I don't even see why this is an argument---it's so completely ridiculous to have recipes in an encyclopedia, barring some famous ones, that I'm baffled people are actually seriously defending the idea.

-Mark

The problem I have is that some people are enunciating as a general principle that encyclopedias should be "descriptive, not proscriptive."

I don't understand that and I believe it runs contrary to the historical concept of an encyclopedia.

It's a question of degree. You, for example, acknowledge the legitimacy of including "some famous ones." I'd be interested in knowing which ones you feel qualify.

Personally, I'd certainly expect "Oysters Rockefeller" to be in an encyclopedia and to include some discussion of whether it does or does not contain spinach, and give at least one representative recipe. Quite possibly one from the "spinach" camp and one from the "no spinach" camp.

In the case of food dishes that are of obvious cultural importance, representative recipes are relevant. Again, as the level of depth of an article increases, we would begin with general principles (hollandaise sauce is made from butter, egg yolks, and lemon juice or vinegar) but obviously an article that gave the ingredients without explaining how you turn them into hollandaise sauce is not covering the topic very fully.

We don't need the recipe for Aunt Nettie's famous Marshmallow-Anchovy Surprise Omelette that she always used to bring to the church potluck, but it should be OK to say more about meatloaf than that it "is made with seasoned ground meat (usually ground beef or a combination of ground beef and other ground meats), which is formed into loaf shape and baked."

Consider the passage from the Encyclopedia Britannica, 11th edition which I will present later below. The whole article about the telegraph is full of detailed circuit diagrams and mechanical diagrams at a level which goes _far_ beyond the schematic illustration of basic principles. Much of the Britannica 11th is at what might be called the engineering-handbook level of depth and detail. From the information given in "Telegraph" article, you could almost build a "Varley's Double Cup insulator," or wire up a sounder for duplex working by either the differential method or the bridge method.

In recent decades, encyclopedias in the United States have been marketed to parents as a means of giving their high-school students an "edge" in school, and the level of detail and maturity of discourse has accordingly been dumbed down. The 11th edition of the Encyclopedia Britannica was clearly written by highly educated adults for highly educated adults. The World Book seems to be written a high-school textbook level; the current Britannica at, at best, a college level. There is little of the "engineering-handbook-level" detail in it.

Like the Britannica 11th, Wikipedia is written (mostly) for adults by adults. And, since Wikipedia is not paper, There should be no artificial restrictions on the level of detail and depth in an article. In many topics, once we get beyond an introduction with schematic diagrams and general principles, we will naturally get into specifics of actual practices.

My copy of the "Handbook of Chemistry and Physics" has recipes in it:

REMOVING CARBON DEPOSITS FROM FLASKS

First rinse flask with acetone or carbon disulfide to remove traces or tar. Add a few grams of magnesium nitrate. Heat gradually over a free Bunsen flame until water is all expelled and magnesium nitrate melts. Rotate the flask to distribute the melt and continue the heating till the brown fumes of nitric oxide cease to evolve. Finally cool and dissolve the residual magnesium oxide in dilute acid by boiling.

That's from an extensive section entitled "Laboratory Arts and Recipes." If that's not a "how-to," I don't know what it is.

Now, here's a tiny part of what the Britannica 11th has to say about telegraph operation. Descriptive or proscriptive?

In order to maintain a system of telegraph lines in good working condition, daily tests are essential. In the British Postal Telegraph Department all the most important wires are tested every morning between 7:30 and 7:45 A.M. in sections of about 200 miles. The method adopted consists in looping the wires in pairs between two testing offices, A and B (fig. 4); a current is sent from a battery, E, through one oil of a galvanometer, g, through a high resistance, r, through one of the wires, I, and thence back from office B (at which the wires are looped), through wire 2, through another high resistance, r, through a second coil on the galvanometer, g, and thence to earth. If the looped lines are both in good condition and free from leakage, the current sent out on line I will be exactly equal to the current received back on line 2; and as these currefits will have equal but opposite effects on, the galvanometer needle, no deflection of the latter will be produced. If, however, there is leakage, the current received on the galvanometer will be less than the current sent out, and the result will be a deflection of the needle proportional to the amount of leakage.

The galvanometer being so adjusted that a current of definite strength through one of the coils gives a definite deflection of the needle, the amount of leakage expressed in terms of the insulation resistance of the wires is given by the formula

Total insulation resistance of looped lines = 3/4R(D/d 3/4);

in which R is the total resistance of the looped wires, including the resistance of the two coils of the galvanometer, of the battery, and of the two resistance coils r and r (inserted for the purpose of causing the leakage on the lines to have a maximum effect on the galvanometer deflections). In practice the resistances r, r are of Io,000 ohms each. The deflection observed on the galvanorneter when the lines are leaky is d, while D is the deflection obtained through one coil of the galvanometer with all the other resistances in circuit; anu assuming that no leakage exists on the lines, this deflection is calculated from the constant of the instrument, i.e., from the known deflection obtained with a definite current. For the purpose of avoiding calculation, tables are provided showing the values of the total insulation according to the formula, corresponding to various values of d. If the insulation per mile, i.e.; the total insulation multiplied by the mileage of the wire loop, is found to be less than 200,000 ohms, the wire is considered to be faulty. The climatic conditions in the British Islands are such that it is not possible to maintain, in unfavourable weather, a higher standard than that named, which is the insulation obtained when all the insulators are in perfect condition and only the normal leakage, dtie to moisture, is present.

There are three kinds of primary batteries in general use in the British Postal Telegraph Department, viz., the Daniell, the bichromate, and the Leclanch. The Daniell Batteries, type consists of a teak trough divided into five cells by slate partitions coated with marine glue. Each cell contains a zinc plate, immersed in a solution of zinc sulphate, and also a porous chamber containing crystals of copper sulphate and a copper plate. The electromotive force of each cell is 1.7 volts and the resistance 3 ohms. The Fuller bichromate battery... (etc. etc.)

-- Daniel P. B. Smith, dpbsmith@verizon.net alternate: dpbsmith@alum.mit.edu "Elinor Goulding Smith's Great Big Messy Book" is now back in print! Sample chapter at http://world.std.com/~dpbsmith/messy.html Buy it at http://www.amazon.com/exec/obidos/ASIN/1403314063/