recipe notation

A method of visually representing recipes

Origins
It strikes me as odd that to date no one has improved on the standard method of organizing and displaying the notation for recipes. What first came to mind was something very similar to music notation – the standard visual representation of events occurring over time capable of indicating a wide range actions. In fact, the design that appeared in my mind as I was falling asleep looked like this:

(click to enlarge)

Next step was really to define the total number of possible actions that occur in cooking. Wash, stir, heat, cut, it seemed like quite a few, but then you must imagine 1) that it is finite and 2) that even if it were more complex than music (and it probably wasn’t) most recipes could still be handled with a basic set and then footnotes could be added for the outside cases.

So what are the benefits? Well, first, music is not as restricted by language. Like mathematics, it travels well because most of the ‘terms’ are visual and exempt from words, local syntax et al. Words that are in musical notation are very limited and can learned quickly. It is certainly less terminology than what is in any complex cookbook – think braise, flambé, sauté, et al.

Second, usability (really this should be first). Recipes are horrible at conveying information. If that great example of Napoleon’s march is a how to display data well (it is able to express time, distance, # of troops dead, outside temperature and some correlation of these things), cooking recipes are presented in a format which actually creates problems for itself.

(click to enlarge)

Why, for example, is there no visual representation of time? You should be able to look at a recipe and see the events over a period – e.g., about when in this two process am I going to be doing the bulk of the cooking. Why is there no obvious visually distinction between types of events, say baking versus marinating. That with the exception of the ingredients list there is nothing other than innumerated narrative text is pretty much absurd. If music was at this stage of notation…well, there would be a whole lot less music.

The third benefit is that we might see things differently and better. At a glance, we might be able to judge a recipe and its complexity. We will probably have a better understanding how we will have to coordinate events over time – “wow, there are a lot of things that are going to be going on in sequence in measures three and four; there’s no way I can be deep frying donuts while that’s going on.” Especially when trying to pull together multiple dishes for a meal, having a visual display of the process change everything.

What to Include

So here’s the first blush and what might be represented:

> Time – we’ve talked about this one, but it’s probably the biggie. Some nice linear format, it seems foolish to walk away from the crutch of the musical stave and its measures.

> Actions – it’s going to be a large collection of terms, but it should be manageable. The question is whether and how to abstract it beyond its current elements. For example, should cutting be removed as a separate function and then use visual ‘adjectives’ to describe types of cutting.

>> Samples: Open/Peel/Wash, Chop/Cut/Tear, Mix/Combine/Stir/Toss, Cover/Pour

> Temperature – this seems like one that can stand on its own. There are some general types based on the type of heat – over, stovetop, deep-fry, open flame, smoke – and then there is temperature in both general (medium-high) and specific terms (425 degrees).

> States of Transformation – Descriptions of ingredients at end states such as smooth, translucent, brown. Often, TIME isn’t the judge of doneness, it’s a visual assessment of the dish in progress.

> Ingredients as themselves – well here’s where I think we can call a spade a spade. Listing ingredients and their measurements is about as straightforward as we could get. We could represent the information differently by the SIZE of the text, but I not sure where that might be valuable. Measurements in Volume (cups, tsp) vs. Weight (pounds) vs. General Measures (a bunch, five) might be distinguished for clarification.

The challenge here will be determining where to stop. Notation will create the most value if it does it adds information AND simplicity to the experience of reading recipes. Symbols made of pictographs would be great, but it’s got to be simple enough (at least at first) that it doesn’t FEEL like a new language. The basics have to be simple to master.

Additionally, this doesn’t necessarily need to be able to articulate EVERY recipe or EVERY preparation technique. It should begin simple. In creating these it probably makes sense to start by trying to document a recipe in process rather than translating one from a text.

The First Draft

Beginning with structure similar to a musical notation scale, I began to layout the recipe for scramble eggs. Here’s what it looked like:

(click to enlarge)

I wanted to see the basic ingredients laid out and show some basic mixing. There was a row at the bottom intended to capture details, but it became evident that it was clearer to put them on the lines.

