The noun in Modern English has only two grammatical categories, number and case. The existence of case appears to be doubtful and has to be carefully analysed.
The Modern English noun certainly has not got the category of grammatical gender, which is to be found, for example, in Russian, French, German and Latin. Not a single noun in Modern English shows any peculiarities in its morphology due to its denoting a male or a female being. Thus, the words husband and wife do not show any difference in their forms due to the peculiarities of their lexical meanings.
Modern English, as most other languages, distinguishes between two numbers, singular and plural.
The essential meaning of singular and plural seems clear enough: the singular number shows that one object is meant, and the plural shows that more than one object is meant. Thus, the opposition is "one – more than one". This holds good for many nouns: table – tables, pupil – pupils, dog – dogs, etc. However, language facts are not always so simple as that. The category of number in English nouns gives rise to several problems which claim special attention.
First of all, it is to be noted that there is some difference between, say, three houses and three hours. Whereas three houses are three separate objects existing side by side, three hours are a continuous period of time measured by a certain agreed unit of duration. The same, of course, would apply to such expressions as three miles, three acres, etc.
If we now turn to such plurals as waters (e.g. the waters of the Atlantic), or snows (e.g. "A Daughter of the Snows", the title of a story by Jack London), we shall see that we are drifting further away from the original meaning of the plural number. In the first place, no numeral could be used with nouns of this kind. We could not possibly say three waters, or three snows. We cannot say how many waters we mean when we use this noun in the plural number. What, then, is the real difference in meaning between water and waters, snow and snows, etc.? It is fairly obvious that the plural form in every case serves to denote a vast stretch of water (e.g. an ocean), or of snow, or rather of ground covered by snow (e.g. in the arctic regions of Canada), etc. In the case of water and waters we can press the point still further and state that the water of the Atlantic refers to its physical or chemical properties (e. g. the water of the Atlantic contains a considerable portion of salt), whereas the waters of the Atlantic refers to a geographical idea: it denotes a seascape and has, as such, a peculiar stylistic value which the water of the Atlantic certainly lacks. So we see that between the singular and the plural an additional difference of meaning has developed.
Now, the difference between the two numbers may increase to such a degree that the plural form develops a completely new meaning which the singular has not got at all. Thus, for example, the plural form colours has the meaning 'banner' which is restricted to the plural (e.g. to serve under the colours of liberty). In a similar manner, the plural attentions has acquired the meaning 'wooing' (pay attentions to a young lady). A considerable amount of examples in point have been collected by O. Jespersen.
Since, in these cases, a difference in lexical meaning develops between the plural and the singular, it is natural to say that the plural form has been lexicalized. It is not our task here to go into details about the specific peculiarities of meaning which may develop in the plural form of a noun. This is a matter of lexicology rather than of grammar. What is essential from the grammatical viewpoint is the very fact that a difference in meaning which is purely grammatical in its origins is apt under certain conditions to be overshadowed by a lexical difference.