De Interpretatione Ch. 2 - "What is a Noun?" - Aristotle

Definition of a Noun

A noun is a sound having meaning established by convention alone but no reference whatever to time, while no part of it has any meaning, considered apart from the whole. Take the proper name 'Good-steed,' for instance. The * steed ' has no meaning apart, as it has in the phrase 'a good steed.' It is necessary to notice, however, that simple nouns differ from composite. While in the case of the former the parts have no meaning at all, in the latter they have a certain meaning but not as apart from the whole. Let us take ' pirate-vessel,' for instance. The 'vessel' has no sense whatever, except as a part of the whole. We have already said that a noun signifies this or that by convention. No sound is by nature a noun: it becomes one, becoming a symbol. Inarticulate noises mean something—for instance, those made by brute beasts. But no noises of that kind are nouns. 

Aristotle, here, defines what a "noun" is. He says that it is "a sound having meaning established by convention alone but no reference whatever to time, while no part of it has any meaning, considered apart from the whole." In other words, a noun is (1) a sound (2) the meaning of which is determined by human agreement, not anything inherent in the sound (3) is not conjugated in time with tense (4) and its parts do not have meaning on their own by anything other than convention. 

Indefinite Nouns and Names
'Not-man'- and the like are not nouns, and I know of no recognized names we can give such expressions as these, which are neither denials nor sentences. Call them (for want of a better) by the name of indefinite nouns, since we use them of all kinds of things, non-existent as well as existing.

Indefinite nouns, and also names, don't have any meaning because they have no relation to something definitive. He uses the example of "not-man" and of "Philonus". What does "not-man" refer to? It could refer to anything that is not a man. What does "Philonus" refer to? It could refer to anyone who happens to be given that name. 

Cases of Nouns

 'Of Philo,' 'to Philo,' and so on are cases of nouns and not nouns. Otherwise we define all these cases as the noun in itself is defined ; but when ' is,' 'was ' or 'will be ' is added, they do not then form propositions, which either are true or are false, as the noun itself always does then. For 'of Philo is' cannot by itself constitute a true or false proposition. Nor yet can 'of Philo is not.'

Also, the cases of any noun are not separate nouns, but simply the ways in which that same noun is used. A declined noun is clearly not another type of noun because even adding the verb to be does not constitute a sentence like the noun does as a subject combined with the verb to be.