Trying another recipe – pasta and a sauce – I changed the format a little to see how small differences in the connecting lines might make things clearer.

(click to enlarge)

It was immediately clear that ingredients made more sense at the top and equipment below. That way you were visually always putting the ingredients into receptacles. Also, the issue of events that aren’t time based, but results based – such as boiling water – added another element into the notation. The end of these events (marked by >>) is really when it reaches a state (boiling). The amount of time is irrelevant and varies based on the stove.

The Second Draft

Let’s make it the most basic it can be. Let’s represent the recipe as simply a line with the major events roughly stamped out in time.

In some ways, this is the best one can hope for – it provides additional info with minimal effort. Having one of these at the top of every recipe would provide new perspective on the process over time. As it was well summed up by a friend, this graph covers Sequence and Duration.

The issues that this ‘line graph’ brings up are: What are the Major events? And what should be excluded? How do we deal with events that are on different paths? Here’s a quick redrawing.

Pretty good. We’re getting closer to something that could be used to support a recipe. We’ve snuck in a few more elements:

1. Time is formally represented in numbers as well as visually
2. Changes of state (e.g., > boiling = until boiling) has a method of description

At the same time, we also begin to feel that the creep of too much text. The thing is showing evidence that all the verbiage could quickly become a burden. Does ‘Pasta’ need to be repeated three times? As we add other elements, we should think about symbols or alternatives to words as ways of describing things.

First Pictograms

The first attempt at incorporating pictograms was when I introduced cooking utensils.

(click to enlarge)

This version began to demonstrate how text could be eliminated without losing meaning.

> Utensils, like ingredients, are outlined on the left and listed in order of use – just in case the icons aren’t detailed enough. The icons are now included in the flow and, in fact, are the start points for each of these flows.

> Heat is represented with ^s. Six of them indicating HIGH, one indicating SIMMER.

> The word ADD has been removed in favor of the + symbol.

Two other interesting things also emerged. There is something of a SPINE that develops as the recipe proceeds. The main cooking receptacle is at the center of the recipes flow. Second, there will have to be a TIPS box. A series of bulleted sentences that covers the nuances (e.g., you can test the pasta by biting into it OR after draining don’t let the pasta sit more that two minutes or it will stick together).

More on Pictograms

Getting Lost
Came across a new feature on Mapquest. In the printable version on the map, they've added ICONS for the major types of ACTIONS (left turn, right turn, merge, etc).

(click to enlarge)

Very similar problem to the recipe notation and suggests other attributes of the problem. When you're driving and trying to read directions, you want to be able to see major ACTIONS (LEFT) and have them be OBVIOUSLY DISTINGUISHED from other less important information (e.g., LEFT on elm st). Most directions (either online or those written out by hand) didn't do much beyond capitalizing major ACTIONS. In writing out directions, I would CIRCLE these kinds of words and/or put them as the first word of each event to try and accomplish this. So the shorthand would look like this:

(L) on elm
go two blocks
(R) on maple
(R) on edgewood

Beyond the (L) and (R), one never invests enough time to define any other MAJOR actions. Mapquest has done a good job with a first attempt. LEFT and RIGHT, however, should be more markedly different from each other at a glance -- maybe make it a < instead of a full arrow and place it more to the right of the sign (for the TURN RIGHT icon).

On last similarity between driving directions and cooking directions is the need to be able to find your place in the sequence AT A GLANCE. Somehow the notation system must make it easier to quickly ID you LAST STEP completed, in order to find your NEXT STEP. From here on out, we'll call this the LET'S GET LOST problem.

