Bread "starters", some preliminary successes and failures

After a few false starts (pun intended) I've had a little luck in using a so-called starter, rather than packaged yeast, to make bread from scratch. The starter is made by mixing flour or some other source of starch with water or milk and, depending on the variety of starter, other ingredients such as salt or sugar. The mixture is allowed to incubate for up to a week depending on the recipe, usually at somewhat elevated temperature (80° to 85° F is typical) although there are room-temperature starters. The idea is to provide a medium for naturally occurring yeasts or bacteria to grow and ferment. Some starter recipes do add yeast at the beginning to kickstart the process, but the idea is that you'll only have to do this once. For, once the starter is ready and fermentation established, you can use part of the starter and then propagate the rest of it with periodic additions of fresh liquid, flour, or whatever else is required. In essence a bread starter is like a stock culture in microbiology, kept continuously alive with periodic refreshment of the growth medium. Some bakers claim to have maintained a single batch of starter for decades, even centuries. Starters can also be refrigerated or even frozen, though before use the cold starter needs to be warmed and then allowed to stand for several hours or up to a day until it shows signs of life.

The most commonly used bread starter is the sourdough starter, which relies upon both yeast and Lactobacilli (and related bacteria). The yeast provides the lift as with ordinary bread; the Lactobacilli produce lactic acid, which gives sourdough its characteristic tang. So far I have failed to produce a working sourdough starter. Another rare and wonderful bread starter is used to make salt-rising bread, in which the leavening is thought to be provided by Clostridium perfringens, also the organism responsible for gas gangrene as it happens. C. perfringens produces a complicated mixture of fermentation products, including among other things propionic and butyric acids, which give salt-rising bread a unique cheesy smell and taste. After several failures I successfully baked a salt-rising bread and it was good, but the recipe is a tricky one--for one thing, the starter has to be maintained for a long time at a fairly high temperature to suppress yeast formation--and at least in all the recipes I've encountered the starter is not propagated.
My successes have come from a yeast-boosted recipe in Bernard Clayton's The Complete Book of Breads (link, though this is to a newer edition than I'm using) called a "Honey Starter" though I did not use honey. The ingredients:

1 package dry yeast
2½ cups lukewarm water
2 tbsp brown sugar
2½ cups all-purpose flour

These are simply mixed together in a container and allowed to sit, covered, at room temperature for five days with occasional stirring. Periodically the cover must be loosened to allow gases to escape. This is not unfortunately a sourdough starter and the active organism is probably simple yeast. The mixture becomes rather frothy for a time before subsiding, and develops a strong alcoholic smell; mine smelled somewhat fruity, too, like a cheap wine. The starter is replenished with 1 cup each of warm water and flour stirred with 1 cup of starter left over from using the rest--although I'm trying a little experiment in altering this formula.

The best way to use a bread starter is by the "sponge" method, which is more time-consuming but more forgiving of a starter that's down on power, so to speak. The starter is mixed in a large bowl with all the liquids called for in the recipe as well as the salt, sugar and so forth, but then only part of the total amount of flour called for is stirred in, enough to produce a thick but still fluid batter. I have found this usually takes somewhat less than half of the total amount--say, 1½ cups of flour if the total amount is 4 cups. Then this mixture is covered with a tea-towel and incubated in the usual "warm spot". (I use my electric often with the light kept on and the heat periodically switched on for perhaps twenty seconds at a time to keep the oven warm.) Slowly, bubbles form in the batter--more slowly if the starter is somewhat anaemic--and eventually it grows light and frothy, becoming the desired "sponge". This might take as little as a half-hour with a really lively culture but an old, weak one might take two hours or more. Eventually the sponge grows too light to support its own weight, whereupon the top of the mixture collapses into the bowl. The sponge must be periodically inspected until this collapse is observed. Then the mixture is removed from its warm spot and the rest of the flour slowly added and kneaded in until an elastic dough is formed that can be handled without too much sticking. For this I use a stand mixer with a bread hook rather than kneading by hand.

I've had decent luck in using the "Honey Starter" in bread recipes that call for ordinary dry yeast. One cup of the starter is substituted for one packet of yeast and whatever liquid was used to bloom the yeast at the start. I have done especially well with English muffins using the recipe from an older edition of The Joy of Cooking. I'm disappointed however that the starter did not produce a sourdough flavor; hence I'm propagating the current batch using 1 cup of milk instead of water. Milk is a more congenial medium for the Lactobacilli desired.