The terms and distinctions Mises settles upon are as follows.
First there is the distinction between money and money substitutes. Claims on money are not considered to be money themselves, but rather are termed money substitutes. An important qualification is that a claim on money is only considered a substitute for that money if it is (a) perfectly liquid (no significant expense must be incurred for the claim to be redeemed), (b) perfectly secure, and (c) payable on demand (that is, it doesn't need to "mature" like, for example, a Treasury bill or a Certificate of Deposit). If the claim does not satisfy the above criteria, then it is not a money substitute. If it is used as a general medium of exchange, then it, by definition, is money proper. But it is obviously not the same kind of money as that which it is redeemed for. It circulates as its own kind of money.
Thus token money is, by definition, a money substitute and not money proper. To conceptually lump a token money with specie as money proper, simply due to its metallic content would be to think as a numismatist, and not as an economist. What matters for economics is the thing's role in market exchanges.
Mises subdivides money proper into three categories:
We may give the name commodity money to that sort of money that is at the same time a commercial commodity; and the name fiat money to money that comprises things with a special legal qualification. A third category may be called credit money, this being that sort of money which constitutes a claim against any physical or legal person.Now a claim on money can only be credit money if it is an "imperfect" claim against other kinds of money. Because, again, if the claim is (a) perfectly liquid, (b) perfectly secure, and (c) payable on demand, then it qualifies as a money substitute, and not as money proper.
The important distinction between commodity and fiat money is that the latter is technologically indistinguishable from things that are not money. A note of non-redeemable fiat paper may have the exact same chemical make-up as a bill of board game play money. It is the legal status of each that is the decisive difference between the two. Gold specie, on the other hand, is money by virtue of being gold. It has its own use value (although no "intrinsic" value as is so often claimed, even by fans of Austrian economics); a use value that was decisive in it initially acquiring exchange value.
A school of monetary theorists which Mises refers to here as "nominalists" (and elsewhere as chartalists and etatists) would have it that all money is fiat money. Moreover, they say the "money nature" of fiat money arises immediately from the state declaring that it is money. Mises demonstrates with references to documentary evidence, that historically money has in general been exchanged by weight and quality, and not by "tale" (denomination), so long as any relative debasement has been detected. Therefore, most money throughout history has been commodity money.
Furthermore, even money that is truly fiat money is not established as money immediately upon state pronouncement. To think otherwise would be to think as a jurist, and not as an economist. What makes a thing a money is that it is actually used as a general medium of exchange by individuals on the market. Legal tender laws may be the deciding factor in convincing individuals to use the government's paper as money, but it is the use that makes it money, and not the law.