Direction vs. Technique

In working through this, it has become apparent that there are a few categories of information belnded in to recipies:

Four Categories of Information

1. Ingredients (e.g., 1 medium onion chopped coarsely)
>> Ingredient Name (onion)
>> Ingredient Amount/Size/Measurement (1 medium)
>> Ingredient Preparation (chopped coarsely)

2. Recipe Direction (First, sautee onion in skillet...)

3. Explanation of Cooking Techniques

4. Specific Tips on Preparation

All the types of information in Number 1 is usually contained in the Ingredients List and is separated from the other text in most recipes. Sometimes some of ingredient preparation is mixed into the narative copy. Numbers 2, 3 and 4 are usually intermixed together in the narative section of the recipe.*

Big Conclusion: To be effective, mapping of the recipe must be focuses on Recipe Directions -- the step-by-step events of completing the recipe. The other items must live in separate sections (like the INGREDIENT LIST does in most existing recipes.)

The other three categories require a narrative explanation or levels of detail that cannot be better replaced by symbols or a system. How much information to document and whether there is a standard method for noting this other information is to be determined.

*Sometimes cooking techniques are assumed, making recipes confusing or incomplete for novice cooks -- "braise the chicken pieces". Sometimes these are handled in a separate section of a cookbook.

What kinds of symbols?

Tablature Problems
In trying to develop some pictograms, you have to make an absolute decision 1) how hard or diffcult you want to make creating the final notation, and 2) whether you want the notation to be creatable with special software or drawing skills -- meaning, do you want to use new icons or combinations of existing alpha-symbols (+ - ^ ). Looking at even a very simple guitar tab gives you an idea of both issues:

Tumbling Dice - The Rolling Stones
key: B with capo on the fourth fret ('open G' tuning)

intro
(f#) |--------------------|-------------|
(D#) |-3------------------|-0-----------|
(B) |-2-h4-p2-0-2--------|-0-----------|
(F#) |-----------2-p0-----|-0-----------|
(B) |----------------4-2-|-0-----------|

The 'p' stands for (p)ull which is a technique for manipulating the guitar string. On traditional sheet music, these are represented by nice, intuitive arcs (which also can represent the duration of the pull). But it has to be translated into pure TEXT model to make it available on the web as text (vs an image) and to give everybody the ability to create it.

Acutally, Guitar tab is more detailed than more common chord notation. Because many player will be happy to learn and play the chords of a song (80/20), that's the dominant model from quantity of music score content on the web. Tab is more complex because it illustrates the details of particular notes as well as some technique.

More unusual (and more interesting) is Piano Tab. It is an attempt to convey the basic melody and timing in a TEXT format that will allow it to be easily published on the internet. It is really could be better described as a data format -- a standard way to code that requires decoding at the other end.

Nightswimming by R.E.M. - Intro:

4|-ga--g--|g---|-ga--g--|a---|
4|d----d--|e---|d----d--|d---|
2|g--g-h-3|c--2|g--g-h-3|d---|
b|1-2-3-4-|1234|1-2-3-4-|1234|

-the numbers (1-7) to the left of the tab, indicates in what octave
(starting at the lowest C) to play the notes in that row. When numbers appear elsewhere in a row, it is to indicate a new octave.
-ordinary letters (cdefgah) symbolices the corresponding notes

>>>from pianotabs.net, credit to Tor Undheim - for full details, click here to go to his site<<<

A person probably wouldn't play this tab as they would sheet music, but instead would work out the song from the tab. Maybe overtime someone could get comfortable playing from this notation, but for now I think this is a good representation of a data format use to conveniently transmit infomation.

Big Conclusion: This is a similar challenge with Recipe Notation. It MUST err more on the side of an INTUITIVE model that requires little training to understand an adopt (one of the benefits of the current narrative approach). That probably eliminates the use of a TEXT ONLY based approach (like the tabs above or this) to the notation and requires a pictogram model (like this).

However, that doesn't mean that some use of simple TEXT symbols may be of good us. As you can see in the pictogram model (here) the '+' and '-' can provide good, clear direction without addition visual details.

Towards that end, I think the notation should try for less (existing TEXT symbols) and use more (pictograms) when the symbols just don't cut it -- which will be often.

Symbols List 1

Symbols List 1

In keeping with the last post, here are some initial symbols that many survive in this text format vs. needing to be turned into more complex pictograms.

(click to enlarge)