the filo pastry leaves are laid out in a tray, handlng them gingerly to avoid ripping them, then oiled with olive oil using a cooking brush (an easy job for a careful and methodical child's hands). Equally, when I wanted to make spanakopita for the first time as an adult I'd just use the memory of eating my yiayia's spanakopita to decide how much to use of spinach, leek, fresh onions and feta -the main ingredients of the filling, again from memory. I never looked up exactly how much e.g. spinach to use- first of all, the spinach I get from the supermarket comes in bags of 1kg and it's obvious that my (yes) baking tray would not take, say, 10kg of spinach.
As to new recipes I've never had before, my strategy is to look up some information about the dish on wikipedia, which never gives precise anything when it comes to dishes, and then see whether I can do something like that at home. Occasionally, yes, I've looked up recipes- but I've never followed one. I look up a recipe to get a feel for relative quantities, not precise, absolute values. But most of the time I don't even do that and use the proportions of ingredients that I feel I'll like.
In general, like I say above, eyballing quantities is part and parcel of knowing how to cook. I'm finding that this is hard to reconcile with some peoples' experience, but this really is the way most people cook- and the way that most people have, traditionally, cooked.
Edit: for the sake of having a honest debate, I confess that there are recipes that I've looked up, specifically recipes for emulsions like bechamel sauce, mayonnaise or a Greek sauce called avgolemono. Those require some precision (though not absolute) because, well, they're emulsions. But those are about 1% of my cooking. If I summarised my approach the other 99% of the time it would be thus: I reverse-engineer the recipes of dishes I've tasted and approximate those of the ones I haven't.
> I'm sorry but what you provide above is not a proof. It's what passes for a "proof" in internet forums.
The proof is empirical, and several-fold:
First, the existence of precise recipes that people use. They work.
Second, your own inability to even define what needs precision and what doesn't. "Meat pie isn't baking because most baking refers to cakes" is laughable at best. As is "well, yes, for sauces you need a more precise recipe, but it's emulsion". As if "emulsion" or "baking" (which you can't even properly define) is somehow different from the rest of cooking.
Third. The "cooking is art" argument doesn't preclude that you can become good at it on precise technique alone. It's not magic. "Until meat is ready" has had a precise temperature measurement Celsius/Fahrenheit for decades. Chemical reactions that cause caramelisation, make emulsions, cause ingredients to tenderise or dry out etc. have been know to decades, can be explained in simple terms, can be observed and measured with instruments that are very commonly available now etc.
But that aside, the most infuriating part of your "argument" is that it's the "draw the rest of the owl" [1] argument. Just because you know how to cook, and you don't use precise recipes, and you draw some ill-defined arbitrary lines between where precision is required and where it's not, this somehow invalidates the experiences and the needs of absolutely everybody else.
A recipe that's vague caters to the self-professed "cooking is art" snobs. A recipe that's precise can and will serve as a foundation for absolutely everyone, regardless of their proficiency with cooking.
This is the same, or was the same for many years in many other areas, by the way. From programming to cycling, from music to martial arts. "It has to come from the heart" has plagued many a hobby. Thankfully, in the age of information it's changing, and people no longer assume that "draw the rest of the owl" is a valid approach.
A favorite quote seen in the introduction of a cookbook
in the 1970s or so,
which unfortunately I have never been able to later attribute:
“The world needs more cooks, not more cookbooks.”
I think this can also apply to some programming environments.
I agree with dinamic 1. Not suitable for a blob on hacker news. It would be cool if it had the text + translations but... Also there are some obvious errors. One item is noted as being from a periodical from the year 4.... Hmm I think not.
This isn't specifically related to this database (since it doesn't contain recipes), but how cool would it be to finetune a GPT model on hundreds of thousands of recipes and have it generate new ones to try.
that is so amazing. A friend found an old Indian cookbook in an obscure language which was 80 years old and realized that even some daily food evolved so much just in the span of a century
Clickbait garbage. First thing I look for...a database of the cookbook recipies. Oh its just a database of cookbooks so basically a crappy version of amazon. Not the actual information in them.
Q: What is not in the Sifter?
A: It does not contain the texts of books or recipes.
Seems to be a bit useless without it.