Warning: The following is a long winded explanation to Darren's question
Starch is generally insoluble in water at room temperature. Because of this, starch in nature is stored in cells as small granules which can be seen under a microscope. Starch granules are quite resistant to penetration by both water and hydrolytic enzymes due to the formation of hydrogen bonds within the same molecule and with other neighbouring molecules. However, these inter- and intra-hydrogen bonds can become weak as the temperature of the suspension is raised. When an aqueous suspension of starch is heated, the hydrogen bonds weaken, water is absorbed, and the starch granules swell. This process is commonly called gelatinization because the solution formed has a gelatinous, highly viscous consistency.
Once in this state the starch molecules can be broken down. Starch molecules are glucose polymers linked together by the alpha-1,4 and alpha-1,6 glucosidic bonds. Depending on the relative location of the bond under attack as counted from the end of the chain, the products of mashing are dextrin, maltotriose, maltose, and glucose, etc. Dextrins are shorter, broken starch segments that form as the result of the random hydrolysis of internal glucosidic bonds. A molecule of maltotriose is formed if the third bond from the end of a starch molecule is cleaved; a molecule of maltose is formed if the point of attack is the second bond; a molecule of glucose results if the bond being cleaved is the terminal one; and so on. The initial step in random depolymerization is the splitting of large chains into various smaller sized segments. The breakdown of large particles drastically reduces the viscosity of gelatinized starch solution, resulting in a process called liquefaction because of the thinning of the solution. The final stages of depolymerization are mainly the formation of mono-, di-, and tri-saccharides. This process is called saccharification, due to the formation of saccharides.
What does all this mean? The principle enzymes active during mashing are alpha and beta amylase. These work by breaking the long strings of glucose molecules in a starch into shorter strings. However these amylases are only capable of breaking alpha-1,4 links. Dextrins, found in crystal malt, also have side branches off the linear chains, which are connected by alpha-1,6 bonds, which can only be broken by another enzyme, dextrinase. Dextrinase is present in malt but is very temperature sensitive, and therefore eliminated in the kilning of even light malts. The result of this is that glucose chains or molecules are broken off the ends of a dextrin until a 1,6 branch is reached, at which point the amylases are unable to break any more bonds. The resulting dextrin is called a limit dextrin because the amylase enzymes cant break it further. Therefore there should be no problem with mashing crystal or dextrin malts because the body and head retention result from limit dextrins which remain even after an extended starch conversion rest.